"What is a wargame" Topic

Bill, you've shucked and jived :))) around the issue of YOUR definition of a "simulation". To date you have failed to provide a working definition of "simulation". How you are using the term is inconsistent with the information I have linked to. I'm sorry you don't like my definition but it then becomes incumbent upon you to provide a better one.

Rich

To date, the only definition I've come up with is that a simulation is based upon computational models. Baring external input, each time the simulation is run the same value is derived. The purpose of this is to vary the computations in order to see how that effects the derived value. The computational models are based up real word systems.

Under this definition, wargaming is not a simulation. The function of a wargame is to provide entertainment. The ramification of this is that more and more detail is not guaranteed to improve the game. It is probably not the direction one wants to go. In fact I would say it is counter productive.

One of the advantages of focusing our attention on the 'game' and not the 'simulation' is that we can then focus on what makes a wargaming entertaining. For one thing, as has been mentioned on TMP, playing a wargame is like watching a movie. When we discuss our games we roll it out as if we were recounting a movie. So I'm going to suggest that to improve our wargaming experience we must find ways to minimize the interruption of the flow of the "movie". More charts, for example, widens the interruption of the "movie" and thereby hurts the flow of the game.

Rich

The last thing I want though is a movie. I want system that examines things with true to life data. Mechanism that model true to life situations. Games that ignore this and have very skewed systems intended to "simplify" maters complicate the game in other ways. The whole chart thing is such a sticking point with me. Its really not that big of a interruption to read a few lines on a page. So much of the simplicity stuff is some attempt to make a miniatures game a computer game or a RPG. It isn't ever going to work. Even the most diluted miniatures game I can think of, Axis and Allies miniatures, has charts in the rule book that need referencing.

I am not intending to be abrasive or flippant. I really think though allot of guys are in the wrong direction for what they want. I think they should try a RPG that uses miniatures to play out battles. RPG's are the games that can give that being in a movie experience. Wargames are based on data and constants that must be displayed to the player. Charts may seem dry to some but they work and always have. These gimmicky games that try to work around the old standard always come up short. I don't measure success on revenue gained. Take the Pinto for example, if Ford sold a billion of them they were still a design "SNAFU"

PC simulator games are not intended to be exact simulations. Whereas NASA shuttle simulator is. That's the difference between a PC game called a simulator and a true simulation.

Supergrover6868:
Having talked with both government [USAF and NASA] and PC simulator designers, and having watched those two groups trade systems, taking games and making them into 'true simulations' while PC simulator game designers lift entire programs and make them into games, I can't agree.

Simulations are 'true' when they do what they are designed to to do, simulating whatever aspects of reality were selected.

Sure the true simulators cost big money compared to PC game budgets but the intent is not to be that exact in the world of entertainment. Wargames we play no matter how detailed or complex are not simulations. The term tends to turn off players to the game and is used in many cases disparagingly to describe a game a player feels is to complicated and/or detailed. I think more new people would come to the hobby if the term simulation wasn't used.

Money doesn't determine what is a true simulation and what isn't. I mean, what dollar amount tips to scales?

A simulation only has to be as exact as necessary, and that depends on what it is supposed to do. Whether a simulation has a tolerance of .0000000001 or .1, it is still a simulation. As for the term 'simulation', I think lots of folks would come to the hobby if:

1. Designers and gamers knew what the hell it meant and provided it in their games.
2. The problem is that designers are putting that label on their games NOW. Why? Because it leads gamers to buy them. If it didn't, you'd see that word disappear.
3. The word 'simulation' hasn't hurt the computer industry with flight simulators etc. It's brought in alot of business. Yet they haven't provided some of the aspects of war that tabletop wargaming can.

Rick wrote:
Bill, you've shucked and jived :))) around the issue of YOUR definition of a "simulation". To date you have failed to provide a working definition of "simulation". How you are using the term is inconsistent with the information I have linked to. I'm sorry you don't like my definition but it then becomes incumbent upon you to provide a better one.

Rick:
Have I? How so? I gave you my definitions ages ago, and you didn't accept them. That's why we went with yours, remember? My definitions of accurate can be and are applied to any number of art forms and disciplines. It works for your definition as well as mine.

To review, my definition is:
A simulation is any system designed to mimic some aspect of reality--and succeeds. It is an artificial construct/model designed to either recreate an event [A static simulation] or an environment [dynamic simulation] Any number of mediums and mechanics can be used to create such a model, and like most functional entities, it can be wheels within wheels, models within a model.

Many games and wargames are simulations. As any game is an artificial environment with it's own internal logic, rules and goals, it can be both a simulation and a game.

To date, the only definition I've come up with is that a simulation is based upon computational models. Baring external input, each time the simulation is run the same value is derived. The purpose of this is to vary the computations in order to see how that effects the derived value. The computational models are based up real word [sp? world] systems.

I think you are confusing a computer program with a simulation. They aren't necessarily the same thing. A Dynamic simulation can only derive the same value if all the decisions within the simulation are the same, which is a static/event producing simulation, not a dynamic one. It is like designing the car with a steering wheel, but only letting run on tracks like a train.

Whether the simulation is a comupational model or not, a simulation is be definition a model of something else, presumably in the real world.

Under this definition, wargaming is not a simulation. The function of a wargame is to provide entertainment. The ramification of this is that more and more detail is not guaranteed to improve the game. It is probably not the direction one wants to go. In fact I would say it is counter productive.

The function of may simulations are entertainment too. And more and more detail doesn't guarantee an improved simulation. The very same problems with game system overload is true of simulations. Even the most complicated computer simulations have limits, beyond which more detail destroys its value. So more detail can be just as counter-productive for simulations as games. Regardless of the 'function', too much data, too much detail will wreck the system. That is a given in all simulation design, just as it is in game design.

In the "Handbook of Simulation", Jerry Banks of the Georgia Institute of Technology and an engineer defines simulations:

"Simulation is the imitation of the operation of a real-world process or system over time…Simulation is used to describe and analyse the behavior of a system, ask what-if questions about the real system. Both existing and conceptual systems can be modeled with simulations [i.e. actual systems and opinions/concepts]"

Games are nothing more than puzzles with players looking for optimum play with 'what-if' questions--for entertainment.

Banks says the above on the first pages of the book. One page 4 he points out that "any simulation design has limits to what can be included and achieved. A simulation should do no more than what it is designed to do, with the least amount of data."

Now, this guy designs simulations to model engineering projects and systems, usually with computers. Yet his understanding of what a simulation does is the same for a PC game designer and me, a training game/simulation designer.

One of the advantages of focusing our attention on the 'game' and not the 'simulation' is that we can then focus on what makes a wargaming entertaining. For one thing, as has been mentioned on TMP, playing a wargame is like watching a movie. When we discuss our games we roll it out as if we were recounting a movie. So I'm going to suggest that to improve our wargaming experience we must find ways to minimize the interruption of the flow of the "movie". More charts, for example, widens the interruption of the "movie" and thereby hurts the flow of the game.

Ah, Rick, a movie is a simulation. The term 'flow', which is used by both game designers and simulation designers is an important one. It describes the balance between boredom and too great a challenge, between game or simulation decisions and game or simulation administration.

Defining simulations as totally divorced from games, and then describing the benefits as seeing the game as a movie, I think illustrates the basic problem in trying to insist they have to be different because games are meant to be fun. Games are artificial environments for fun and simulations are artificial environments that model real ones, which can be for research, fun, or training… or whatever. I use games to train adults. They are games, but their function is only partially meant to be fun, they are supposed to instruct too.

When it is fun to learn about history, face similar tactical and strategic challenges as historical general in game form, a simulation can also be a game.

Wargames don't 'have to' be simulations, but considering what most designers claim to be doing in designing the game [recreating Napoleonic battle for instance] they can't avoid being simulations too and still achieve those goals.

Theys are true when they do what the are designed to do. Yes a PC game is designed to entertain not train a man to fly the Aicraft repsented in the program as opposed to military training simulation that is designed to do just that, along with real world flight training,of course.

Many times I hear "simulation" used in war gaming to assert that the game is trying to 100 percent model reality and thus is to complicated to play. I don't believe any war game is. It may be to complicated to be enjoyable for certain persons tastes though.

Whether the term is being used properly or not. It is too often(IMO) used disparagingly to say a game is to complicated. I see that as chasing away potential players.

This is an example of a common problem in English where terms become mutated and mean different thing to different people. The Dictionary changes to accommodate the new use or even creation of words. It ends up usually causing much confusion. A comical example would be the use of the word "Heavy" like in the Movie back to the future where the scientist is confused by the term constantly used by the lead character.

I once thought that setting some industry standards in a glossary would be a great thing but then realized that it probably would never be followed by all gamers and the confusion and debate would still occur. So instead of trying to correct people on there use of a term I try to understand what their definition of it is and go from there.

Under my definition, right or wrong in the eyes of others, Commercial wargaming and PC games don't produce conflict simulations or Flight, Tank, and Naval simulators. The PC game industry does produce a class of program that is categorized as simulators due to the games similarity with actual simulators used by the military to train people on certain systems, or that your in First person view of a vehicle like a race car. The difference being as you mentioned, how much is modeled. In commercial games not everything has to be. The PC industry does use marketing tags like "realistic physics engine" to sell games. But any games that taut this realistic physics have omissions, errors, or other short cuts to keep programing costs down.

The only claims of crossover between Actual Simulators for the Military and PC entertainment I can think of is Falcon 4.0. Where the Spectrum Holobyte claimed they had a discontinued program from the USAF and used the basics of it to create the Game.They claimed certain aspects of the code were blacked out due to Security issues.

I read some blurb about a more detailed version of Steel Beasts that stated it was sold to Military Clients only. Which in my mind removes it's classification as a entertainment title and to a tool for the Military.

Many times I hear "simulation" used in war gaming to assert that the game is trying to 100 percent model reality and thus is to complicated to play. I don't believe any war game is. It may be to complicated to be enjoyable for certain persons tastes though.

Whether the term is being used properly or not. It is too often(IMO) used disparagingly to say a game is too complicated. I see that as chasing away potential players.

Supergrover6868:

I will agree that there is a very persistent belief in wargame circles that simulations have to be complex, and those no fun as games.

In other words, the more detail injected into a simulation, the more accurate/realistic/better it is as a simulation. That is utter nonsense, and any professional simulation designer will tell you the same. The amount of information included in a design is not the defining quality of any simulation.

This one belief keep wargamers and hobby wargame designers from making much sense of wargame design--which is one big reason why simulations is a derogatory term in many quarters.

As far as a simulation goes, I don't care how much money or detail is injected into a design, it still can't model 100 percent of reality. No simulation ever built or ever can be built will do that. That is a simulation's main strength and abolute limit. A 100 percent accurate wargame of WWII would have to include 60 million+ dead etc. etc. The wonderful thing about a simulation is that you can isolate SOME parts of reality and avoid the rest. The point being is that no military or NASA simulation ever even approaches 100 percent accuracy/reality, nor would the thing be of any value if it did.

My son is a engineer at JPL working on the 2011/ was 2009 Mars Rover. Their simulations are great, and I've seen a few, but they are limited, usually on purpose, because complexity can ruin a simulation. It wouldn't do what it is supposed to do.

So, while military simulations may be more complex or cost more in some cases, they are no different in kind than any game simulation, and often are identical.

The only claims of crossover between Actual Simulators for the Military and PC entertainment I can think of is Falcon 4.0. Where the Spectrum Holobyte claimed they had a discontinued program from the USAF and used the basics of it to create the Game.They claimed certain aspects of the code were blacked out due to Security issues.

Actually, there are entire game companies that do nothing but take military and science simulations PC programs and turn them into games, with very little reduction in content--for entertainment--and vice versa.

I read some blurb about a more detailed version of Steel Beasts that stated it was sold to Military Clients only. Which in my mind removes it's classification as a entertainment title and to a tool for the Military.

So a simple sale of a game stops it from being entertaining when used as a tool for the military? A tool is a tool, it doesn't lose its character simply because someone decides to use it for another purpose. I always see von Riesswitz's Kriegspiel as a classic example. It was specifically designed to be a tactical training tool for Prussian officers, but von Riesswitz states in his foreward how surprised he was that officers enjoyed the game and found it entertaining…

If a 'more detailed' version of Steel Beast was sold to the military, that could mean any number of things, including the injection of 'classified' or the ability to inject new material. The military might have wanted a more detailed version, but that doesn't mean all or even most of the military's wargames are more detailed than say F&F or FoW.

Whether a game or simulation is entertaining is an inherent quality of the design, not something turned off and on when someone decides it is, or when the design is used for a different purpose.

Certainly entertainment can be a goal of a game design, but as we know, that doesn't guarantee it will be entertaining. And if that design purpose isn't entertainment, that doesn't keep the result from providing entertainment just the same. Lots of examples of that in the computer game industry.

There is a great deal written about what creates game and simulation entertainment, but those qualities are never defined as exclusively inhabiting games and PC simulation games, while all other kinds of simulation designs or wargames are ipso facto, completely devoid of them.

As long as we believe that entertainment is something that can be defined into a design, or defined out if it by mere use or complexity, then we really aren't going to get to what is entertaining.

Bill:

To review, my definition is:
A simulation is any system designed to mimic some aspect of reality--and succeeds. It is an artificial construct/model designed to either recreate an event [A static simulation] or an environment [dynamic simulation] Any number of mediums and mechanics can be used to create such a model, and like most functional entities, it can be wheels within wheels, models within a model.

Thank you for the definition. My question is how does this differ from the definition of a game. If they are the same then why use the term 'simulation'?

By the way it's Rich not Rick. No big thing, just a niggle.

Rich

Rich:
Pardone me. Just a typo--movin' the fingers too quickly and not re-reading it. They aren't the same, though a design can be both a game and a simulation. Here again are the working definitions I have known and used. They are working definitions because they help designers design… Not because they are the final word on the subject.

I've also provided a few variety of views of games and simulations from professionals that I am sure you can see are very similar, though expressed by folks in very different disciplines over several decades.

A Game is a system, an artificial environment with its own set of rules for a set of decisions towards pre-determined goals. It's a puzzle of choices that lead to winning decisions for the purpose of entertainment or training.

A [dynamic] Simulation is an artificial environment that mimics portions of a real environment where 'what if' alternatives can be explored for the purpose of research, training and skill development, or entertainment. Simulating is mimicking something else.

A Simulation Game is an artificial environment with a set of rules that mimic a real environment or even theories, opinions etc. providing real-world decisions towards pre-determined goals within that simulated environment. [which are part of the reality mimicked.]

A Wargame is an artificial environment with a set of rules that mimic some chosen aspects of war, from the principles of war to simulating historical battles and wars.

The designer determines whether a particular design is a game, a simulation game or a wargame, or all three.

Battle Cry and Memoirs '44 were designed to provide "stylized history". Axis and Allies was designed to "offer some of the strategic decisions of WWII, not the details of history." They are wargames offering some of the principles of war, but none were meant to simulate actual military history.

On the other hand, the designers of Fire and Fury, Empire 1-5, Piquet, Armati, Fields of Glory etc. state that their games were designed to simulate particular historical battlefield environments.

They were all designed to entertain.

"The object of any wargame (historical or otherwise) is to enable the player to recreate a specific event and, more importantly, to be able to explore what might have been if the player decides to do things differently."
--Jim Dunnigan, Chapter 1, "What is a Wargame?"
The Complete Wargame Handbook, 1980, 1993, and 2000

"The Primary Rule Of Wargaming: Nothing may be done contrary to what could or would be done in actual war."
--Fred T. Jane (Of Jane's Fighting Ship's fame)

"A game is a series of interesting decisions."
--Sid Meiers, Designer of Civilization I-IV

"A game is one or more causally linked challenges in a simulated environment"
--Ernest Adams and Andrew Rollings, PC game designers
On Game Design.

"A simulation is the imitation of the operation of a real-world process or system over time."
--Jerry Banks, Handbook of Simulation

"A simulation allows players to safely make real-world decisions and develop skills in an unreal environment."
-David Bartlett, former chief of operations,
Defense Modeling and Simulation Office.

Bill, thanks for the response. You have given me some thing to think about.

Rich

1. Jim Dunnigan, Chapter 1, "What is a Wargame?" 
The Complete Wargame Handbook,
"The object of any wargame (historical or otherwise) is to enable the player to recreate a specific event and, more importantly, to be able to explore what might have been if the player decides to do things differently."

I have no problem with Dunnigan. He is defining a game

2. Fred T. Jane (Of Jane's Fighting Ship's fame)
"The Primary Rule Of Wargaming: Nothing may be done contrary to what could or would be done in actual war."

Again this is accomplished through a game..

3. Sid Meiers, Designer of Civilization I-IV
"A game is a series of interesting decisions."


Ditto

4. Ernest Adams and Andrew Rollings, PC game designers, On Game Design.
"A game is one or more causally linked challenges in a simulated environment"

For PC games computational models are required to obtain a simulated environment.

Except for Adams and Rollings, the quotes fall within the field of gaming. I do not see the need for the term simulation.

5. David Bartlett, former chief of operations, 
Defense Modeling and Simulation Office.
"A simulation allows players to safely make real-world decisions and develop skills in an unreal environment."

Bartlett points out what a simulation allows a player to do. He does not state how a simulation is created.

Let's turn to Jerry Banks. You have quoted him twice. You quote from the Handbook of Simulation:

"Simulation is the imitation of the operation of a real-world process or system over time…Simulation is used to describe and analyse the behavior of a system, ask what-if questions about the real system. Both existing and conceptual systems can be modeled with simulations [i.e. actual systems and opinions/concepts]"

I believe that Banks is thinking in terms of computational models. This is why: He followed the comments above with an example of a computational model of arriving and servicing of bank patrons. He has a vertical axis and a horizontal axis. He has # of customers running down one side and time categories across the top. At the crossing of each axis is a number.

Also notice his example was for the purpose of investigating time responses. It was not designed for entertainment.

Banks continues,

"This small simulation can be accomplished by hand, but there is a limit to the complexity of problems that can be solved in this manner. Also, the number of customers that must be simulated could be much larger than 20 and the number of times that the simulation must be run for statistical purposes could be large. Hence, using the computer to solve real simulation problems is almost always appropriate."

It seems to me Banks clearly implies that simulations are defined in terms of computational modeling.

