"Naval Gunnery pre-dreadnought period" Topic

What's the most important factor for naval gunnery in this period – distance over the water, or the technical qualities of the gun?

Part of what I've heard seems to indicate that regardless of a gun's max range, the chances to hit are about the same b/c the sighting and other gear are primitive. Of course, the maximum range of a gun can be an issue, but it seems that all the guns in this period are used at well within max range.

Another way to phrase the question, if we're firing at 1000 yard increments, should larger guns hit better at longer ranges? Should small guns hit better at shorter ranges? And so on.

At the Battle of Santiago in 1898 in the Spanish American war, the range was 5,000 yards or under and the overall hit rate was under 3%. Most hits were scored by 5,6 and 8 inch guns more as a function of rate of fire than accuracy. All guns were essentially fired on local comtrol and the optics on a 12" weren't any better than on a middle size gun. Overall the guns could shoot a good deal farther than they could reliably aim. I don't belive the after battle analysis showed the American 12 or 13 inch guns to be any more accurate than the 5.6 or 8 inch.

By the time of the Russo-Japanese War the big guns were firing at 10,000 yards. I don't know how the acuracy improved but despite the Russians having more big guns, the real killer at Tsushima were the Japanese amored cruisers and the 'swamp them with shells' approach. I believe that lead to the prevalence of extreme mixed batteries (12" mains, 9.2" secondaries) just in time to be useless for WW1 when director control made the big guns truly deadly at long ranges.

Agree w/ McKinstry, have read the average engagement distance was approx. 8ooo yds during Russo-Japan War.

Please forgive my uninvited--and decidedly preachy--response, but I feel like I've just heard "safe sex" advice being given by a kid on a street corner. The apparent lack of real knowledge has provoked me, and I can't just let it pass.

There is simply so much shoveling of smoke, and piling it on top of more smoke in naval game gunnery, that it boggles the mind.

Well, my mind.

You've been warned, so skip on to something else if you don't want to feel lectured to by someone with an agenda. Two agendas, to be honest--one to enlighten, and one to point the way to something better, in time.

Still here? Okay, you were warned!

To answer the first question, the "technical qualities" of a gun are everything. If you know these, it is possible to "predict" the chances of hitting a given target at a given range. Range itself is irrelevant except as it produces a smaller "Danger Space" for the incoming shell, and lower chances of hitting.

Big guns are NOT "better" at long range, while all guns tend to be at "close range." All guns--of whatever size--are more accurate at "close range", and not because of better visibility, but the sheer physics of shell flight.

Also, choosing to figure the game chances to hit in thousand yard increments seems exceptionally severe, tending to make all guns look more alike than they actually are.

The effectiveness of any gun starts with its basic interior ballistics (partly a function of caliber), but as war gamers, we are concerned with exterior ballistics, and these are the sum of variables, most influenced by the gun's muzzle velocity.

Leaving aside the obvious relevance of the presence or not of Director Control, gun crew training, combat experience, and their physical/moral condition, you can graph the behavior of a given gun by the relation between shell weight, shape, muzzle velocity, the width and height of the target above the water, and angle of fire above the horizontal.

Other variables that war games must ignore include atmospheric pressure, humdity, temperature, and more esoterica which must be assumed to be a "wash" for prediction purposes.

The likelihood of one shot striking a target at any given range is a function of how high the shell flies above the water in feet, and the angle of its trajectory. A flatter angle shell is more likely to hit a target than one falling more steeply.

Commonly, chances to hit in naval war games are concocted from macro performance (such as looking at the percentages of hits scored for shots fired in a given action). This is rather like one race car engine being rated by the number of cars in the race and the order in which they finish--none of which describes the performance of a given engine at all.

The actual measure of a gun's potential is its "Danger Space" against a given target, a factor almost totally ignored by game designers, and understood--or so it seems!--by almost no one.

Danger Space is that area behind a target which if a shell would fall into actually hits the target first. The wider this area, the greater the chances to hit. The taller the target, the wider the Danger Space at all ranges. As ranges increase, and falling shells assume steeper angles near the end of their flight, the Danger Space narrows, and a point finally comes where a given shell would have to fall directly on the target to hit--effectively impossible in the Pre-Dreadnought era.

At longer ranges, all shells begin to fall more steeply from a greater height, limiting their chances of hitting to falling directly upon the ship itself, like an aerial bomb. The problem with this "plunging fire" is that it tends to occur at or over the horizon, a feature beyond gunnery control until after WW I, and then only with the use of aircraft to spot and direct fire. And due to the low maximum angles of ships' guns prior to the 1920's and 30's, plunging fire was effectively impossible in the Pre-Dreadnought and WW I eras for reasons of physics--not doctrine or "training."

I'm not saying that a much more accurate system can't be developed. Indeed, it already has been years ago by myself and colleagues (some now gone). "Cordite and Steel" is in re-design to eliminate the need for a team of Games Masters to adjudicate fire effects, and to fine tune rules for damage control, as well as to make the exhaustive ship/gun/shell data more accessible. The new game plays far faster than the original version, and can finally be moved from the floor to the table top, even without the loss of "true scale" for the Pre-Dreadnought era in 1:2400 or smaller.

I can hope that 2012 will be the year it finally comes all together again!

But, please, don't design your rules, or modify the ones you have and like, based on hearsay. Do some research, look up Exterior Ballistics on the Net, and seek grounding in fact, not speculation or "logic" based on nothing in particular.

Thus endeth the lesson.

TVAG

And then start to take into account the quality of the shells, the wear on the guns, the quality and consistency of the charge, the ability of crews to accurately determine range, the accuracy of the data the gunners have been provided with to lay their guns and you will, effectively, be back to square 1.

Sorry TVAG but it isn't just Physics, there are many more variables to apply in this period that are VERY difficult to make allowance for.

The lack of hard data on actual gun performance in action makes this a particularly difficult period for the prediction of 'game' performance. The theoretical performance is an excellent starting point, I'm with you on that, but the few action we have data for imply that actual performance was considerably worse than predicted and we really do not know the cause(s) for certain, at least not enough to quantify them.

What about the gun platform?

Surely ship design and stability also improved accuracy as time passed?

Sounds like the rant was more about some old rules set TVAG wants to resurrect in 2012.

The importance of fire control cannot be overestimated. All the penetration ability in the world wonīt amount to anything if you canīt hit the target.
What it comes down to is effective fire control solutions using the best equipment available by crews trained to the highest level.
Of course, quick firers were invaluable in that with all that lead in the air you had to have hit something, and the Japanese showed time and again how lethal these weapons were. But still, if either the Chinese in 1894 or the Russians in 1904-05 had been better trained, especially in fire control, things could have turned out in quite a different way.

Oh, and Happy New Year as well!

All well and good, but if you're doing a fleet game, you need fairly simple mechanics to represent "typical" shooting.

From what I read above, it _SOUNDS_ like there is a much more lethal "flat trajectory zone" which is basically close range – in other words, the gun is fired fairly level, the shell moves at enough velocity that there's