Now lets look at your definition,

"A simulation is any system designed to mimic some aspect of reality--and succeeds. It is an artificial construct/model designed to either recreate an event [A static simulation] or an environment [dynamic simulation] Any number of mediums and mechanics can be used to create such a model, and like most functional entities, it can be wheels within wheels, models within a model."

You state:

System to mimic some aspect of reality. Dunnigan: recreate a specific event (action over time). I see no functional distinction here between your definition and Dunnigan's. Therefore the term ‘game' would be preferable to simulation because of the computational baggage simulation brings to the discussion.

You state, "mimic some aspect of reality". But you are not very specific about how simulations go about creating these models of reality. You simply say, "Any number of mediums and mechanics can be used ". I've only been able to identify two different mechanics. Simulations use computational models. Wargames use physical models (model hills, trees, soldiers, etc.). What needs to be done is to come up with a third way simulations mimic a portion of "reality".

Thank you for the definitions. I set aside the Game definition since you provided a definition of a wargame.

"A Wargame is an artificial environment with a set of rules that mimic some chosen aspects of war, from the principles of war to simulating historical battles and wars." Dunnigan: to be able to explore what might have been if the player decides to do things differently. [I added Dunnigan's comments to round out your definition.]

A [dynamic] Simulation is an artificial environment that mimics portions of a real environment where 'what if' alternatives can be explored for the purpose of research, training and skill development, or entertainment.

A Simulation Game is an artificial environment with a set of rules that mimic a real environment or even theories, opinions etc. providing real-world decisions towards pre-determined goals within that simulated environment. [which are part of the reality mimicked.]

I'm a bit confused with "providing real-world decisions towards pre-determined goals within that simulated environment". Are you saying the simulation game provides decisions, or, within the simulation one can make various decisions leading to pre-determined goals? If it is the latter, then one is making various decisions that are based on how well they reach the predetermined goals. This would be a variation on the "what if" scenario.

You also wrote, "Simulating is mimicking something else." Therefore I think we can write that out as Simulating = Mimicking = Imitating (at least in my dictionary).

So a Wargame is
* artificial environment
* imitates a portion of the [military] environment
* to explore what if.
* (mine) purpose entertainment

A Dynamic Simulation.
* artificial environment
* imitates a portion of the environment.
* to explore what if.
* purpose research, training, etc.

A Simulation Game
* artificial environment
* imitates a portion of the environment
* to explore what if
* no purpose is provided.

Condensing each definition, in this manner, helps to show there is really no functional difference between your definition of Wargame, Dynamic Simulation, and Simulation game. Therefore ‘simulation' adds nothing to our understanding of wargames that the definition of ‘wargame' doesn't already provide. Given the computational baggage that comes with ‘simulation', use of the term simply muddies the understanding of ‘wargame' rather than clarifies our understanding.

This leaves us without a functional difference between a wargame and a simulation other than a simulation is built upon computational models and a wargame is not. What is needed is to show the tools and techniques of a simulation and how they differ from that of a wargame.

Rich

Rich wrote:
So a Wargame is
* artificial environment
* imitates a portion of the [military] environment
* to explore what if.
* (mine) purpose entertainment
A Dynamic Simulation.
* artificial environment
* imitates a portion of the environment.
* to explore what if.
* purpose research, training, etc.
A Simulation Game
* artificial environment
* imitates a portion of the environment
* to explore what if
* no purpose is provided.
Condensing each definition, in this manner, helps to show there is really no functional difference between your definition of Wargame, Dynamic Simulation, and Simulation game. Therefore ‘simulation' adds nothing to our understanding of wargames that the definition of ‘wargame' doesn't already provide. Given the computational baggage that comes with ‘simulation', use of the term simply muddies the understanding of ‘wargame' rather than clarifies our understanding.
This leaves us without a functional difference between a wargame and a simulation other than a simulation is built upon computational models and a wargame is not. What is needed is to show the tools and techniques of a simulation and how they differ from that of a wargame.

Rich:

My point exactly. IF a wargame is meant to simulate a real battlefield environment, a simulation is designed to simulate a real environment, and a game is meant to simulate a real environment, they all are simulations. [i.e. designs whose purpose is to mimic, copy, represent, recreate, imitate etc. etc. some aspects of reality, past or present.]

They are all simulations. They share a common goal and that goal dictates aspects of each structure. One is designed to train for war, one is designed to do research or train, and one is designed for entertainment, they are all still simulations.

That means that the many of the tools and techniques for designing a simulation do not differ--in fact can not differ-- from those of designing a wargame or game as long as each is supposed to mimic a real environment.

*That is why Jim Dunnigan speaks of wargames, but calls his company Simulations Publications, Inc. and designs simulation games for the Military.

*That is why Fred Jane's says ""The Primary Rule Of Wargaming: Nothing may be done contrary to what could or would be done in actual war." He is stating that the Primary rule of wargaming is simulating reality: actual war.

*That is why Sid Mieirs can call his Civilization game a simulation.

*That is why Ernest Adams and Andrew Rollings, game designers, speak of simulated environments when they simply design PC games—and not even 'simulation' games.

*That is why, David Bartlett of the Defense Modeling and Simulation Office speaks of simulations when his office designs wargames of all kinds.

*That is why Jerry Banks' description of a simulation agrees with other, non-engineer and non-research AND non-computer definitions of wargames and simulations.

The Point:
IF you, me and Jerry Banks all decide to design a simulation, one for gaming purposes, one for training purposes, and one for research, and you use table top miniatures, and I use role playing scripts and Banks uses computers, our efforts are still all focused on the same thing: Simulation.

That means there are inherent challenges, design concepts, techniques, and methodologies that all three hold in common because all of us are designing a simulation, regardless of the design purpose, and regardless of the medium [miniatures, RPG scripts, and computers]

That means when Jerry Banks speaks of time as a unique arbitrator of simulations, that applies to ALL simulations. Time, the progress of activities, whether in a game or a computer program, is a core issue for a simulation.
Whether miniatures, scripts, or computational systems, they all share the same simulation challenges, each medium provides unique benefits and problems in simulating—and unique technical requirements, but they all still face the same challenges in simulating, and use the same methods and techniques to solve them.

Simulations come with "computational baggage" only when folks insist that it must be attached. Computers are a way of simulating, period. Not THE way, and not inherently better than other methods depending on what the designer wants to simulate.

And when I say this, I am speaking from experience, not just reading books or throwing together a miniatures game. I have sat with trainers, researchers, game designers at the same table to discuss those common simulation design issues, even though we were all using different mediums to create those simulations, computers, RPG scripts, etc for different purposes.

And while there are many things that are different because of the different mediums, there were a great many similarities among our efforts.[similar: root word of simulation, nach]
That has been the catalyst for a great many advances in simulation design, and why simulation design can be seen as a separate study from computers or games, but applies directly to both…

When I said that simulations should "providing real-world decisions towards pre-determined goals within that simulated environment", it was confusing.

Sid Mieirs said that games are "a series of interesting decisions". I was saying simulation games provide a series of interesting decisions "found in the real world being simulated."

If I designed a simulation of Waterloo, but the only decisions the player could make were related to uniform colors, I don't have a simulation, no matter how well the environmental simulation systems work…

Whether a game or a wargame is a simulation, begins with whether that is the goal of the designer. They are the ones who establish the function differences between game, simulation and wargame in their particular design. Not that our hobby designers ever bother--they like to smear the issue so any design can be whatever fits at the moment. In the end the designer is the arbitrator--or they should be.

A game can have an artificial environment completely divorced from reality, like checkers or poker or Chutes and Ladders.

A wargame, on the other hand, has to simulate something in war to be a wargame, so the only difference is that is is a simulation of war. This connection is strong enough that dictionaries define wargames as simulations.

Once the similarities are recognized and used, a great deal of valuable methodology and techniques, developed over the last 50 years, can be used in our hobby to improve the simulation aspects of our games. That doesn't mean they would be any less fun, simple, or more technical for the player. It does demand that designers actually do what they say they are doing.

Just one instance: A well-known game designer wrote this recently:

"The scientific notion of testability provided an additional motivation – not that I make any claims as to the game being particularly precise, objective, or scientific. All rules sets portray the capabilities of different armies as – unavoidably – the author sees them. But without being able to set up and re-fight entire historical battles, it is nearly impossible to find any sort of objective standard against which to measure the author's judgments."

The sad fact is through simulation development and decades of experience, what he says about 'any sort of objective standard' being 'nearly impossible' is absolutely false.

There are objective, proven measures that have be rigorously tested that will provide those measures without re-fighting entire battles.

Simulation designers are constantly faced with creating simulations that can't have their validity tested against the real world, but still have to accurately mimic it all the same. Lots of ways to do that.

First time I heard about those methods was from an engineer who had to create a multi-million dollar production line with several innovations. Similar systems did not exist, his was the first. If his simulations were wrong, millions would be wasted in producing a system that didn't work.

The same methods can be used on our hobby games to provide those 'objective standards'. It isn't impossible at all.

[Remember the innovative baggage-handling system that didn't work when the Denver Airport opened? The designers decided they didn't need any simulator's help… they were called in to help fix the problem…]

Our hobby designers, like the one quoted above, keep presenting the same problems over and over again as insolvable, when they have been for a long-long time. All it takes is accepting that simulation design is now a discrete disciple, with methodologies applicable to ANY effort to design a simulation, game or not.

If we insist that the definitions for wargame, simulation, and game have to be mutually exclusive, we have defined out of the ability to use what are similar between them… particularly when all three are used to simulate, mimic, recreate etc….

Bill

The Scotsman wrote:

…the more detail injected into a simulation, the more accurate/realistic/better it is as a simulation. That is utter nonsense, and any professional simulation designer will tell you the same.

This is completely true; but sometimes I get depressed at how people who should know better still make the tyro's error of imagining that adding more detail is making a better simulation.

All it takes is accepting that simulation design is now a discrete disciple, with methodologies applicable to ANY effort to design a simulation, game or not.

Also true. I suspect that the reason for the perpetuation of the tyro's error mentioned above is that most simulations are done by people who are no, first and foremost, simulation modellers.

I have to ask, are you a professional simulationist? Or do you just read Jerry Banks for fun?

All the best,

John.

But Bill, I don't think you are being consistent. You presented definitions for Game, Dynamic Simulations, Simulation Games, etc. to distinguish one from the other as guidelines for others. When I point out that the definitions all say the same thing you come back with "my point exactly." But Bill, definitions are meant to define something.

You go on to say "they all are simulations. [i.e. designs whose purpose is to mimic, copy, represent, recreate, imitate etc. etc. some aspects of reality, past or present.]"

Under this definition, model railroading is simulation, painting landscapes is simulation. Making and sailing model boats is simulation. Carving animals is simulation. Building model houses with toothpicks is simulation. They all imitate some aspect of reality. Your definition lacks the one thing all definitions need to do. It fails to define.

That means when Jerry Banks speaks of time as a unique arbitrator of simulations, that applies to ALL simulations. Time, the progress of activities, whether in a game or a computer program, is a core issue for a simulation.

This would make chess a simulation., Monopoly, Life, etc.

"Simulations come with "computational baggage" only when folks insist that it must be attached."

Jim Dunnigan works with computational models. Sid Mieirs works with computational models. Ernest Adams and Andrew Rolling work with computational models. David Bartlett works with computational models. Jerry Banks works with computational models. All these men work with simulations. How can simulations not come with computational baggage? And none of these men are creating hobby wargaming rules.

"A wargame, on the other hand, has to simulate something in war to be a wargame, so the only difference is that is is a simulation of war. This connection is strong enough that dictionaries define wargames as simulations."

1. Here is where I think things get a bit "wookie" [a technical term]. To imitate something does not make that something into a simulation. There must be more to it than that.
2. Modern wargames, with their computational models, are simulations.

"Simulation designers are constantly faced with creating simulations that can't have their validity tested against the real world, but still have to accurately mimic it all the same."

The real world is the validity test. Once built, that production line either works or it doesn't. Wargaming, on the other hand, is based on history. History is not an objective science. I don't care what Ranke intended, history is a subjective enterprise. This means you do not have the real-world objective validation one has with simulations and their computational models.

Bill, what you seem to be doing is creating an overly broad definition of simulation so you can include wargaming within the scope of simulation. Now had you said wargaming could benefit from some the new techniques being developed by simulation designers, I might have gone there with you. I think there are things wargaming, as a game, can learn from developments in simulations. But to make the definition of simulation so broad that it loses its ability to discriminate is not the way I would have gone. That's why I think how gamers and simulators go about their creations is as important as the fact that they are engaged in trying to mimic aspects of "reality".

Rich

Supergrover6868 wrote:
I once thought that setting some industry standards in a glossary would be a great thing but then realized that it probably would never be followed by all gamers and the confusion and debate would still occur. So instead of trying to correct people on there use of a term I try to understand what their definition of it is and go from there.

Supergrover:

I meant to respond to this, but got sidetracked. What I find odd with the wargame hobby is its lack of descriptions for what is the core of the hobby: the wargame. The RC model plane hobby has 'standards' for free flight, scale and true scale planes followed by hobbiests all over the world, and none of them question what RC modeling means…

The same is true of stamp collecting and golf and making making jewelry. Definitions and standards are useful, so they are used. I can't think of a single hobby that is as out-to-sea about what the hobby is about, how the core aspects are technically described as historical wargames.

Can you imagine being on a RC model plane list where the members debate what an RC model plane is, or where a good number of the members would say no one will ever agree on it and it doesn't matter. All that matters is if whatever is done is fun. Or quilt makers, or bridge players…?

Very strange…

John S.:

How embarrassing. That line "All it takes is accepting that simulation design is now a discrete disciple, with methodologies applicable to ANY effort to design a simulation, game or not" should read discipline not disciple…. but you got the idea.

Yes, I have been a professional simulationist? And I read Jerry Banks when it was suggested to me. I was having problems with a simulation design issue and a colleague recommended the book as containing some solutions--which it did.

What is important to know is that:
1. I designed training simulation games that never used computers, mostly RPGs and scenario/matrix designs in training such things as communication skills and management.
2. The colleague that recommended it was a PC simulation designer for building designs and flow patterns.
3. Obviously, Banks was writing about engineering test simulations using PC programs, though other contributors to the book were not.

Actually, while I am always amazed by how universal simulation design is, I can't imagine anyone reading Jerry Banks' book for 'fun' ;-j

Bill

Wargaming is an historical interpretative endeavor. Historical interpretation is an individualistic interpretative endeavor. The interpretation depends on the person doing the interpreting. It would be like telling gamers how they can interpret history.

That isn't to say the hobby doesn't have standards. It's just that the standards are not codified. Nor are they ridged. The standards are our expectations of what a wargame is and what should be included. Game designers must meet these expectations (standards) if the rules are to be accepted.

Rich

Rich

Rich wrote:
But Bill, I don't think you are being consistent. You presented definitions for Game, Dynamic Simulations, Simulation Games, etc. to distinguish one from the other as guidelines for others. When I point out that the definitions all say the same thing you come back with "my point exactly." But Bill, definitions are meant to define something.

Rich:
Well, I certainly have bouts of inconsistency now and then. Howsomever, I didn't define Game, Dynamic Simulations, Simulation Games to distinguish one from the other by offering mutually exclusive parameters. I meant that all four items Games, Wargames, Dynamic Simulations, and Simulation Games share components.
Again, saying it another way:

•Games are a series of decisions within an artificial environment towards a goal.

•Wargames are a series of decisions within an artificial environment of war towards military goals such as capturing terrain, the destruction of the enemy etc.
•Dynamic Simulations are artificial environments that model real environments offering alternative 'what if' choices.

•Simulation Game is a series of decisions within an artificial environment that model a real environment towards goals found in that modeled reality.

How these different and similar should be obvious. They are not the same thing, but do have the share components, which is why all of the folks I quoted from a wide variety of venues seem to be using those words interchangeably—because they do share many of the same components.

You go on to say "they all are simulations. [i.e. designs whose purpose is to mimic, copy, represent, recreate, imitate etc. etc. some aspects of reality, past or present.]"
Under this definition, model railroading is simulation, painting landscapes is simulation. Making and sailing model boats is simulation. Carving animals is simulation. Building model houses with toothpicks is simulation. They all imitate some aspect of reality. Your definition lacks the one thing all definitions need to do. It fails to define.

Those things are 'models' of some aspect of reality, but actually I would place them in the Static simulation department if at all, where an event or particular thing is being represented, not an environment per se. As I didn't ever deal with Static simulations, I don't know what technical differences there are between them and simple models [or replicants if you want.] There entire table of a model railroad could be called a simulation, particularly if the modeler had choices to make within the setup that mimic the real choices faced by engineers etc.

That means when Jerry Banks speaks of time as a unique arbitrator of simulations, that applies to ALL simulations. Time, the progress of activities, whether in a game or a computer program, is a core issue for a simulation.
This would make chess a simulation., Monopoly, Life, etc.

Why? Simply because time is the arbitrator of the progress of
activities in a game doesn't make it a simulation because simulations share that design trait. It is a shared trait and because of that, there are shared technical aspects. Games don't have to simulate anything in reality, regardless of what design similarities there are.

"Simulations come with "computational baggage" only when folks insist that it must be attached."
Jim Dunnigan works with computational models. Sid Mieirs works with computational models. Ernest Adams and Andrew Rolling work with computational models. David Bartlett works with computational models. Jerry Banks works with computational models. All these men work with simulations. How can simulations not come with computational baggage? And none of these men are creating hobby wargaming rules.

The baggage I was talking about is the belief that simulations have to contain 'computational models' to be simulations. Not true. Computational systems are a way of simulating, and a very popular choice, but not the only way. For instance, give two people a scenario for an office conflict and ask them to reach an agreement based on the goals given the each person in their version of the scenario. They are to use certain verbal skills in the simulation. This is a game with goals simulating a particular office environment and problem.

The actual skills and methods developed in the simulation are shown to directly aid participants in being successful in real situations through rigorous validation, which establish that the skills learned in the simulation have direct, real world application in situations similar to the simulation. No computational models are necessary.

When David Bartlett speaks of simulations, his organization had developed any number of simulations for the military that require no computational models in the simulation. So I will repeat, a valid simulation does not require computational models.

"A wargame, on the other hand, has to simulate something in war to be a wargame, so the only difference is that is is a simulation of war. This connection is strong enough that dictionaries define wargames as simulations."
1. Here is where I think things get a bit "wookie" [a technical term]. To imitate something does not make that something into a simulation. There must be more to it than that.
2. Modern wargames, with their computational models, are simulations.

Who says there must be more to it? Lots of simulation professionals don't agree. Technically, once you start trying to imitate a real environment, you run into the very same problems that ALL simulation designers face, so why does there need to be more to it? If the purposes are the same, if the technical problems are the same, and the methodologies work for all of them, and the result is often indistinguishable from each other, why not say they are the same thing?

"Simulation designers are constantly faced with creating simulations that can't have their validity tested against the real world, but still have to accurately mimic it all the same."
The real world is the validity test. Once built, that production line either works or it doesn't. Wargaming, on the other hand, is based on history. History is not an objective science. I don't care what Ranke intended, history is a subjective enterprise. This means you do not have the real-world objective validation one has with simulations and their computational models.

Sorry, not true. Part of the reason is that you need to understand what problems simulators face in creating simulations that will be validated by 'the real world.'
Many have no more reliable information than any history. For instance, I am an architect designing a stadium, one with many innovations never seen before. I have to accurately predict how crowds will act in different scenarios including outright panic to design safe and effective traffic paths.

All I have is others' descriptions of how other crowds have acted in the past in other environments. AND the subjective reports of the events. I can't test my design because it hasn't been built yet, but it is different from any other design so comparisons are difficult too. I certainly can ask 30,000 people to show up so I can run flow exercises. How do you simulate something as subjective as 'crowd behavior?' So, without risking hundreds of millions of dollars in a faulty design, how do I simulate crowd activities in my stadium design so I can be confident the design will work once built?

This kind of problem is faced all the time by professional simulators. Over the decades, by experimentation and study, they have validated methods that actually allow a designer to confidently design a simulation of a reality not open to testing, which will none the less prove to be an effective simulation of that environment. Most often such simulations are based in part on subjective interpretations, an absence of critical information etc. etc. Methods have been developed to over come such obstacles. ALL of them are applicable to wargame design of historical events.

Wargame designers don't have to use such methods if they don't want to, but that is a choice, not because such validation is impossible.

Bill, what you seem to be doing is creating an overly broad definition of simulation so you can include wargaming within the scope of simulation. Now had you said wargaming could benefit from some the new techniques being developed by simulation designers, I might have gone there with you.

I did say that. But, how can wargames benefit from simulation design methods if they aren't simulations? I say that any design meant to simulate, mimic, or recreate a past or present environment is by default a simulation, and will invariably face the same design issues present in all simulation design, whether computational or not, whether research or not, whether a wargame or not.

I say this with a lot of confidence because that has been my experience and the experience of a vast number of simulation and game designers. That is why you find Bartlett calling his wargames simulations, while Meiers, [I never get the i's and e's right] Adams and Rollings stating their games simulate environments.

I think there are things wargaming, as a game, can learn from developments in simulations. But to make the definition of simulation so broad that it loses its ability to discriminate is not the way I would have gone. That's why I think how gamers and simulators go about their creations is as important as the fact that they are engaged in trying to mimic aspects of "reality".

What thousands of simulation and game designers have found over several decades is that their definitions of simulations and games have overlapped more and more. Why? It is a direct result of their technical efforts to develop simulations and games. By far the largest amount of money generating these discoveries has been spent in the computer game and research industries.

It isn't strange that these overlaps would be revealed first in computer designs or that many would see simulations as requiring computational programs, even if the it isn't true. [We are talking 100+ Billion Dollars in just the last decade alone.]

A game is different from a simulation if the designer doesn't create his game to model reality in any way. A simulation isn't a game if it contains no participant goals or competitive decisions within the recreated environment. A wargame doesn't have to be simulation for the same reason. Most wargames ARE simulations to be of any use, and many game designers chose to create games that model reality. They are ipso facto simulations.

Best Regards,
Bill

Rich wrote:
Wargaming is an historical interpretative endeavor. Historical interpretation is an individualistic interpretative endeavor. The interpretation depends on the person doing the interpreting. It would be like telling gamers how they can interpret history.

Rich:
Not really. Parts of history have to be interpreted, but not the large majority of it. For instance, Napoleon was the Emperor of the French, Lincoln was President during the AWC. The French infantry wore blue during much of the 19th Century. I feel quite comfortable in telling gamers how to 'interpret' this kind of history.

To claim these are open to interpretation would be rather silly. To state that Napoleon was a bad emperor, or that his military prowess depended solely on his generals, or that Lincoln meddled too much in military affairs are certainly open to interpretation. However, even there it is necessary to interpret *something*, and that something is the historical evidence. The difference between knowing that Napoleon was Emperor and determining that he was a bad emperor has a lot to do with what evidence is available and the general consensus of historians etc. It is the question "is it beyond a reasonable doubt?" based on the evidence. Historians are constantly moving bits of history into this
"beyond a reasonable doubt" category through research and study. We simply accept it as 'true', but it has been hard won. Lots is open to interpretation, but lots isn't.

That isn't to say the hobby doesn't have standards. It's just that the standards are not codified. Nor are they ridged. The standards are our expectations of what a wargame is and what should be included. Game designers must meet these expectations (standards) if the rules are to be accepted.

What standards are those? Beyond a wobbly size standard for miniatures [How many standards are there for a 15mm figure--from 12mm to 18mm?] or an equally loose description of games as grand tactical, tactical, skirmish, brigade-sized etc. etc. [How many lengthy discussions have been on this list over those distinctions to no avail?]

There aren't any standards or categories or functional definitions for that matter. Expectations aren't standards, and certainly in this hobby are all over the map. I am not sure if even a solid dozen would agree on their 'expectations' for anything meaningful other than the vague 'it's gotta be fun.'

What expectations do designers meet that you would qualify for 'standards', other than miniature's games need to use miniatures…? For instance, the history of the HMGS is replete with examples of the results created by an absent standards/definitions/categories for whatever we are doing.
That isn't taking away one whit of the good things they do for the hobby, but it is still true that a major stumbling block for their efforts is that no one can come up with an acceptable description of what it is we do.

It's like the National Association of RC Modelers being unable to come up with a definition of what an RC Model is or what different kinds exist, free flight, scale, true scale etc. It would be a very different hobby if they couldn't come up with some. Ours would be very different if we did.

Best Regards,
Bill

I have re re evaluated my ideas on a some sort of industry standard of definition. I think they would be helpful for player and game designer. Maybe a wiki should be made or something.

This is a good thread. I heartily agree with pretty much everything Scotsman has said.

I have re re evaluated my ideas on a some sort of industry standard of definition. I think they would be helpful for player and game designer. Maybe a wiki should be made or something.

Wiki is probably the best tool for a project like that. Was there already a thread about those definitions somewhere?

Wiki is probably the best tool for a project like that. Was there already a thread about those definitions somewhere?

Perhaps, most topics come around at least twice. I have started a wiki for project Im going to start posting terms for discussion

Sorry for the lateness of the reply. I have been reading up more on the subject. Let's begin with the obvious. A simulation is a noun. Nouns name a person, place, or thing. In this case simulation names a thing. Simulate is a verb. There is no automatic relationship between simulate and simulation any more than there is between replicate and simulation.

Simulations come in two forms: dynamic and static. Both are involved with real world processes not environments (more about that later). The difference between the two is ‘time'. Dynamic simulations model real world processes from the beginning of the process to the end. Static simulations model a slice of that process. For example, you have the beginning of the manufacturing process, you have resources added to the processes at certain points and you have the finished product at the end. Dynamic simulation models the process from beginning to end. It is dynamic because it takes a certain amount of time for the manufacturing process to be performed. On the other hand, if you want to model what happens when resources are added to the manufacturing process, this is called a static simulation. For the sake of the simulation, it is considered to occur instantly. At one moment the process lacks this resource. At another it has the resource.

Both dynamic and static simulations can be broken into deterministic simulation and stochastic simulation. Deterministic simulations have no random variables and stochastic simulations do have random variables. For examples, if the resources, added to the production process, come as fixed amounts then the model of this would be called a deterministic static simulation. If, on the other hand, the amount varies based on what is happening in the production process prior to the addition, this is a stochastic static simulation.

The term ‘simulation' is simply a term used to describe the creation of a model(s) that attempts to replicate a real world process. No model, no simulation. The model itself can be composed of logic elements or computational elements.
Now Game Theory, for example, is not a simulation. What Game Theory does is to set up a real world environment for decision-making. An environment is not a process. It is a set of conditions by which certain decisions can be made. There is no real world process being mimicked and therefore no dynamic or static simulations. One can object and say that game theory simulates this decision-making environment. Therefore, it is a simulation. As has been pointed out, simulation is a noun and requires a model. Simulate is a verb and can be used with a simulation or with other creative endeavors. The use of the term ‘simulate' does not automatically refer to a simulation anymore than the term replication automatically refers to simulation. Therefore, simulating something is not enough to make it a simulation.

Does game theory use models? Of course. But they are models of the decision-making environment not models of real world processes. There is processing going on when a decision is made but the processes occurs in the brain. Our knowledge of the decision-making process of the brain is nowhere close to allow us to model that process.

How do you simulate something as subjective as 'crowd behavior?' So, without risking hundreds of millions of dollars in a faulty design, how do I simulate crowd activities in my stadium design so I can be confident the design will work once built? …. Most often such simulations are based in part on subjective interpretations, an absence of critical information etc. etc.

Bill, how much do you know about simulations? Your stadium architect should have access to crowd flow models that can predict areas of likely congestion and analyze options for improvement. Stadiums and railroad facilities have been at the forefront of this type of simulation development.

I say that any design meant to simulate, mimic, or recreate a past or present environment is by default a simulation.

I think we have covered this. Simulations, dynamic and static, model real world processes not environments unless we are talking about computer programs that are designed to create environments. Also, where there is no model there is no simulation.

A simulation, dynamic or static, depend upon the modeling of real world processes not environments. Past events can be modeled if those events were caused by processes still in existent today. For, example, you can model the processes by which a dam failed 10 years ago because the processes which caused the dam to fail are processes still valid today. The physics hasn't changed.

Not really. Parts of history have to be interpreted, but not the large majority of it. For instance, Napoleon was the Emperor of the French, Lincoln was President during the AWC. The French infantry wore blue during much of the 19th Century. I feel quite comfortable in telling gamers how to 'interpret' this kind of history.

That is NOT history Bill. It maybe historical facts but it is not history. There is no such thing as ‘Parts of History'. History is what historians create. There is no history until the historian creates it. You don't just string together a number of historical facts and call that history. I got at least that much from doctorial program in History at the University of Washington. A historian gathers a number of historical facts, interprets them, writes down the interpretation and the result is history.

For example, it was readily accepted that the English beat the French, during the Napoleonic wars, through superior musket firepower. Paddy Griffith comes along, after studying the same source material and say no. The English beat the French because they combined an initial volley with a bayonet charge. From his analysis of his source material, Griffith re-interpreted what the sources had to say. History is the end product of the interpretation of source material. History is an interpretive endeavor.

What wargame rule writers do is very similar. From his study of his source material he develops an interpretation of how wargame battles should unfold. This interpretation guides him in the creation of rules by which certain aspects of battle can be recreated on the tabletop with miniature figures. This is not a simulation. It is not game theory. It is historical interpretation.

Rich

Rich Knapton wrote:

[Snips]
Let's begin with the obvious. A simulation is a noun. Nouns name a person, place, or thing. In this case simulation names a thing. Simulate is a verb.

It's pretty dubious to base any argument on the distinction between a noun and a verb in English, where the two are much less clearly distinct than in more disciplined languages. (Consider "running", or even "simulating". Is it a verb, a noun, or an adjective? The answer is "yes, it is", as it is the form respectively of the present participle, the gerund, and the gerundive.)

There is no automatic relationship between simulate and simulation any more than there is between replicate and simulation.

It's not merely dubious, it's flat out wrong, to claim that "simulate" and "simulation" are unrelated. They are the verbal and nominal forms from precisely the same etymological root. They are necessarily related most strongly. To claim otherwise indicates a profound falure to understand the elementary workings of language.

Simulations come in two forms: dynamic and static.

That's one way of classifying them, sure. There are plenty of others (I must finish off my paper on "A classification of simulation classifications" some time soon). As has already been indicated, one fairly common convention treats "simulation" as referring specifically to dynamic models; but it need not be so. If you accept the division of simulations into static and dynamic, then it is worth bearing in mind that it is possible to create static models of dynamic systems (e.g. Lanchester's laws of combat), and dynamic models of static systems (e.g. Buffon's needle to find an approximate value for pi). This reinforces the point that it is in the nature of simulation modelling for a simulation to be able to have different properties from the system it simulates. A trivial point, but one often misunderstood.

Both are involved with real world processes not environments (more about that later).

Or imaginary processes. No principle of simulation says that the simulation has to be based on a real system; some of the most profitable uses of simulation are for considering systems that don't exist yet.

The difference between the two is ‘time'. Dynamic simulations model real world processes from the beginning of the process to the end. Static simulations model a slice of that process.

The distinction between "dynamic" and "static" has nothing to do with whether a process (itself an arbitrary construct) is followed from "end to end" or only in part. "Dynamic" and "static" mean simply what they say; either the model moves, or it doesn't. You can have a static model of an "end-to-end" process, just as you can have a dynamic model of a "slice of a process".

Both dynamic and static simulations can be broken into deterministic simulation and stochastic simulation.

That's another axis of classification, yup.

The term ‘simulation' is simply a term used to describe the creation of a model(s) that attempts to replicate a real world process. No model, no simulation. The model itself can be composed of logic elements or computational elements.

Apart from the fact that it needn't be a process, and it needn't be in the real world, yes. Certainly it is true that simulation is fundamentally a modelling activity.

Now Game Theory, for example, is not a simulation.

Why not? Game theory models competing strategies. If we allow that simulations can be static, I see no reason that a game-theoretic approach cannot be considered a simulation. What's more, there are extensions to game theory, such as the theory of moves and drama theory, that add dynamic elements. In many cases these models purport to represent strategic choices to be made by decision-makers in the real world. I can think of no ground on which they should not be considered simulations; static deterministic simulations, certainly, but if you don't exclude static models from the category of simulation, you can hardly exclude these.

What Game Theory does is to set up a real world environment for decision-making.

No it doesn't. Game theory is a branch of applied mathematics, and mathematicians maintain a sturdy independence from the real world, and never use words like "environment".

An environment is not a process.

Why not? Whether something is a process or not is merely a question of how you choose to look at it. You can look at, say, a building as a static (noun-like) thing if you like. Or you can recognise that it exists in time, there was a time when it wasn't here, there will be a time when it isn't here again, and you can look at it as a process. Such categories are imposed by the way you choose to look at the world, they are not an immutable part of external reality.

There is no real world process being mimicked and therefore no dynamic or static simulations.

Not true at all. When Schelling and Rapoport used game theory to discuss the wisdom or otherwise of different strategies for nuclear deterrence and disarmament, they very clearly and explictly intended to mimic (in some important respects) the real-world processes of nuclear strategy.

One can object and say that game theory simulates this decision-making environment. Therefore, it is a simulation. As has been pointed out, simulation is a noun and requires a model. Simulate is a verb and can be used with a simulation or with other creative endeavors. The use of the term ‘simulate' does not automatically refer to a simulation anymore than the term replication automatically refers to simulation. Therefore, simulating something is not enough to make it a simulation.

This is terrible nonsense. The term "simulate" is very clearly joined at the hip to the term "simulation", and simulating something is exactly and exclusively what makes a simulation.

Does game theory use models? Of course. But they are models of the decision-making environment not models of real world processes.

In what way is decision-making not a "real-world" process?

Your arguments seem to depend on creating completely false distinctions, and then using them to prove that something is not the same as itself.

There is processing going on when a decision is made but the processes occurs in the brain. Our knowledge of the decision-making process of the brain is nowhere close to allow us to model that process.

You seem to have failed to understand the essential nature of modelling. Simplification is the essence of modelling; you don't need to understand the neurological processes of human congition to create a model of decision-making, and, indeed, if you did possess such understanding, you wouldn't have any need to make the model in the first place.

Bill, how much do you know about simulations? Your stadium architect should have access to crowd flow models that can predict areas of likely congestion and analyze options for improvement. Stadiums and railroad facilities have been at the forefront of this type of simulation development.

Reputable simulationists are very careful about using words like "predict". Prediction sounds suspiciously like foretelling the future, which is a job for oracles, soothsayers, astragallomancers and sybils, not for simulation modellers.

[Snips]
That is NOT history Bill. It maybe historical facts but it is not history. There is no such thing as ‘Parts of History'. History is what historians create. There is no history until the historian creates it. You don't just string together a number of historical facts and call that history. I got at least that much from doctorial program in History at the University of Washington. A historian gathers a number of historical facts, interprets them, writes down the interpretation and the result is history.

And a simulation modeller gathers a number of facts, interprets them, expresses the facts and interpretations as an executable model, and the result is a simulation.

And, just as a history that says that says Napoleon Bonaparte was a Swiss punk rocker would be poor history, so a simulation model that showed helicopters spontaneously transmogrifying themselves into spanners would be a poor simulation (and I've seen one that did that, before it was corrected). But neither the history nor the simulation is expected to reflect its prototype in its full detail and variety.

What wargame rule writers do is very similar. From his study of his source material he develops an interpretation of how wargame battles should unfold. This interpretation guides him in the creation of rules by which certain aspects of battle can be recreated on the tabletop with miniature figures.

Yes -- all quite right…

This is not a simulation.

…up until this bit. It is precisely a simulation; it is a simplified executable model that tries to capture the essential elements of the system under study.

I'm still not entirely sure just what it is you are failing to understand about simulation, but you seem to me to have got hold of the (not uncommon) mistaken view that a simulation is some sort of mechanistic, interpretation-free mapping of an objective system into a model, which can be in some absolute and reductionist sense "correct" or "incorrect". It isn't.

All the best,

John.

John D Salt: [Hi John]

It's pretty dubious to base any argument on the distinction between a noun and a verb in English, where the two are much less clearly distinct than in more disciplined languages. (Consider "running", or even "simulating". Is it a verb, a noun, or an adjective? The answer is "yes, it is", as it is the form respectively of the present participle, the gerund, and the gerundive.)

It doesn't matter how ‘simulation' became a noun. The fact of the matter is it is a noun.

It's not merely dubious, it's flat out wrong, to claim that "simulate" and "simulation" are unrelated.

John you're going to have to do a lot better job than that of reading what I wrote.

This reinforces the point that it is in the nature of simulation modelling for a simulation to be able to have different properties from the system it simulates.

However, I noticed that the only two terms you used were static and dynamic. This rather reinforces my point. There may be combinations of the two but there are only two forms static or dynamic.

"Dynamic" and "static" mean simply what they say; either the model moves, or it doesn't.

As I wrote in that same paragraph: "For the sake of the simulation, it is considered to occur instantly."

[mine: What Game Theory does is to set up a real world environment for decision-making.] No it doesn't. Game theory is a branch of applied mathematics, and mathematicians maintain a sturdy independence from the real world, and never use words like "environment".

I'm afraid it does John. While your back was turned economists and social scientists swiped the mathematicians' Game Theory for their own purposes. One is to mathematically capture social and economic environments in which decisions are made.

Why not? Game theory models competing strategies. If we allow that simulations can be static, I see no reason that a game-theoretic approach cannot be considered a simulation.

I don't see how a model of various strategies equates to a static model of a real world process. The reason I use the term real world process is that this discussion is in relation to historical wargaming. Simulations of non-real processes have no relevance to this discussion.

Not true at all. When Schelling and Rapoport used game theory to discuss the wisdom or otherwise of different strategies for nuclear deterrence and disarmament, they very clearly and explictly intended to mimic (in some important respects) the real-world processes of nuclear strategy.

It seems to me that what Schelling and Rapoport wrote about was the development of models by which nuclear strategy can be created. In other words, in the absence of any effective real world process for nuclear strategy they were attempting to develop such strategy. If that is true, there was no effective process for them to model. Game theory yes, but hardly a simulation. Simulation presupposes an already existing process. Remember we are speaking in terms of real world processes.

This is terrible nonsense. The term "simulate" is very clearly joined at the hip to the term "simulation", and simulating something is exactly and exclusively what makes a simulation.

John, you must keep in mind the frame of reference in which this discussion is taking place. We are discussing modeling real world processes and how that relates to wargaming. Simulation is the noun form of the word simulate which is itself a verb. In all likelihood, the noun form was created based on the verb form. So one would expect a relationship. However, they are not synonymous. To assume that each time one uses the term simulate one is talking about a simulation is an error in logic.

In what way is decision-making not a "real-world" process?

Once again you misread my posting.

Your arguments seem to depend on creating completely false distinctions, and then using them to prove that something is not the same as itself.

On the contrary, I make "technical" distinctions and then compare them.

You seem to have failed to understand the essential nature of modelling. Simplification is the essence of modelling; you don't need to understand the neurological processes of human congition to create a model of decision-making, and, indeed, if you did possess such understanding, you wouldn't have any need to make the model in the first place.

I think John you have lost sight of what is being modeled. Schelling did not propose the modeling of existing real world decision-making processes. What he was doing was to help provide a functional decision-making process.

Reputable simulationists are very careful about using words like "predict". Prediction sounds suspiciously like foretelling the future, which is a job for oracles, soothsayers, astragallomancers and sybils, not for simulation modellers.

Let me turn to my trusty little red dictionary: Predict: "say or estimate that (a specified thing) will happen in the future or will be a consequence of something." In other words, "estimate what will happen." The whole function of the models is to create information about likely future events so intelligent decisions can be made. This has nothing to do with soothsaying.

And a simulation modeller gathers a number of facts, interprets them, expresses the facts and interpretations as an executable model, and the result is a simulation.

Historians do not create executable models. What they do is interpret human events that occur over time. One can say historians look at human events dynamically. Anthropologists, on the other hand, look at human events statically.

And, just as a history that says that says Napoleon Bonaparte was a Swiss punk rocker would be poor history,

No, that would not be history.

But neither the history nor the simulation is expected to reflect its prototype in its full detail and variety.

There are no historical prototypes.

…up until this bit. It is precisely a simulation; it is a simplified executable model that tries to capture the essential elements of the system under study.

What you fail to understand John is that in history you cannot get to "the essential elements of the system under study." The best that the historian can do is get to what others believed were the essential elements and left record of that belief. Therefore, the historian does not create models that try "to capture the essential elements of the system under study".

I'm still not entirely sure just what it is you are failing to understand about simulation, but you seem to me to have got hold of the (not uncommon) mistaken view that a simulation is some sort of mechanistic, interpretation-free mapping of an objective system into a model, which can be in some absolute and reductionist sense "correct" or "incorrect". It isn't.

John, if you are going to state, as Bill has, that the discipline of simulation has much to teach the writers of wargame rules, it is incumbent upon us to define what that discipline is. To say that anything that is simulated is a simulation reflects no discipline. The discipline referred to are those involved with dynamic or static modeling of real world processes. The question is, does that describe what wargame rule writers to do? My opinion is it does not. Historical interpretation (which is what rule writers do) is not the same as modeling real world processes.

Rich

Rich,

Out of curiosity, what books have you been reading about the subject?

You are getting the basics mostly right now, but somehow I feel that you are just making things overcomplicated with some weird distinctions.

Also, are you sure that your frame of reference in valid? I was pretty sure this thread was about wargaming in general and not only about historical games (actually I was the first one who even mentioned the phrase "historical wargame" and that was in the middle of page one). Wargaming is often about what-ifs and possible future events or even completely imaginary subjects, not just "real-world processes". Same applies to simulations.

Rich wrote:
Sorry for the lateness of the reply. I have been reading up more on the subject. Let's begin with the obvious. A simulation is a noun. Nouns name a person, place, or thing. In this case simulation names a thing. Simulate is a verb. There is no automatic relationship between simulate and simulation any more than there is between replicate and simulation.

Rich:
Hi. No problem. I was off in the hinterland anyway. Hmmm. "No automatic relationship."

Okay, so linguistics, semantics, word construction and derivation, epistemology, and history, as well as the dictionary don't apply. From my experience, when someone sets out to simulate something, they end up creating simulations. Regardless of 'how' they do that, they encounter and have to solve technical and conceptual problems that are inherent in simulation design. It doesn't matter how you say it. It has also been my experience that when I talk about simulations 'simulating' something or others talk to me, no one seems to have a problem understanding.

Whether that is an automatic relationship or not, we are talking about something people do. They create simulations which are designed to simulate. If you believe that there are simulations that don't simulate, and efforts to simulate that never become simulations, you're welcome to. I tend to find little problem with the automatic relationship between simulations and simulating in communicating with simulation designers and must other folks.

The real question is how such non-automatic relationships aid simulation game design, if at all.

Simulations come in two forms: dynamic and static. Both are involved with real world processes not environments (more about that later). The difference between the two is ‘time'. Dynamic simulations model real world processes from the beginning of the process to the end. Static simulations model a slice of that process. For example, you have the beginning of the manufacturing process, you have resources added to the processes at certain points and you have the finished product at the end. Dynamic simulation models the process from beginning to end. It is dynamic because it takes a certain amount of time for the manufacturing process to be performed. On the other hand, if you want to model what happens when resources are added to the manufacturing process, this is called a static simulation. For the sake of the simulation, it is considered to occur instantly. At one moment the process lacks this resource. At another it has the resource.

Time is a core component in any simulation, determining the procession of activities, but both static and dynamic simulations can have their system run on the same time increments. Rich, it seems you've defined static simulations as a 'part' of a dynamic simulation. And both dynamic and static simulations can have things happen 'instantly' within the simulation. The amount of simulation time processed does not define static and dynamic simulations. Where did you get that idea?

My understanding of Static simulations is any simulation recreated to model a single event and only one event, which means the outcome is always the same result. Dynamic simulations recreate an environment that can be manipulated with a variety of outcomes. Wargame designers are constantly confusing the two, trying to force 'historical events' on players while creating a dynamic simulation that can offer alternative events.

Anyone who designs simulations for a living that I have met or read recognizes those definitions of Dynamic and Static Simulations. I haven't seen your definitions before.

Both dynamic and static simulations can be broken into deterministic simulation and stochastic simulation. Deterministic simulations have no random variables and stochastic simulations do have random variables. For examples, if the resources, added to the production process, come as fixed amounts then the model of this would be called a deterministic static simulation. If, on the other hand, the amount varies based on what is happening in the production process prior to the addition, this is a stochastic static simulation.

Rich, where are you getting this? Now you have simulations types not based on the system, but based on what kinds of inputs are added to or used in a simulation: No random variables: Deterministic and random variables: Stochastic.
Or outputs: deterministic static or stochastic static. This seems to imply that one simulation, deterministic, will always produce the same result because there are no random variables, and the other won't. Seems to be the Static and Dynamic simulations according to my definitions.

It strikes me as defining a mixer by what fruit you put in it. Unless the mixer was specifically designed to only puree strawberries, you would hardly change the name of the mixer with each kind of fruit you blend.

You may be thinking of Static and Dynamic VV&T Techniques, which is closer to your definitions. Such techniques are methods for assessing accuracy of a static or dynamic model, which can be part or all of a simulation. It is done by controlling the 'input' to test how the system runs, and thus what 'products' it creates.

Research, testing, training and wargames require a simulation model provide for a wide variety of variables or decisions, which is a dynamic simulation. If you want to define simulations by input and output, that is your privilege, but I don't see how that helps one design simulations games. I am sure a "deterministic static simulation" would be a real bore.

The term ‘simulation' is simply a term used to describe the creation of a model(s) that attempts to replicate a real world process. No model, no simulation. The model itself can be composed of logic elements or computational elements.

I certainly can agree with that, other that to say that I think the model can be composed of more than just logic elements or computational elements. It depends on what is being modeled and how. Emotional and physical elements can be major components.

Now Game Theory, for example, is not a simulation. What Game Theory does is to set up a real world environment for decision-making. An environment is not a process.

Game Theory doesn't set up anything but a theory on games and decision-making. SOME game designers use Game Theory to set up real world environments for decision-making, but I don't think that is a requirement of Game Theory.

Most simulation designers would disagree with you about a process not being an environment. In fact, any process that has an entry/beginning and exit/end, from a roller coaster ride to a computer model of chemical reactions must create an environment with its own internal rules, systems, logic and cause and effect—in other words an environment. Chess creates an environment. The robotic construction of a car on an assembly line creates an environment and neither would work if they didn't.

Set up a process, such as driving a screw into a piece of wood. All of a sudden what wood, how it is laid out, the type of screw and driver, angle of the hole etc. etc. all come into play—if we are talking about something real and not theoretical. Context is impossible to avoid when anyone discusses practical and/or physical results.

It is also very difficult to imagine an environment of any sort without processes. If the environment responds, requires, or produces anything, it does it through a process. That is how environments express themselves, how they exist. Even the landscape on the moon has processes that erode the rock specific to that environment, that help create that environment.

Again, practical applications here for game design, not semantics or theories.

It is a set of conditions by which certain decisions can be made. There is no real world process being mimicked and therefore no dynamic or static simulations. One can object and say that game theory simulates this decision-making environment. Therefore, it is a simulation. As has been pointed out, simulation is a noun and requires a model. Simulate is a verb and can be used with a simulation or with other creative endeavors. The use of the term ‘simulate' does not automatically refer to a simulation anymore than the term replication automatically refers to simulation. Therefore, simulating something is not enough to make it a simulation.

What? Game Theory doesn't set up a real world environment for decision-making. Game Theory is just that, a theory on how decisions are made. It certainly has spawned simulations, but it has also generated games that have little relationship to any particular reality. They are just decision-making games. All the variations of tic-tac-toe for instance.

Does game theory use models? Of course. But they are models of the decision-making environment not models of real world processes. There is processing going on when a decision is made but the processes occurs in the brain. Our knowledge of the decision-making process of the brain is nowhere close to allow us to model that process.

You've lost me there. You just wrote "What Game Theory does is to set up a real world environment for decision-making." And you have models being environments. Who said anything about modeling the brain?

We are talking simulation game design, yes?

Bill, how much do you know about simulations? Your stadium architect should have access to crowd flow models that can predict areas of likely congestion and analyze options for improvement. Stadiums and railroad facilities have been at the forefront of this type of simulation development.

? I am not sure what you are saying here. Those flow models certainly do exist and are used [some created from other simulations], but they often can only suggest—not determine—likely areas of congestion for new structures. If the models that architects 'should have access to" already answered all possible design questions, there would be no need for more simulations of such things. But they don't, which is why "Stadiums and railroad facilities have been at the forefront of this type of simulation development."

I still say that any design meant to simulate, mimic, or recreate a past or present environment is by default a simulation. And because of that, I can expect to find similar design issues involved regardless of the medium. I haven't been disappointed yet.

I think we have covered this. Simulations, dynamic and static, model real world processes not environments unless we are talking about computer programs that are designed to create environments. Also, where there is no model there is no simulation.

Argh! We have covered this. I defy you to build a "real world process" without creating an environment for it to operate in. That process will have limits, internal rules, values, inclusions and exclusions, products etc. If it is a 'real world' process, where does the 'real world' come into play, if not as an environment for the process? Computers aren't the repository for all created environments. A wargame is an artificial environment. A chess game is an artificial environment.

That is NOT history Bill. It maybe historical facts but it is not history. There is no such thing as ‘Parts of History'. History is what historians create. There is no history until the historian creates it. You don't just string together a number of historical facts and call that history. I got at least that much from doctoral program in History at the University of Washington. A historian gathers a number of historical facts, interprets them, writes down the interpretation and the result is history.

Let's not get sidetracked by the academic definition of 'history' as exclusively what historians write. When I say history, I am talking about what has happened in the past and what we can know of it, however limited that might be, and then how that can be modeled in a simulation game.

Historians analyze, organizing and present research and books on what they have discovered of the past, which is wonderful stuff and produce most of the growth of historical understanding. They refer to that creation as 'history', but it isn't the total sum of what we can know about the past and not the limit of what I am talking about, as much as their labors add to our understanding.

For example, it was readily accepted that the English beat the French, during the Napoleonic wars, through superior musket firepower. Paddy Griffith comes along, after studying the same source material and say no. The English beat the French because they combined an initial volley with a bayonet charge. From his analysis of his source material, Griffith re-interpreted what the sources had to say. History is the end product of the interpretation of source material. History is an interpretive endeavor.

History is indeed an interpretive endeavor, but that is there is much, much more to that 'product' than just interpretation. It doesn't describe the entire activity or even most of it. And you would know that from a doctorial program.

Actually, having discussed with Paddy some of his work back in the 1980s, I know that he was and is doing far more than 're-interpreting' the known, but adding to the known with new and overlooked source material—not simply the same materials rehashed with a 'new interpretation.'

*Napoleonic French infantry fought with muskets and
bayonets.
*The Battle of Borodino was fought in Russia.
*Wellington commanded the Allied army at Waterloo.

If you believe those historical statements are simply interpretive products of some historian, then we really are going to have to agree to disagree.

What wargame rule writers do is very similar. From his study of his source material he develops an interpretation of how wargame battles should unfold. This interpretation guides him in the creation of rules by which certain aspects of battle can be recreated on the tabletop with miniature figures. This is not a simulation. It is not game theory. It is historical interpretation.

Perhaps, if they do more than simply read and accept a historian's own interpretation.I know of at least two wargames that are based solely on Ryan's interpretation of history in "A Bridge Too Far" and nothing more. That isn't necessarily bad, but it sure isn't the game designer's interpretation of history. It certainly would be his interpretation of how the game mechanics portray Ryan's history. Another issue. The designer is free to depend on the historian's interpretation, but that interpretation has no historical value unless it is based on historical data.

Rich, the game designer is interpreting *something*, and he is then claiming that his wargame simulates that *something*. Unless we are talking free-flight fantasy, such efforts are based on *something* other than simple imagination. That certainly is a canon from my college days in the history department. Paddy claimed that his 'interpretation' was a clearer picture of 'what really happened.' He defended his interpretation by providing historical facts to support it, so his interpretation was based on *something*, and in analyzing and producing 'history', that relationship had to be established. You know, all those footnotes and bibliographies and stuff. I agree that many wargame designers claim to be simulating his interpretation of history—but that isn't always true, it can be someone else's interpretation. Even so, *something* has to exist to be interpreted… That makes it at least half of a historian's work.

So where does that leave us? If a designer claims [operative word here] to have designed a game that simulates, recreates, represents, mimics, or models some part of history, a Napoleonic battle or whatever, his interpretation has to have some relationship to historical evidence, historical sources. That simulation game must somehow model that evidence, those sources, to be a simulation. It must simulate *something.*

I have been endeavoring to establish how that is done, technically, practically, not philosophically. If the claim is recreating history in a game, how is that claim achieved, established, justified, determined and proven? It is a technical claim and demands technical answers.

Bill H.

Wargame, n. A game about war, or some aspect of it, such as a battle, real or imaginary.

Wargame, n. A game about war, or some aspect of it, such as a battle, real or imaginary.

DonL:

Yep, seems pretty straight forward to me.

Karsta (Hi Karsta)

You are getting the basics mostly right now, but somehow I feel that you are just making things overcomplicated with some weird distinctions.

That's probably a factor of just now learning what all this information means. I need to take the concepts I'm learning and put them into expressions I'm comfortable with. I guess the result is a bit weird.

Also, are you sure that your frame of reference in valid?

The frame of reference is simulations as models of real world processes and its relation to wargaming. My assumption is that real world processes precludes other forms of wargaming than historical.

Wargaming is often about what-ifs and possible future events or even completely imaginary subjects, not just "real-world processes". Same applies to simulations.

Yes but this is not a philosophical treatise on wargaming and simulation. In other words it is not about all the ramifications of the terms ‘wargame' and ‘simulation'. Implied in the use of the term ‘wargame' is historical miniature wargames. And, implied in the use of the term simulations is the modeling of real world practices.

Rich

Rich Knapton wrote:

It doesn't matter how ‘simulation' became a noun. The fact of the matter is it is a noun.

But that is not the question at issue, is it? Nobody disputes the nounness of the word. The problem arises when you derive an absurd conclusion from that fact, following no detectable process of argument.

John you're going to have to do a lot better job than that of reading what I wrote.

What, exactly, are you claiming that I have misunderstood? You claimed that there was no "automatic" relation between "simulation" and "simulate", a statement for which you produced no evidence. I pointed out that there is a "necessary" relation between them, based on elementary etymology. You may perhaps have some strange definition of "automatic" that makes it different from "necessary". However, I suspect that you don't.

[I wrote]
This reinforces the point that it is in the nature of simulation modelling for a simulation to be able to have different properties from the system it simulates.

However, I noticed that the only two terms you used were static and dynamic. This rather reinforces my point. There may be combinations of the two but there are only two forms static or dynamic.

No, it does not reinforce your point. The reason I used the example of the dynamic/static classification was that it was one you had introduced yourself, and therefore presumably understand. It is faulty reasoning to assume that therefore there are no others. There are, as I stated, many different axes of classification for simulation models; static/dynamic, open/closed, stochastic/deterministic, decision-support/training, interactive/batch, discrete-event/continuous, standalone/federated, and any others people may care to devise. Saying that "there are only two forms" is about as useful as saying that the world is divided into two classes of people, those who divide the world into two classes of people, and those who don't.

[I wrote]
"Dynamic" and "static" mean simply what they say; either the model moves, or it doesn't.

As I wrote in that same paragraph: "For the sake of the simulation, it is considered to occur instantly."

…which is quite a different thing. Lots of things occur instantly in dynamic simulations; it's the key trick behind the discrete-event approach.

[Snips]
I'm afraid it does John. While your back was turned economists and social scientists swiped the mathematicians' Game Theory for their own purposes. One is to mathematically capture social and economic environments in which decisions are made.

That sounds suspiciously to me as if you are admitting that game theory is used for modelling purposes.

It seems to me that you set some peculiar value on a distinction between an "environment" and a "process". Perhaps you should attmept a clearer definition of what you mean by these. I suspect that there isn't any essential difference, given the traditional system-theoretic definitions which allow systems to be environments for other systems, and make it clear that drawing the system boundary (the division between "system" and "environment") is an arbitrary act. Stafford Beer is fun to read on this sort of thing (and also one of the fathers of simulation modelling, although sadly he didn't write much about it).

[Snips]
I don't see how a model of various strategies equates to a static model of a real world process.

What don't you see? Decision-making can clearly be seen as a process, albeit in the form usually treated by game theory rather a simple one. At one stage there is a decision to be made, but which decision it is is undefined; then, a decision is taken, and the process is completed. Of course, in things like series-play prisoner's dilemma, or anything in the theory of moves, there are a series of decisions to be made, which makes the processy nature of the thing clearer still (although no longer of course a static model).

The reason I use the term real world process is that this discussion is in relation to historical wargaming. Simulations of non-real processes have no relevance to this discussion.

That smacks rather of moving the goalposts, but it does put you in considerable difficulty in defining what you mean by a "non-real process". As has been pointed out elsewhere in this thread, their "what-if" nature is essential to wargames simulation models. If Rommel gets defeated at Gazala, or Wellington at Waterloo, is that somehow not a "real process"? If so, that leaves you a very small area to discuss.

It seems to me that what Schelling and Rapoport wrote about was the development of models by which nuclear strategy can be created.

So, you agree that game theory in this case is used for modelling, yes?

In other words, in the absence of any effective real world process for nuclear strategy they were attempting to develop such strategy. If that is true, there was no effective process for them to model. Game theory yes, but hardly a simulation. Simulation presupposes an already existing process.

No, Rich, simulation does not presuppose an already existing process. You are quite wrong. When I wrote simulation models for Eurotunnel, the terminal whose operation I was modelling was a bulding-site. When I wrote simulations on the operation of EH101 flights off Type 23 destroyers in the Denmark Strait, neither the EH101 nor the Type 23 had entered service. I've just finished writing a simulation model about a tactical comms system not due to enter service until about 2012. Furrther counter-examples could be provided ad lib. The systems on which all these simulations were based had at the time they were written only theoretical existence; which is not what people normally mean by "real world" (a pretty unhelpful phrase for most simulation purposes, as it turns out).

Remember we are speaking in terms of real world processes.

You are; but that is because you don't understand simulation.

[I wrote]
This is terrible nonsense. The term "simulate" is very clearly joined at the hip to the term "simulation", and simulating something is exactly and exclusively what makes a simulation.

John, you must keep in mind the frame of reference in which this discussion is taking place. We are discussing modeling real world processes and how that relates to wargaming.

And how exactly does that affect the price of fish? It still remains true that simulating something is exactly and exclusively what makes a simulation.

Simulation is the noun form of the word simulate which is itself a verb. In all likelihood, the noun form was created based on the verb form. So one would expect a relationship. However, they are not synonymous.

"Quick" and "quickly" are an adjective and an adverb, and hence are not synonymous. So what? Can you really not see how very silly such a form of argument is?

To assume that each time one uses the term simulate one is talking about a simulation is an error in logic.

The observation, not assumption, is based on usage and etymology, not on logic. If you genuinely believe that you can adduce a logical argument to show that "to simulate" does not result in "simulation", please do present it.

[I wrote]
In what way is decision-making not a "real-world" process?

Once again you misread my posting.

In what way? It's no good you just sitting there saying "you need to do a better job of reading" and "you misread my posting". Explain in what way you think I have misunderstood.

On the contrary, I make "technical" distinctions and then compare them.

The distinction between a noun and a verb is neither particularly technical, nor, since it belongs in the category of grammar, is it one on which semantic comparisons can meaningfully be made.

If you are really interested in definitions, you might do better to accept the ones that have been pretty well agreed in simulation modelling as a professional discipline, rather than seeking to invent your own.

[Snips]
Let me turn to my trusty little red dictionary: Predict: "say or estimate that (a specified thing) will happen in the future or will be a consequence of something." In other words, "estimate what will happen." The whole function of the models is to create information about likely future events so intelligent decisions can be made. This has nothing to do with soothsaying.

I think that's not a bad description of the purpose of simulation models (decision-support models, anyway; training models are quite different). But you'll notice that "create information about" is a very different thing from "predict". That's why the old SCS strapline always used to be "Simulation: For understanding", not "Simulation: For prediction".

[I wrote]
And a simulation modeller gathers a number of facts, interprets them, expresses the facts and interpretations as an executable model, and the result is a simulation.

Historians do not create executable models.

Ah, now I fear you're going to have to do a lot better job than that of reading what I wrote. I wrote "Simulation modeller", not "historian". The historian generally expresses him or her self in narrative writing, with the odd map if we're lucky (and wads of tables from John Ellis, and, yes, executable models from Phil Sabin). The simulation modeller uses a different medium. But both are interpretive activities.

[I wrote]
And, just as a history that says that says Napoleon Bonaparte was a Swiss punk rocker would be poor history,

No, that would not be history.

So, all history that is factually incorrect is not history? Not, I think, a principle anyone is likely to accept, given the sheer bulk of historical work it would disqualify.

[I wrote]
But neither the history nor the simulation is expected to reflect its prototype in its full detail and variety.

There are no historical prototypes.

If there are no historical prototypes, then how can you possibly claim to know that Napoleon was not, in fact, a Swiss punk rocker? You're not one of those "La guerre d'Irak n'aura pas lieu" types, are you?

[I wrote]
…up until this bit. It is precisely a simulation; it is a simplified executable model that tries to capture the essential elements of the system under study.

What you fail to understand John is that in history you cannot get to "the essential elements of the system under study." The best that the historian can do is get to what others believed were the essential elements and left record of that belief. Therefore, the historian does not create models that try "to capture the essential elements of the system under study".

What are the "essential elements" obviously always depends on your own perception and the purpose opf the study. However, if you seriously believe that working historians do not try to capture the essence of what they are studying, then I think your grasp of how history works is as poor as your grasp of simulation.

John, if you are going to state, as Bill has, that the discipline of simulation has much to teach the writers of wargame rules, it is incumbent upon us to define what that discipline is.

First, I'm not going to state that; I happen to think it works more the other way round, and professional simulationists could learn a lot from the more intelligent amateur wargamers (and when MoD hires Phil Barker to do work for them, that seems to show that they agree). Second, I think it is not incumbent upon us to define the discipline so much as it is to describe it. Now, you seem to have decided to create your own defintion of "simulation" that is starkly at variance with those understood in the industry since its inception, and which I have beenm accustomed to in practice over the last twenty years.

To say that anything that is simulated is a simulation reflects no discipline.

So what? Simulation modelling is an enormously broad church, and very few of the people who practice it are aware of the full breadth of the subject. Systems Dynamics folks think it's what they do. Discrete-event simulationists think it's what they do. Electronic engineers who make flight simulators think it's what they do. Chemical engineers, for goodness' sake, think they do it when they solve systems fo flow equations. All of them (with the exception of the chemical engineers if you believe that simulations must be dynamic) are right. Yet sor some peculiar reason, some wargamers, who most certainly do do (manual, stochastic, dynamic) simulation by any defintion, are dogged in their refusal to accept that it's what they do.

The discipline referred to are those involved with dynamic or static modeling of real world processes. The question is, does that describe what wargame rule writers to do? My opinion is it does not. Historical interpretation (which is what rule writers do) is not the same as modeling real world processes.

Sculpture is not the same as painting in oils, but both are recognisably art.

How you can claim that wargames rules do not seek to model the processes of maneouvre, attrition, command, supply and so on is simply beyond me. It seems obvious on its face that this is what they do. Yet you seem determined to introduce Humpty-Dumpty redefinitions from which to chop logic to disprove this obvious fact. Why? What's so terrible about accepting that wargamers do simulation?

All the best,

John.

Donlowry:

Wargame, n. A game about war, or some aspect of it, such as a battle, real or imaginary.

TheScotsman:

Wargame, n. A game about war, or some aspect of it, such as a battle, real or imaginary. DonL:

Yep, seems pretty straight forward to me.

And it seems straight forward to me also. It's only when you throw in the term ‘simulation' that everything goes to hell.

From my experience, when someone sets out to simulate something, they end up creating simulations. Regardless of 'how' they do that, they encounter and have to solve technical and conceptual problems that are inherent in simulation design. … I tend to find little problem with the automatic relationship between simulations and simulating in communicating with simulation designers and must other folks.

Are you saying is that someone who simulates a unicorn in wood and someone who simulates a real world processes are, in essence, are doing the same thing? After all, if you are simulating something you are making a simulation. That is so simplistic as to be non-functional.

I still say that any design meant to simulate, mimic, or recreate a past or present environment is by default a simulation.

As mentioned above, this is so overly broad as to be non-functional.

Time is a core component in any simulation, determining the procession of activities, but both static and dynamic simulations can have their system run on the same time increments. Rich, it seems you've defined static simulations as a 'part' of a dynamic simulation. And both dynamic and static simulations can have things happen 'instantly' within the simulation. The amount of simulation time processed does not define static and dynamic simulations. Where did you get that idea?

Anyone who designs simulations for a living that I have met or read recognizes those definitions of Dynamic and Static Simulations. I haven't seen your definitions before.

Let's correct that.

Handbook of Simulation
"Discrete-event models are dynamic; that is, the passage of time plays a crucial role. Most mathematical and statistical models are static, in that they represent a system at a fixed point of time.

Manufacturing systems design and analysis
"A static simulation model is a representative of a system at a particular point in time. Models of this kind are defined as having structure without activity, that is to say, the simulation process will not evolve over time."
Dynamic simulation models combine structural components with activity, that is to say, a dynamic simulation model is a representation of a system as it evolves over time.

GSM, GPR and Edge Performance
"Maybe the most important building blocks of this kind of dynamics are simulation time and user mobility which are not relevant in static simulations."

Mobile broadband multimedia networks.
"System simulators can evaluate system performance according to two different methods: static and dynamic simulations. The former provides snapshots of network performance when resource allocation reaches an equilibrium state. The latter [dynamic reproduces system behaviours for a time interval of network working."

Simulation-based lean six-sigma and design for six-sigma
"Finally, and based on the nature of model evolvement with time, models can be static or dynamic. In Static models the system state does not change over time … Most operational models are, however, dynamic. System state variable often change with time, …"

Computer applications in the social sciences
"In constrast to static models, dynamic ones represent a process and operate it over time."

And I can get lots more. All I had to do is go to Google Books and type in ‘static and dynamic simulations'." This was a fairly random selection of quotes and all reference processes or systems. None reference environments.

I can now make the distinction a bit more pithier. Static simulation models take a snapshot of the process. Dynamic simulation models looks at the process over time. Or, static models do not do not include time and dynamic models do.

Game Theory doesn't set up anything but a theory on games and decision-making. SOME game designers use Game Theory to set up real world environments for decision-making, but I don't think that is a requirement of Game Theory.

At this point I have to admit I was wrong about Game Theory. Game theory does not create an environment. Here are some quotes about game theory

Stanford Encyclopedia of Philosophy,
Game theory is the study of the ways in which strategic interactions among rational players produce outcomes with respect to the preferences (or utilities) of those players, none of which might have been intended by any of them.

Principia Cybernetia Web
Game Theory is a branch of mathematical analysis developed to study decision making in conflict situations.

A Course in Game Theory
Game theory is a bag of analytical tools to help us understand the phenomena that we observe when decision-makers interact. The basic assumptions that underlie the theory are that decision malers pursue well-defined edogenous objectives (the are rational) and take into account their knowledge or expectations of other decision-makers' behavior (they reason strategically. … We treat game theory not as a branch of mathematics but as a social science whose aim is to understand the behavior of interacting decision-makers.

Game Theory at Work
Game Theories studies how smart, ruthless people would act and interact in strategic settings.

Game Theory
Game Theory may be viewed as a sort of umbrella ir 'unified field' theory for the rational side of social science. … [it] does not use different, ad hoc constructs. … it develops methodologies that apply in principle to all interactive situations… if one accepts that interaction is the essence of social life, then … game theory provides solid microfoundations for the study of social structure and social change.

Game Theory – Myerson
Game Theory can be defined as thestudy of mathematical models of conflict and cooperation between intelligent rational decision-makers. Game theory provides general mathmatical techniques for analyzing situations in which two or more individuals make decisions that will influence one another's welfare.

A History of Game Theory
For the purpsoes of this book, we define game theory as the (relatively) rigorous analysis of situations of strategic interdependence.

If we consider the situations in which two or more individuals make decisions a real world process, then what game theory does is study that process and creates a model (generally mathematical but doesn't have to be) of that real world process.

If this is correct, then game theory is a simulation just as John's other examples are simulations. And, it reinforces the concept that simulations model real world processes.

Most simulation designers would disagree with you about a process not being an environment.

There is a relationship there but it is not a one to one relationship. An environment is: "the setting or conditions in which a particular activity [process] is carried on."

Now a process is "a series of actions or steps taken in order to achieve a particular end." Obviously this must occur in some kind of environment and modeling a process can set up a kind of environment, however the two are not synonymous. A process is not an environment.

In fact, any process that has an entry/beginning and exit/end, from a roller coaster ride to a computer model of chemical reactions must create an environment with its own internal rules, systems, logic and cause and effect—in other words an environment.

However, creating an environment is quite a bit different from being an environment. A process is not an environment.

We are talking simulation game design, yes?

Actually no. We are talking about the modeling of real world processes, called simulations, and how this relates to wargaming.

I still say that any design meant to simulate, mimic, or recreate a past or present environment is by default a simulation.

Again. I think you are confusing process with environment. The simulation or rather the model design is meant to mimic a process or part of a process not an environment. A particular environment may be created as result of the modeling but it is not the "thing" being modeled.

Let's not get sidetracked by the academic definition of 'history' as exclusively what historians write.

You've got it backward Bill. Anyone who selects certain historical facts to analyze and interpret; who then puts this interpretation on "paper" in narrative form is an historian. This activity began long before academic historians arose. Certain historians would probably challenge me for my limiting history to narrative form. But narrative form captures the essence of the flow of time that I think is integral to history.

*Napoleonic French infantry fought with muskets and
bayonets.

*The Battle of Borodino was fought in Russia. 

*Wellington commanded the Allied army at Waterloo.

If you believe those historical statements are simply interpretive products of some historian, then we really are going to have to agree to disagree.

You've once again confused ‘historical facts ‘with ‘history'. Those facts you wrote are not history. They are not the creative product. They are the groundwork by which the creative product emerges. It is like building blocks and a house. The blocks are not the house but the house is built from the blocks.

Even so, *something* has to exist to be interpreted… That makes it at least half of a historian's work.

What gets interpreted is the relationship that exists among these facts in relation to the flow of time.

I have been endeavoring to establish how that is done, technically, practically, not philosophically. If the claim is recreating history in a game, how is that claim achieved, established, justified, determined and proven? It is a technical claim and demands technical answers.

It seems to me what we have been doing is what must always be done in such a discussion. We have been defining terms. It has been your contention that what simulation designers do have a lot of relevance to what wargame rules writers do because wargames are simulations. To establish if that is true one must determine what simulation designers do when they create a simulation. We then establish what a wargame rule writer does when he creates a wargame, we can then compare the two to see if these two endeavors are doing the same thing. If they do the same thing, then we can say that a wargame is a simulation. If they don't then we can't.

A simulation, it seems, is a model(s) of a real world process. The model may include the passage of time in which case the model is dynamic. That is to say, there is a start time, intervening time, and an end time. Or, the model may not include the passage of time; it may be a snapshot, in which case it becomes a static model. There seems to be two crucial elements in a simulation of this sort. First there must be a real world process. The process doesn't actually have to exist, but is must operate as though it exists in the real world. The creator of the nonexistent process is not the simulator. It may be the same person but that person is not operating as a simulation designer when he creates his real world process. The simulation designer then comes along and creates a model of that created process.

Second, that process may be modeled using logic elements or computational element. Without the elements of real world process and model, there is no simulation.

It has been said that this is not the only definition relative to simulating. That's correct. Those other definitions are not relevant to what simulation designers do. For example, you can simulate a fictitious creature, such as a unicorn, in wood. And then say, "See, I have a simulation of a unicorn." But this is not what simulation designers do. They do not carve wooden unicorns, unless for a hobby in which case they are not operating in the capacity of a simulation designer.

Is this what wargame rule writers do? In my estimation, they do not. To begin with, there is no real world process for them to model. Those processes may have existed in the past but they no longer exist for the rule writer to model. The rule writer must go and research historical evidence and then recreate that process or event. There is no preexisting real world process from which a model can be constructed. Once the wargame rule writer has recreated as much of the process as he thinks important he presents it to the gamer as an environment in which the gamer must make certain behavioral choices (not his behavior but that of his army). This just says, a wargame is a game.

Thus I feel very comfortable in saying that a wargame is not a simulation, that is to say they do not model real world processes. It is an historical interpretive effort that creates an environment, in this case a military environment, in which gamers are forced to make certain decisions. That is to say rule writers create conditions by which the gamer must make certain decisions. Those conditions do not originally exist in real time. One must go back to the past and attempt to recreate those conditions. He does not model, or mimic, those conditions because, until he recreates them they don't exist. Once created, he doesn't model them.

Rich

John:

What, exactly, are you claiming that I have misunderstood?

I wrote, "There is no automatic relationship between simulate and simulation." You wrote, "It's not merely dubious, it's flat out wrong, to claim that "simulate" and "simulation" are unrelated" You were not paying attention and misread my post. I never mentioned they were unrelated. You failed to read what I wrote correctly.

No, it does not reinforce your point. The reason I used the example of the dynamic/static classification was that it was one you had introduced yourself, and therefore presumably understand. It is faulty reasoning to assume that therefore there are no others.

John you are not paying attention. I never said there was one definition. I kept saying this was the definition used in this discussion.

Now, you seem to have decided to create your own defintion of "simulation" that is starkly at variance with those understood in the industry since its inception, and which I have been accustomed to in practice over the last twenty years.

Once again John, you are not paying attention to the previous threads of this topic. It was Bill who said, "What I am referring to is a model of a process or system." It was Bill who repeated what Jerry Banks wrote: "Simulation is the imitation of the operation of a real-world process or system over time." I'm simply using Bill's and Jerry's definition. If you think that's a bad definition, then come up with your own.

Simulation modelling is an enormously broad church, and very few of the people who practice it are aware of the full breadth of the subject. Systems Dynamics folks think it's what they do. Discrete-event simulationists think it's what they do. Electronic engineers who make flight simulators think it's what they do. Chemical engineers, for goodness' sake, think they do it when they solve systems for flow equations. All of them (with the exception of the chemical engineers if you believe that simulations must be dynamic) are right.

John, you're not paying attention to what YOU write. There is a reason why these are simulations. All of them model real world process. I don't think any of them place magic into their models (at least I hope not). Therefore, I don't really see why that definition is so inappropriate.

You did present a definition of simulation when you wrote: "[a] wargame is a simulation because it has a manual, is stochastic and dynamic." Poker has rules, a random probability distribution and constant change. Are you then saying poker is a simulation? See, it's not that easy to come up with a workable definition.

You chastised me for describing a simulation only in terms static and dynamic. You didn't pay attention. I mentioned other terms. The reason I chose those terms was to correct Bill's statement, "A simulation is any system designed to mimic some aspect of reality--and succeeds. It is an artificial construct/model designed to either recreate an event [A static simulation] or an environment [dynamic simulation]" I simply went on to point out that this is not what defines these two types of simulations. I never said they were the ONLY types.

John, you're not paying attention to the context in which I'm writing. Go back and read the previous posts so you will be able to place my comments in their correct context. I didn't invent the definition of simulation. Bill and Jerry Banks did (although I think it's a good one). If you have a better definition the let's see it. I was not the one who singled out dynamic and static simulations. I was simply responding to Bill

Second, I think it is not incumbent upon us to define the discipline so much as it is to describe it.

John, describing what all these people have in common IS defining it.

How you can claim that wargames rules do not seek to model the processes of maneouvre, attrition, command, supply and so on is simply beyond me.

You simply cannot model a process that doesn't already exist. There must be something there to be modeled. Wargame rules do not model anything. These processes must be recreated from the historical record. The gamer then presents them as part of the environment in which the game is to be played.

If you want to call wargames a simulation then it is incumbent upon you to define what you mean by the term ‘simulation'. When you present a term like this, you are implying that every simulation has something in common with all other simulations and that wargames share this common something. Otherwise you have to state what kind of simulation you're talking about. You can't say, "some simulations do this and some do that" because the reader won't know what kind of simulation you're talking about. There must exist a concept that all simulations share. Otherwise no one knows what you are talking about. If no one knows what you are talking about then quit using the word.

Rich

Rich,

First there must be a real world process. The process doesn't actually have to exist, but is must operate as though it exists in the real world.

Is this what wargame rule writers do? In my estimation, they do not. To begin with, there is no real world process for them to model. Those processes may have existed in the past but they no longer exist for the rule writer to model.

Process can be modelled even if it doesn't actually exist, but one that did exist previously can't be modelled? Yes, this is what I would describe as a weird distinction. I really can't understand your reasoning why things like maneouvre, attrition, command or supply can't be modelled.

Model is one of those words that have a plethora of uses. Since the frame of reference is modeling real world processes, when I say these things cannot be modeled, I mean they can't be modeled as simulation designers do when they model processes. Here a model is the creation of only the most important aspects of the process. It is not the recreation of the whole process. Model, then, has a technical aspect to it. It is the extraction of the most important data from a process and put in a form that mimics the process as a whole. Therefore, a model needs a pre-existing process from which to extract the most important data. No pre-existing process, no model.

Process can be modelled even if it doesn't actually exist, but one that did exist previously can't be modelled?

It can't be modeled because it simply doesn't exist for the rule writer to be able to create a model of it. However, physical existence is not required for a process to exist. If the work has never been done, one can still draw up the process for it. That process can later be modeled.

If the process existed in the past, but not now, then the process must be recreated from the historical record. Then a model can be created from it. But this is not what is done. The rule writer does not recreate the process and then make a model of it. He simply transposes the historical data into data he can use in the rules.

In marching, for example, the rule writer does not look at the whole process of marching (whatever that may mean) and then take only the most important aspects of that process to model. What he does is to find some reference to rate of marching and converts that to units of movement for the game.

So, when I say they do not model, I mean they have no pre-existing process from which to extract only the most important aspects. Nor do they recreate that process in order then to create a model.

Rich

Rich:
Why would anybody trying to model a battle give a damn about "process of marching" or how powder burns in a cannon? Such detail is really not needed; there are far more important things that has to be taken into account.

If a mechanical engineer adds an electric motor to his simulated machine, do you think he cares how magnetic fields work in the motor? For him it's mostly just a box with certain dimensions and a rotating axle that gives him certain torque at certain speed. Yet he can build pretty good models of machine systems.

And please, could we keep this discussion about wargames and simulations in here and not spread it too much over to "engineers vs. novelists"-thread?

Are you saying is that someone who simulates a unicorn in wood and someone who simulates a real world processes are, in essence, are doing the same thing? After all, if you are simulating something you are making a simulation. That is so simplistic as to be non-functional.

Rich:
I have said several times that simulation models are systems, processes, such as computer programs and tabletop games. In that sense, it would require a unicorn [or player as unicorn] to operate in an environment. It is simple, and you seem to want to complicate it.

If you attempt to simulate either events or decision-making environments, you will invariably run into the issues that all simulation designers face regardless of the medium.

I wrote: I still say that any design meant to simulate, mimic, or recreate a past or present environment is by default a simulation.
You responded: As mentioned above, this is so overly broad as to be non-functional.

"Functional" is the operative word. The definition becomes very operational when you attempt to create a simulation.

Handbook of Simulation
"Discrete-event models are dynamic; that is, the passage of time plays a crucial role. Most mathematical and statistical models are static, in that they represent a system at a fixed point of time.

Manufacturing systems design and analysis
"A static simulation model is a representative of a system at a particular point in time. Models of this kind are defined as having structure without activity, that is to say, the simulation process will not evolve over time."
Dynamic simulation models combine structural components with activity, that is to say, a dynamic simulation model is a representation of a system as it evolves over time.

Rich, you have misunderstood what Banks is saying, Big Time. Static simulations have one set of inputs and one set of results, which is why he says 'at a particular point in time' without activity. It doesn't evolve over time. A dynamic simulation combines components with activity. As he says, it will evolve [read change] over time, because the activities [unlike static simulations] generate the evolution. What he is saying is that with variables, such as different decisions in that 'activity' creates change--over time obviously.

A static simulation is meant to recreate a particular event at a particular point in time, like a battle, or a factory system. No matter how often you repeat the static simulation, you will always get that one 'particular point in time.'

GSM, GPR and Edge Performance
"Maybe the most important building blocks of this kind of dynamics are simulation time and user mobility which are not relevant in static simulations."
Mobile broadband multimedia networks.
"System simulators can evaluate system performance according to two different methods: static and dynamic simulations. The former provides snapshots of network performance when resource allocation reaches an equilibrium state. The latter [dynamic reproduces system behaviours for a time interval of network working."

Simulation-based lean six-sigma and design for six-sigma
"Finally, and based on the nature of model evolvement with time, models can be static or dynamic. In Static models the system state does not change over time … Most operational models are, however, dynamic. System state variable often change with time, …"

Again, Rich, in a static simulation where only one event is represented. That means for Dynamic Simulations "the building blocks of this kind of dynamics are simulation time and user mobility which are not relevant in static simulations."

Got it? User Mobility and Time are not relevant to static simulations. Why? Because there is neither user mobility or any evolution in processes that require a time monitor.

hat's why he says " Static models the system state does not change over time", but most 'operational models are dynamic, in that systems provide for variables and because of that change with time.

A movie is a static simulation. It provides 'a snapshot of a particular point in time.' It never changes, so there are no structural components that provide 'activities' for participants, researchers, etc. so no need to have a time component clocking variable activities.

Computer applications in the social sciences
"In constrast to static models, dynamic ones represent a process and operate it over time." And I can get lots more. All I had to do is go to Google Books and type in ‘static and dynamic simulations'." This was a fairly random selection of quotes and all reference processes or systems. None reference environments.

Read what Banks says. A dynamic simulation represents a process which is operated [note: operated as in manipulated, played, etc.] OVER TIME. There has to be an internal time component for any operations to that place. The static simulation isn't operated or manipulated within any time structure--it isn't operated. It's a johnny one-note system that is run or not run period.

A movie of Gettysburg does not present a process nor does it operate over time. It just is--an event. That means that if the movie is two hours long, that time has no relation to actual event time in the simulation nor is there a process requiring a time component built in it. [Even '24' fudges on this point.]

Dynamic Simulations DO require a time component within the system that allow participants to act on it within structural processes—which a simulation game of Gettysburg would require. To be pithy, a movie of Gettysburg has no need for a turn/time sequence, a simulation game of Gettysburg does.

I can now make the distinction a bit more pithier. Static simulation models take a snapshot of the process. Dynamic simulation models looks at the process over time. Or, static models do not do not include time and dynamic models do.

Wrong. That is not what Banks was saying, at all. He was saying that the SYSTEM must provide a process over time because the activities will change over time, unlike a static simulation, which doesn't ever change--but does include a time component, but always the same, never varying. You quoted him as saying "Simulation is the imitation of the operation of a real-world process or system over time." Not a dynamic simulation, but a simulation, any simulation. That's all I've been saying….

A static simulations is a single event, a single operation, whereas a dynamic simulation evolves over time, and those has a functioning time component for the varying processes.

There is a relationship there but it is not a one to one relationship. An environment is: "the setting or conditions in which a particular activity [process] is carried on."
Now a process is "a series of actions or steps taken in order to achieve a particular end." Obviously this must occur in some kind of environment and modeling a process can set up a kind of environment, however the two are not synonymous. A process is not an environment.

Keep in mind how this description of 'a process' applies to what you quoted of Bank's descriptions. No, a process isn't an environment. However, if you are simulating a process, attempting to represent a 'real world process', you will be hard pressed not to provide the context for it, you know, that 'real world.'

We are talking simulation game design, yes?
Actually no. We are talking about the modeling of real world processes, called simulations, and how this relates to wargaming.

In fact, such a free-floating process will lose much of it's predictive and process value because it has no context, and can change radically when placed in one.

I still say that any design meant to simulate, mimic, or recreate a past or present environment is by default a simulation.

Again. I think you are confusing process with environment. The simulation or rather the model design is meant to mimic a process or part of a process not an environment. A particular environment may be created as result of the modeling but it is not the "thing" being modeled.

If you feel you can simulate 'real world' processes without ever representing the 'real world', you go for it.

Let's not get sidetracked by the academic definition of 'history' as exclusively what historians write.

You've got it backward Bill. Anyone who selects certain historical facts to analyze and interpret; who then puts this interpretation on "paper" in narrative form is an historian. This activity began long before academic historians arose. Certain historians would probably challenge me for my limiting history to narrative form. But narrative form captures the essence of the flow of time that I think is integral to history.

From my understanding, basic to historiography is the idea that there is History A and History B. History A is what happened, the facts, data, opinions and relics that remain, including folks' opinions on what happened. History B, starting with Herodotus, is the narratives that historians weave of such information. I am simply not willing to include all of History under 'B', however integral it is to the study and expression of history.

*Napoleonic French infantry fought with muskets and
bayonets.

*The Battle of Borodino was fought in Russia. 

*Wellington commanded the Allied army at Waterloo.
If you believe those historical statements are simply interpretive products of some historian, then we really are going to have to agree to disagree.

You've once again confused ‘historical facts ‘with ‘history'. Those facts you wrote are not history. They are not the creative product. They are the groundwork by which the creative product emerges. It is like building blocks and a house. The blocks are not the house but the house is built from the blocks.

I am not confused with the two. Again, Rich, I am not willing to place all of history under "B", regardless if it's importance. Those building blocks, as you call them, are the same material that historical wargame designers use to build their 'houses,' that creative product you speak of. For me, that isn't 'history', but someone's interpretation of the facts of history. My world includes History 'A' and History 'B', both the raw materials and the houses.

It seems to me what we have been doing is what must always be done in such a discussion. We have been defining terms. It has been your contention that what simulation designers do have a lot of relevance to what wargame rules writers do because wargames are simulations.

[snip a definition of simulations and wargames that I don't recognize.]

Is this what wargame rule writers do? In my estimation, they do not. To begin with, there is no real world process for them to model. Those processes may have existed in the past but they no longer exist for the rule writer to model. The rule writer must go and research historical evidence and then recreate that process or event. There is no preexisting real world process from which a model can be constructed. Once the wargame rule writer has recreated as much of the process as he thinks important he presents it to the gamer as an environment in which the gamer must make certain behavioral choices (not his behavior but that of his army). This just says, a wargame is a game.

This statement demonstrates a real lack of understanding concerning how simulation designers often operate: They have to create a simulation of a process, an environment, a system that does not exist when they have to design the simulation. The exact same problem faced by a wargame designer.

This notion that simulation designers can ONLY simulate existing real world processes would negate about two-thirds of the reasons for creating simulations… If a wargame rules writer is 'recreating as much of the process as he thinks important', then he is doing exactly what a simulation designer does.

Thus I feel very comfortable in saying that a wargame is not a simulation, that is to say they do not model real world processes. It is an historical interpretive effort that creates an environment, in this case a military environment, in which gamers are forced to make certain decisions. That is to say rule writers create conditions by which the gamer must make certain decisions. Those conditions do not originally exist in real time. One must go back to the past and attempt to recreate those conditions. He does not model, or mimic, those conditions because, until he recreates them they don't exist. Once created, he doesn't model them.

If those conditions in a wargame 'didn't originally exist in real time' at some point in some fashion, then what is history? And if they can't be modeled—then you have not only asserted that most all wargame designers, both board and tabletop have failed to accomplish many of the basic goals they claim for their historical wargames, but that what they are attempting is impossible: i.e. to simulate, recreate, or model a past reality in some fashion.

You have also are also claiming the same thing about a large percentage of professional simulation designers.
Rich, when you say "he does not model, or mimic, those conditions because, until he recreates them they don't exist. Once created, he doesn't model them" you have just defined most all functional simulations out of existence.

To say 'if the real environment doesn't exist at this moment, then it can't be simulated' is to deny the validity of most all simulations, even when they have proven to do just what you say they can't.

Banks makes several observations as do others in his edited "Handbook" about simulating things that 'don't yet exist' or have existed in the past. It is one of the major, practical applications of simulations, for Pete's sake. Simulations can and do have real value when they simulate or model environments that have or do not yet exist. They do it all the time!

Rich, Jerry Banks is describing something that is DONE all the time, regardless of how you define it…

To date, wargame designers certainly haven't been simulating military history or warfare, though most all of them claim to be doing it in some fashion—that claim is a fundamental attraction of historical wargames. They could be actually simulating if they wanted to.


Best Regards,
Bill H.

Rick Knapton wrote:

I wrote, "There is no automatic relationship between simulate and simulation." You wrote, "It's not merely dubious, it's flat out wrong, to claim that "simulate" and "simulation" are unrelated" You were not paying attention and misread my post. I never mentioned they were unrelated.

It remains entirely unclear to me what you mean by an "automatic" relationship. However, I don't think it much matters whether the manner in which things are related is automatic, manually-assisted, or operated by little green goblins; your argument was plainly attempting to show that the verbal and nominal forms were in some way not related. So, since you have not deigned to enlarge upon what you mean by the qualification "automatic", I shall content myself with saying that, to the extent that you claim that they are unrelated, to that extent, your argument is wrong.

Still, since you don't seem to have offered any counter-argument to my point that meaning cannot be adduced from what part of speech a word is, I take it that you have conceded the main point here.

You failed to read what I wrote correctly.
[…]
John you are not paying attention.
[…]
Once again John, you are not paying attention to the previous threads of this topic.
[…]
John, you're not paying attention to what YOU write.
[…]
You didn't pay attention.
[…]
John, you're not paying attention to the context in which I'm writing.

I'd like to see a little less of this kind of nonsense in your postings, Rich. You make a mistake if you assume that I am not paying attention merely because I disagree with you. You become a very dull interlocutor when you feel the need repeat something six times in a single posting. Argument works better than repetition.

It is especially galling when you can't even be bothered to pay attetion to the context of the thread yourself, and come out with:

You did present a definition of simulation when you wrote: "[a] wargame is a simulation because it has a manual, is stochastic and dynamic."

…which is not something I have ever written, in this thread or anywhere else.

Now, back to the definitions of things: First, defintion itself:

John, describing what all these people have in common IS defining it.

Clearly untrue -- I can describe something by saying that it is blue and two metres tall, but I do not thereby define it. A description of the common features of a set of things is not sufficient to define whether a new thing belongs to that set or not based on a new set of features. Definition is stricter than description.

Next, the static/dynamic thing.

I never said there was one definition. I kept saying this was the definition used in this discussion.

Your exact words were:

"However, I noticed that the only two terms you used were static and dynamic. This rather reinforces my point. There may be combinations of the two but there are only two forms static or dynamic."

Perhaps you meant to say that the classification into static and dynamic is an exhaustive one, which is true enough. But the point I was making was that many other classifications exist. If you are now happy to accept that, as apparently you are, fine.

[Snips]
It was Bill who said, "What I am referring to is a model of a process or system." It was Bill who repeated what Jerry Banks wrote: "Simulation is the imitation of the operation of a real-world process or system over time." I'm simply using Bill's and Jerry's definition. If you think that's a bad definition, then come up with your own.

No need. I think Karsta's definition, posted early in this thread, was good enough for all practical purposes. It's a defintion you have rejected, for the odd reason that it is "too broad". I always think that generality is a useful property of definitions, rather than the reverse.

Incidentally, out of historical interest, I checked my copy of Keith Tocher's "The Art of Simulation Modelling", which I believe was the first book ever to be published on the subject of discrete-event simulation. The great man seems to see no need to define the term "simulation"; he gives a brief account of its historical origins, and then gets right on with showing people how to do it.

There is a reason why these are simulations. All of them model real world process.

No, this hang-up about "the real world" means that you are about 180 degrees out in understanding simulation. I've already pointed out that I have, in my job as a professional simulation modeller, many times written simulations of processes that have no "real-world" existence -- something you've chosen to ignore and flatly contradict, but not, I notice, something you have presented any counter-argument to. That's not a frightfully convincing mode of argumentation, and if you can't do better, then you might "Think, in the bowels of Christ, that you might be wrong".

The essence of simulation modelling is asking the question "what if"? Simulationists like Jerry Banks with a queueing systems focus deal with questions like "What if we replace the two low-rate flange-pressers with one high speed one? Combat simulation folks deal with things like "What if we invest in improved fire-control for out MBT fleet? Or would add-in armour be better?" Flight simulationists ask "What if you were piloting a Nimrod out of Lossiemouth and the following flight emergency occurred?"… and so on, and so forth. The common thread in all sorts of simulation is its counter-factual nature.

When giving basic lectures on the subject, I often like to quote the dialogue "Contrafactus", from Douglas Hofstadter's wonderful book "Godel, Escher, Bach". This introduces the Subjunc-TV, a device quite like an ordinary TV, except that when you watch the action replays, you can twiddle knobs to follow different hypothetical paths -- the subjunctive instant replay shows you what might have happened if Tedzillinger hadn't stepped out of bounds, or if Palindromi hadn't been tackled. Alan Bennet makes similar mention of the idea of "subjunctive history" in his splendid play "The History Boys", which I've recently watched on DVD. This idea of subjunctivity applies to wargames, too, except perhaps for the very restricted set of fixed-outcome games. Remember the old AH game ads asking if you could do better than Rommel?

Usually, the motivation for using simulation is (and I'm quoting from memory here, but from Isi Mitrani's "Simulation techniques for dicrete-event systems", my first and favourite simulation book) that studying the real system is "too difficult, too expensive, or too hazardous" (and I'm sure much the same applies even if the intent is entertainment rather than study). The underlying assumption is (and I think it was Ian Stewart who first made me aware of this in his "Pi in the Sky") that if we can obtain precise results from an approximate representation of a system, these will be just as good as approximate results from a precise representation of the system". It sounds as if it ought to be so, but unfortunately there's no way of proving it.

Given that simulation is always an exercise in the counter-factual, it does not actually make any essential difference to simulation itself whether the method is applied to the study of things that once existed, things that might have once existed, things that exist now, things that are expected to exist in the future, or things that never could exist, like unicorns.

So,

You simply cannot model a process that doesn't already exist.

is simply wrong, and about as thoroughly wrong as it is possible to be.

However

There must be something there to be modeled.

is perfectly reasonable. The objective system, system under study, or whatever you want to call it, need not have real existence. Indeed, if you read somethign like the introductory chapters of Weinberg's "Introduction to General Systems Theory", you will undertsand that a "system" has no "real-world" existence, but is merely a framework placed over our perceptions of the world. If you want to go deeper still, there's the sceptical epistemology argument from people like Peter Unger, but I think there's no need to go that far, and we can cheerfully assume that the external universe does really exist much as we sense it.

Wargame rules do not model anything. These processes must be recreated from the historical record. The gamer then presents them as part of the environment in which the game is to be played.

So, in this case, wargame rules model the processes extracted from the historical record, no? We have an objective system, and a simulation of it that can be manipulated by the users.

I don't think any of them place magic into their models (at least I hope not).

I often think that simulation modelling has got a lot more in common with sympathetic magic than most practitioners would care to admit. Should you ever encounter a strange and interesting book called "Sex Secrets of the Black Magicians Explained", which is worth a quick read, you will notice that the author repeatedly points out that a good magician should not believe his own magic. Likewise, a professional simulationist should never believe his own simulations.

All the best,

John.

The Scotsman wrote:

A movie is a static simulation.

Not using the classifications I'm used to. It's a dynamic simulation (assuming that the actors do actually move and speak, anyway), but a deterministic one.

Of course, if your definition of "simulation" requires evolution over time, then "static simulation" is a bit of a nonsense.

To date, wargame designers certainly haven't been simulating military history or warfare

Why do you say that? What element of simulation have they been missing?

All the best,

John.

Bill:

I have said several times that simulation models are systems, processes, such as computer programs and tabletop games.

Yes and you are wrong. The model is not a system or process. The model is a model of that system or process.

Both John, who builds industrial simulations, and I agree your definition is static and dynamic is simply wrong. What I think you are thinking about is deterministic simulation and stochastic simulation. A deterministic model is such that every time it is run it will return the same value. Not so a stochastic simulation. "Whichever way one chooses to simulate them there will be a stochastic component, namely it will be based on probability distributions. … The starting point for stochastic simulaltion is the construction of a random number generator." Marckov Chain Monte Carlo by Gamerman and Lopes

Both static and dynamic simulations can be deterministic or stochastic. That is to say a static model, in which time is not an aspect of the model, can be deterministic and return the same value or stochastic in which change is built into the simulation.

If you feel you can simulate 'real world' processes without ever representing the 'real world', you go for it.

To be more accurate one makes a model of the real world process. If I have created a model of a real world process then by definition the process exists. Otherwise, I cannot make a model of it.

From my understanding, basic to historiography is the idea that there is History A and History B. History A is what happened, the facts, data, opinions and relics that remain, including folks' opinions on what happened. History B, starting with Herodotus, is the narratives that historians weave of such information. I am simply not willing to include all of History under 'B', however integral it is to the study and expression of history.

Prior to Herodotus there were chronicles, poems, triumphalist writings (the chronicling of success by the king), economic data, religious writings, laws, etc. but no History. I think what you may be thinking of is "the historical record.": that information by which historians create history. You're A is not history. It is the historical record. We can't go around creating our own definition for things.

This statement demonstrates a real lack of understanding concerning how simulation designers often operate: They have to create a simulation of a process, an environment, a system that does not exist when they have to design the simulation.

Wrong again Bill. Designers set out to create a model of a process. An environment may be a bi-product of the model but unless the environment is directly germane to the functioning of the process, they do not set out to model any environment.

This notion that simulation designers can ONLY simulate existing real world processes would negate about two-thirds of the reasons for creating simulations.

What in the world are you talking about? There is only one reason to build a model of an existing real world process. It is to provide management a tool to do "what if" scenarios. You may want to do models of non-existing process to see if your process is like to perform as expected. That's it.

If a wargame rules writer is 'recreating as much of the process as he thinks important', then he is doing exactly what a simulation designer does.

Once again you're confused. Simulation designers do not create processes. They create models of that process.

Rich, when you say "he does not model, or mimic, those conditions because, until he recreates them they don't exist. Once created, he doesn't model them" you have just defined most all functional simulations out of existence.

Only because you fail to make the distinction between a process and a model. System designers create systems (processes). Simulation designers create models of those systems.

Banks makes several observations as do others in his edited "Handbook" about simulating things that 'don't yet exist' or have existed in the past.

I don't have a problem with that. I'm simply saying that is not what the rule writer does. He doesn't first work as a system designer and build the process, then go back as a simulation designer and model that process.

So where are we now?

* Static simulations do not deal with time.
* Dynamic simulations include time in the model.
* A deterministic model always returns the same value unless the model itself is changed.
* A stochastic simulation has randomness built into it. (John may want to fine-tune this)
* Simulation is what we call it when a model is created from a process.
* A process is different from the model of that process.
* Simulation designers do model environment although an environment may be a by-product of a model
* There is no History A. It's called the historical record.

Rich

John S.

It could be that you and I learned different definitions of the same thing. There are a lot of simulation professionals out there in a lot of different disciplines and I learned my stuff back a decade ago.

The reason that a static simulation is static and not dynamic is the lack of any time monitoring, not movement. Play a movie several times and you get the very same simulation action with the very same results. Static. A dynamic simulation, because it actually as an interface with a user through activities DURING the simulation, it requires a time monitoring process. Time becomes the spine on which the system operates.

My understanding of static and dynamic has to do with how the simulation system functions, only one way, static, or able to provide a host of variables and alternative inputs, dynamic. Obviously, a computer system can be re-programed to do both, but that's the way I learned it…

Bill

Yes and you are wrong. The model is not a system or process. The model is a model of that system or process.

Rich:
Honestly Rich, I can't believe you wrote that. If a system is being modeled, wouldn't the model have to operate as and for all intents be 'a system' itself? This is hairsplitting with a vengeance. Of course the model and the thing modeled are different.

Both John, who builds industrial simulations, and I agree your definition is static and dynamic is simply wrong. What I think you are thinking about is deterministic simulation and stochastic simulation. A deterministic model is such that every time it is run it will return the same value. Not so a stochastic simulation. "Whichever way one chooses to simulate them there will be a stochastic component, namely it will be based on probability distributions. … The starting point for stochastic simulaltion is the construction of a random number generator." Marckov Chain Monte Carlo by Gamerman and Lopes

Rich:
As I told John, it could be we are talking about the same thing with different terms, and I know Jerry Bank's work, but there are a lot of folks out there simulating and thinking up new ways to define things. [Present company included.]

Part of my experience is that a random number generator was not something needed in my simulations for the most part, so
I didn't use 'stochastic component' as required to have a stochastic simulation.

To be more accurate one makes a model of the real world process. If I have created a model of a real world process then by definition the process exists. Otherwise, I cannot make a model of it.

Well, hello circular reasoning. And this gets us where?

This statement demonstrates a real lack of understanding concerning how simulation designers often operate: They have to create a simulation of a process, an environment, a system that does not exist when they have to design the simulation.

Wrong again Bill. Designers set out to create a model of a process. An environment may be a bi-product of the model but unless the environment is directly germane to the functioning of the process, they do not set out to model any environment.

So, if it is a 'real world' process, Am I safe in assuming the environment IS germane to the process?

This notion that simulation designers can ONLY simulate existing real world processes would negate about two-thirds of the reasons for creating simulations.

What in the world are you talking about? There is only one reason to build a model of an existing real world process. It is to provide management a tool to do "what if" scenarios. You may want to do models of non-existing process to see if your process is like to perform as expected. That's it.

So no one designs simulations of weather patterns for 2050 or crowd flow patterns for building that do not exist, or simulations of biological functions that haven't been determined yet, or the movement and musculature of animals that haven't existed for millions of years, etc. etc. etc.?

So no one does 'what if' scenarios concerning anything but
"an existing real world process?"

If a wargame rules writer is 'recreating as much of the process as he thinks important', then he is doing exactly what a simulation designer does.

Once again you're confused. Simulation designers do not create processes. They create models of that process.

Well, once again we are speaking of game processes not being actual processes, but only models of processes which then model with processes. Hope that is clear.

Rich, when you say "he does not model, or mimic, those conditions because, until he recreates them they don't exist. Once created, he doesn't model them" you have just defined most all functional simulations out of existence.

Only because you fail to make the distinction between a process and a model. System designers create systems (processes). Simulation designers create models of those systems.

Yep, I did fail to make that distinction. I always found that the model of a process usually ends up being a process too. Silly me.

Banks makes several observations as do others in his edited "Handbook" about simulating things that 'don't yet exist' or have existed in the past.

I don't have a problem with that. I'm simply saying that is not what the rule writer does. He doesn't first work as a system designer and build the process, then go back as a simulation designer and model that process.

What? Rich, I thought you insisted that it was impossible to simulate something that didn't exist in the real world. You lost me here.

So where are we now?
* Static simulations do not deal with time.
* Dynamic simulations include time in the model.

I can agree if we are saying a static simulation doesn't have a method for monitoring time for activities within it, while dynamic simulations do.

* A deterministic model always returns the same value unless the model itself is changed.
* A stochastic simulation has randomness built into it. (John may want to fine-tune this)

Again, I think we are talking about the same qualities with different definitions.

* Simulation is what we call it when a model is created from a process.
* A process is different from the model of that process.
* Simulation designers do model environment although an environment may be a by-product of a model.

Yes, a process is distinct from the model of that process, if only because the model/simulation is artificial in nature. Regardless, to work, the model has to be a process too…

* There is no History A. It's called the historical record.

Right. No History A. I just heard the A and B history description repeated by Dr. Elizabeth Vandiver of the University of Maryland in a lecture on Herodotus. I'll update her and my history professors as long as we can agree that "the historical record" is separate from historians' narratives using that record.

Now, about designing simulation games…

Bill H.

your argument was plainly attempting to show that the verbal and nominal forms were in some way not related.

Even in the Queen's English, to say that there is no automatic relationship implies there is a relationship of some form it's just not automatic. So, when you claimed that I said there was no relationship you were wrong. I can't make it any plainer than that.

Still, since you don't seem to have offered any counter-argument to my point that meaning cannot be adduced from what part of speech a word is, I take it that you have conceded the main point here.

No. You can take it that I find it silly to continue discussing this point.

I'd like to see a little less of this kind of nonsense in your postings, Rich. You make a mistake if you assume that I am not paying attention merely because I disagree with you. You become a very dull interlocutor when you feel the need repeat something six times in a single posting. Argument works better than repetition.

It looks like 6 times was not enough. You still didn't get the point. I don't like to be misquoted or assumed that I made arguments I never made. I would rather mark that up to inattention than purposefully misquoting me to try to make points or that you are having mental problems of some sort. However, the next time you do it I'll come up with something different although I can't guarantee you'll like it.

Rich: You did present a definition of simulation when you wrote: "[a] wargame is a simulation because it has a manual, is stochastic and dynamic."

John…which is not something I have ever written, in this thread or anywhere else.

You absolutely correct. I was taking liberties with your direct quote. Nevertheless, it seemed to me you did present that as a definition of a simulation. Your direct quote: "Yet sor some peculiar reason, some wargamers, who most certainly do do (manual, stochastic, dynamic) simulation by any defintion, are dogged in their refusal to accept that it's what they do."

I can describe something by saying that it is blue and two metres tall, but I do not thereby define it.

I'm sorry but I failed read where you said those people had something blue and two meters tall in common.

But the point I was making was that many other classifications exist. If you are now happy to accept that, as apparently you are, fine.

Since I had already mentions other classifications such as deterministic and stochastic I find your comment a bit silly. Let's move on.

No, this hang-up about "the real world" means that you are about 180 degrees out in understanding simulation. I've already pointed out that I have, in my job as a professional simulation modeller, many times written simulations of processes that have no "real-world" existence

John, I'm becoming real concerned about you. This forgetfulness of what I've written could mean the early onset of dementia. You better have it checked out. You see I already wrote: "First there must be a real world process. The process doesn't actually have to exist, but is must operate as though it exists in the real world."

Rich: You simply cannot model a process that doesn't already exist.

John: is simply wrong, and about as thoroughly wrong as it is possible to be.

John: However

Rich: There must be something there to be modeled.

John: is perfectly reasonable. The objective system, system under study, or whatever you want to call it, need not have real existence.

OK I'll give you that one. What I should have said was "You simply cannot model a process that doesn't already exits in some form or other". I stand corrected.

The essence of simulation modelling is asking the question "what if"?

I think you left out a very important fact. It is created as a management decision-making tool. Jerry Banks doesn't create "what if simulations" for the hell of it. It helps decision makers to learn, to some extent, the implications of change. But this is not the only way to play the "what if" game. Historians can also play that game but with very different tools. In fact that has become very popular lately.

So, in this case, wargame rules model the processes extracted from the historical record, no?

No! What are the processes that are extracted and how are they modeled? Take for example marching rates. We read in Delbruck that an army can march at a certain rate per day. The rule writer converts that rate to what it would be for the time period of each turn. If the rule writer determined that each turn represents 15 minutes of real time, he must convert the day rate into a 15minute rate. Say Delbruck says an army can march 20 miles a day (I don't know what he really said). The rule writer then asks how far the army would march in 15 minutes. If the army marches 10 hours a day, this means in 15 minutes it could cover half a mile. To this point we haven't modeled a thing. All we've done is convert a 20mile a day rate into a 15 minute rate. Now the rule writer must determine what represents half a mile on the tabletop. He now transposes the half-mille rate into a per-turn rate. Say he determines that this represents 10 inches. In his rules he states that a unit can march 10 inches per turn. (This is all hypothetical. Any relation to real rates is purely coincidental.) He has yet to model anything. All he has done is transpose the rate of march per 15 minutes, into inches per turn. He is taking the real world rate and transposing into the gaming rate. This is what he does for the whole game system. This process of transposing creates an environment (conditions) in which opposing players need to make certain decisions. I believe it was von Neumann who defined a game as a set of players, a set of moves [as defined by the environment or conditions of the game] and a payoff. Wargaming adds one more component. It is played for entertainment. In other words, a game is not a model. If it is not a model then it is not a simulation.

Rich

Rich Knapton wrote:

your argument was plainly attempting to show that the verbal and nominal forms were in some way not related.

Even in the Queen's English, to say that there is no automatic relationship implies there is a relationship of some form it's just not automatic.

No, wrong. Basic propositional logic tells you this. If there are no bunnies at all in the garden, then there are obviously no pink bunnies. Similarly, if there is no relationship at all, then there is clearly no automatic one. So while you may draw the inference you say, it's not one that would be drawn by somebody unaccustomed to sloppy thinking.

So, when you claimed that I said there was no relationship you were wrong. I can't make it any plainer than that.

Perhaps you would like, for the second time of asking, to make clear what you intend by "automatic". It makes no sense to me. Perhaps you prefer not to be pinned down, for fear of having your argument refuted; in which case it's not really on for you to be quite as hysterically repetitive as you have been about my not understanding you, when your thoughts are expressed in such a fudged and incoherent manner.

Incidentally, since you seem to prefer undergraduate logic-chopping to reasoned debate, you might care to notice that I did not say that you said there was "no relationship". I said that to claim there was no relationship would be wrong. I assume from your reaction that you agree. Good; that's my point made.

Still, since you don't seem to have offered any counter-argument to my point that meaning cannot be adduced from what part of speech a word is, I take it that you have conceded the main point here.

No. You can take it that I find it silly to continue discussing this point.

Good again. Another point settled.

I'd like to see a little less of this kind of nonsense in your postings, Rich. You make a mistake if you assume that I am not paying attention merely because I disagree with you. You become a very dull interlocutor when you feel the need repeat something six times in a single posting. Argument works better than repetition.

It looks like 6 times was not enough. You still didn't get the point.

And you think that bone-headed repetition is a promising method to get your point over, do you? Even the Bellman from "The Hunting of the Snark" was happy to leave it at three times. Try a reasoned argument, Rich, it really will work much better.

I don't like to be misquoted or assumed that I made arguments I never made.

Nor do I. If I've ever done either, then I apologise. However, I am reasonably certain that I have never misquoted you. If I have, please say where. In fact, I insist that you say where, or withdraw the accusation.

As to arguments you haven't made, please bear in mind that I'm having a good deal of difficulty in working out what your arguments actually are. For the removal of doubt, let me just point out that I have been professionally employed as a simulation modeller these past 20 years, hold a doctorate in the subject (supervised by the first professor of simulation modelling in the UK), and have three times been invited to give keynote speeches at international simulation conferences. While I do not approve of arguments from authority, you may do well to assume that if we are failing to understand each other, it is not because of any lack of understanding of simulation on my side. Got that?

I would rather mark that up to inattention than purposefully misquoting me to try to make points or that you are having mental problems of some sort. However, the next time you do it I'll come up with something different although I can't guarantee you'll like it.

Again, where have I misquoted you?

And spare me the sarcasm. You really can't do it properly that side of the Pond. And you wouldn't like me when I get sarcastic.

Rich: You did present a definition of simulation when you wrote: "[a] wargame is a simulation because it has a manual, is stochastic and dynamic."

John…which is not something I have ever written, in this thread or anywhere else.

You absolutely correct. I was taking liberties with your direct quote.

"Taking liberties" is right. "Completely misrepresenting" would be right, too -- I couldn't even guess what you were referring to.

Nevertheless, it seemed to me you did present that as a definition of a simulation.

What earthly reason could you possibly have had for doing such a manifestly absurd thing?

Your direct quote: "Yet sor some peculiar reason, some wargamers, who most certainly do do (manual, stochastic, dynamic) simulation by any defintion, are dogged in their refusal to accept that it's what they do."

I'm damned if I can see anything there that says, or even mildly hints at, the idea of this being offered as a definition. All it is is a classification of a wargame as a particular type of simulation, using three of the classification axes I've mentioned. And "manual" there means a manual simulation as opposed to a machine simulation, it doesn't mean "has a manual". Twit.

I can describe something by saying that it is blue and two metres tall, but I do not thereby define it.

I'm sorry but I failed read where you said those people had something blue and two meters tall in common.

Your syntax seems to have broken down. Would you like to try expressing that thought again?

[Snips]

No, this hang-up about "the real world" means that you are about 180 degrees out in understanding simulation. I've already pointed out that I have, in my job as a professional simulation modeller, many times written simulations of processes that have no "real-world" existence

John, I'm becoming real concerned about you. This forgetfulness of what I've written could mean the early onset of dementia. You better have it checked out. You see I already wrote: "First there must be a real world process. The process doesn't actually have to exist, but is must operate as though it exists in the real world."

I'm more than a little baffled by the idea of a "real world" thing that "doesn't actually have to exist". But I think…

John: is perfectly reasonable. The objective system, system under study, or whatever you want to call it, need not have real existence.

OK I'll give you that one. What I should have said was "You simply cannot model a process that doesn't already exits in some form or other". I stand corrected.

…that means we're getting nearer to sorting things out.

For the next trick, I would need a better idea of the meaning you put on "process". I see no particular need for the thing to be simulated to be a process. In agent-based modelling, and the sort of stuff they like at the Santa Fe insitute, it is possible to specify the rules governing the atomic behaviour of an agent or cellular automaton, and then watch as the higher-level behaviours emerge (you might find things of interest googlign for Sugarscape or Boids).

The essence of simulation modelling is asking the question "what if"?

I think you left out a very important fact. It is created as a management decision-making tool. Jerry Banks doesn't create "what if simulations" for the hell of it. It helps decision makers to learn, to some extent, the implications of change. But this is not the only way to play the "what if" game. Historians can also play that game but with very different tools. In fact that has become very popular lately.

As to the hypothetical history -- yes, exactly (but do you consider this "real history", or in the same class as the Napoleon was a Swiss punk rocker?). However, Jerry Banks' motivation (and mine, when I'm at work) for creating simulation models is not the only possible one. As well as decision support, simulations may also be created for training purposes -- and a great deal of annoyance and confusion can result when simulations intended for one purpose get used for the other. And simulations can also be created for fun. There may be other motivations, but those seem to me to be the big three. And, although we tend to think of wargames as being "just for fun", they have been created both for training and for decision support.

So, in this case, wargame rules model the processes extracted from the historical record, no?

No! What are the processes that are extracted and how are they modeled? Take for example marching rates. We read in Delbruck that an army can march at a certain rate per day. The rule writer converts that rate to what it would be for the time period of each turn.

Well, there's your process right there. We'll call it "Marching". Extracting the process was so easy for you, you did it without conscious thought -- and you've already made assumptions about the time advance mechanism so that you can model it using your chosen representation. It's a remarkably simple process, at the moment; all it does is start, change the value of a positional variable at a fixed rate, and stop, but it's recognisably a process. Probably it will get a bit more complicated when it can be interrupted by opportunity fire, varied in rate by terrain or formation, or compulsorily halted every hour for rest breaks. Some designers might make the move distance dependent on a dice roll, and so make it a stochastic process; all will have to have some rule about what entities can participate in it, because moving troops is OK, but moving terrain pieces usually not. When we get on to modelling the command processes, we might rule that certain conditions, possibly involving the dreaded command radius, must be satisfied before the movement process can be started or stopped, and so we start dealing with process interactions.

If you don't see that, then you must have some rich and strange definition of "process" that is different from any of the three I am used to. You'll have to say what it is if you expect to be understood. As ever, you stand a better chance if you adopt a widely-recognised existing defintion instead of making up your own.

This process of transposing creates an environment (conditions) in which opposing players need to make certain decisions.

But it's not mere transposition. Even in your brief worked example, a shedload of essential modelling decisions have been made -- even if many of them were probably unconscious.

I believe it was von Neumann who defined a game as a set of players, a set of moves [as defined by the environment or conditions of the game] and a payoff. Wargaming adds one more component. It is played for entertainment. In other words, a game is not a model. If it is not a model then it is not a simulation.

The words "game" and "model" clearly mean different things; but that does not make them mutually exclusive categories, and nothing in the definitions you've given states or implies that they are. I see no reason that something can't be both a game and a simulation, which would be why there are hundreds of boxes on my shelves with the words "A conflict simulation game" written on them.

All the best,

John.

The Scotsman wrote:

John S.

It could be that you and I learned different definitions of the same thing.

Almost certainly. I have been fed numerous different definitions of all sorts of things. I'm just strangling the memory glands of my brain to see if I can recall anyone using the ones you use in my hearing, and I can't -- but then I never seem to go to Jerry Banks' presentations on the too-rare occasions I get to WinterSim.

This is why I prefer not to be too dogmatic about definitions of terms. You'll no doubt notice that I didn't say your defintion was wrong (although Rich decided to take a short holiday from his concern about accurate quotation to claim that I did), just that it didn't match what I was used to.

There are a lot of simulation professionals out there

The thing that depresses me is that there are a hell of a lot of people writing and using simulations, and not so many of them are simulation professionals. ;-)

in a lot of different disciplines and I learned my stuff back a decade ago.

I stated to learn mine two decades ago, and I'm not finished yet, not by a long way.

Have you seen Paul Fishwick's book with the elephant and the monkeys on the cover (I don't recall the title right now)? It's the best thing I've yet seen for covering the sheer variety of things that fall under the rubric of "simulation".

The reason that a static simulation is static and not dynamic is the lack of any time monitoring, not movement. Play a movie several times and you get the very same simulation action with the very same results. Static. A dynamic simulation, because it actually as an interface with a user through activities DURING the simulation, it requires a time monitoring process. Time becomes the spine on which the system operates.

OK, so that seems different again from the distinction I thought was intended. Whether or not a simulation permits interaction with the user is a classification axis I would label "interactive/batch", whereas getting the same result or different results from each run because of (pseudo-)random effects interval to the simulation is the "determinsitic/stochastic" difference I originally thought you meant.

And, weirdly, neither of these is particularly to do with time. It's possible to have (using my classification here) a static representation of time, at least for steady-state systems, using techniques such as finding the balance equations of Markov chains. It's also possible to have a dynamic representation of a system where time is not represented, as in a basic Petri net. Which of these would you class as static and which dynamic using your classification scheme?

I ask merely out of idle curiosity, of course. Over the years I have come to realise that simple-minded classifications like this don't get you very far. They are like the Royal Navy's classification of all birds as sparrows, ehawks, or arse-up ducks, which isn't too helpful to the serious birder. There are a whole bunch of different levels of statis to be recognised -- most people would count a stochastic discrete-event simulation using Poisson arrivals as pretty dynamic, but if it uses a stationary Poisson process it isn't as dynamic as it could be.

All the best,

John.

Thank you for your contribution John.

All the best.

Rich

Honestly Rich, I can't believe you wrote that. If a system is being modeled, wouldn't the model have to operate as and for all intents be 'a system' itself? This is hairsplitting with a vengeance. Of course the model and the thing modeled are different

I'm sorry you think that was hair splitting. It was not intended to be. If we start to call both the system and the model "processing", it presents yet another communications problem. That's all I was saying. I don't play "gotcha" games.

Rich: To be more accurate one makes a model of the real world process. If I have created a model of a real world process then by definition the process exists. Otherwise, I cannot make a model of it.

John: Well, hello circular reasoning. And this gets us where?

What I meant by that was, if I have a model of a process this presupposes that the process itself exists.

So, if it is a 'real world' process, Am I safe in assuming the environment IS germane to the process?

If the model is a mathematical model, the fact that the process is under one roof is irrelevant.

So no one designs simulations of weather patterns for 2050 or crowd flow patterns for building that do not exist, or simulations of biological functions that haven't been determined yet, or the movement and musculature of animals that haven't existed for millions of years, etc. etc. etc.? So no one does 'what if' scenarios concerning anything but 
"an existing real world process?"

All those are real world processes. As I said previously, they don't have to physically exist to be a real world process.

Well, once again we are speaking of game processes not being actual processes, but only models of processes which then model with processes. Hope that is clear.

Well, no. It is not clear. What is a game processes from which a model is created?

Yep, I did fail to make that distinction. I always found that the model of a process usually ends up being a process too. Silly me.

Again, this is not a "gotcha". I think for the sake of our discussion it is important to make the distinction between those who design processes and those that design models of that process, even if it is the same person. They are two distinctly different functions.

What? Rich, I thought you insisted that it was impossible to simulate something that didn't exist in the real world. You lost me here.

As I have said, a real world process does not have to exist physically. It can exist on paper or, more likely, in a computer. It doesn't even have to exist there. It can be as simple as the laws that pertain to man's interactions with nature given the tools he has to work with.

Right. No History A. I just heard the A and B history description repeated by Dr. Elizabeth Vandiver of the University of Maryland in a lecture on Herodotus. I'll update her and my history professors as long as we can agree that "the historical record" is separate from historians' narratives using that record.

Feel free to send my comments to her and my email address to she can point out where I am wrong: rwknap(at)comcast(dot)net. In the mean time, let me quote M.C. Lemon, University of Ulster in Philosophy of History to the question of what is history: "what historians actually do, today" Implied in the answer was there must be a historian in order to have history. If we accept Herodotus as the first historian then prior to him there was no history. There were certainly records upon which historians draw in order to create history. Those we can call the historical record. (i.e. the things historians can draw upon in order to create history.)

Now, about designing simulation games

Not so fast. Trying to get a fast one by me, are we? Lets talk a bit about what a game is.

The words "game" and "model" clearly mean different things; but that does not make them mutually exclusive categories, and nothing in the definitions you've given states or implies that they are. I see no reason that something can't be both a game and a simulation, which would be why there are hundreds of boxes on my shelves with the words "A conflict simulation game" written on them.

Let me ask you, of those ‘conflict simulation games', how many are for the computer? Those are not what we are discussing. I've already admitted that games played on computers are simulation games. So, I see no reason why a game cannot be a simulation. But that's not the discussion. What we are discussing is wargames (written set of rules, figures, model terrain). Are wargames simulations? In other words, wargame = simulation.

We can define a game as "as a set of players, a set of moves [as defined by the environment or conditions of the game] and a payoff. A wargame is that plus the environment is a military environment and it is played for entertainment. We can then define a simulation as a model of a real world process. These are created as decision-making tools. As you pointed out ‘game' and ‘model' mean different things. Therefore we cannot say wargame = simulation (damn, I couldn't find the key stroke that has an equal sign with the line through it.) Therefore, we can say a wargame is not a simulation although there maybe some common aspects to both.

Rich

Rich Knapton wrote:

[Snips] would be why there are hundreds of boxes on my shelves with the words "A conflict simulation game" written on them.

Let me ask you, of those ‘conflict simulation games', how many are for the computer?

Rats, I've misquoted the boxes -- now I wander across the hall to check, I find that what thet actually say is "An Historical Simulation Game". None of them are for the computer; they are all SPI board wargames. One or two say "A Future-History Simulation", and some that say "An Historical Simulation Game" are pretty clearly about hypothetical history, for example "Seelöwe".

Those are not what we are discussing. I've already admitted that games played on computers are simulation games. So, I see no reason why a game cannot be a simulation. But that's not the discussion. What we are discussing is wargames (written set of rules, figures, model terrain). Are wargames simulations? In other words, wargame = simulation.

I can see no hard distinction between the nature of games according to the board and counters they are played with. They seem to be an accidental, rather than an essential, feature. For example, I've played Risk, Chess and Reversi on the computer. I've played DBA with counters; and I believe there is a computer version available. I know lots of people have attempted games like "Squad Leader" using miniatures. And, back in the good old days of discrete-event simulation, PE Consulting used to sell a package called HOCUS. This signified the Hand Or Computer Universal Simulator. The recommended way to proceed was to produce a simulation model of the system of interest first by hand, using the formalism of Activity-Cycle Diagrams. These can be used to produce a sort of game that players can play using counters and cardboard shapes. Then, when people were happy that the manual model was a fair representation of the system, it could be translated by a competent programmer into the computer representation, and executed as many times as was needed to collect statistics.

[Snips]
We can then define a simulation as a model of a real world process. These are created as decision-making tools.

Among other reasons.

As you pointed out ‘game' and ‘model' mean different things. Therefore we cannot say wargame = simulation (damn, I couldn't find the key stroke that has an equal sign with the line through it.) Therefore, we can say a wargame is not a simulation although there maybe some common aspects to both.

Game != model, we agree (or game <> model, if you want to write it like that).

But all games are not wargames. Wargames are clearly a proper subset of all games. I would say that, as wargames are that subset of games that seek to mimic some aspect of war, so that mimicry makes them simulations.

If I could only do the ASCII art necessary to produce a Venn diagram, I would now draw one showing wargames as a subset of the intersection of the sets "games" and "simulations" -- since it seems clear that there are also simulation games that do not mimic war, such as business games, and other things SPI used to classify under the slightly odd heading of "non-wargames".

All the best,

John.

But John, what marketing puts on the cover of a game is not the defining element of what a simulation is.

I agree with you that a wargame is a subset of games. If we break them down it would be games, games for entertainment, games for entertainment with a military environment (conditions).

But all games are not wargames. Wargames are clearly a proper subset of all games. I would say that, as wargames are that subset of games that seek to mimic some aspect of war, so that mimicry makes them simulations.

This is the key point of contention between you and Bill and myself. I do not believe a wargame seeks to mimic some aspect of war as models seek to mimic some aspect of a real world process. I think they can be used that way as in military wargames. But, I do not believe that is what rule writers do for the entertainment market.

I think your use of a Venns diagram is appropriate for games involving the representation of a process, an event, or an activity, [We have been using process because that's the way the discussion was set up. However, if one wants to argue modeling events or activities is more appropriate than process then build a good argument for that.] These would be games for decision making (business or military) and computer assisted games (EA for example). However, computer assistance can range from all aspects of the game down to random number generation. In the first, the overlap is almost complete. In the second, the overlap hardly exists.

However, I don't think a Venns diagram is appropriate visual metaphor for wargames. The reason is that rule writers do not build models of processes, events, or actions. They transpose the data directly from the historical record to the game. There is no middle step of model building.

One might say that the whole rules system is a model but I don't think so. To me it seems to be a set of instructions (environment, conditions) for the interactions of the players. The difference I see between a wargame and a simulation is that in a simulation the model is changed in some respect in order to see what affect that has on the model's output. This change creates "what if" scenarios so decision makers might get a sense of what actions will result in what effects. Wargames are not designed this way. The rules provide an environment in which the players and not the model can create "what if" scenarios.

Rich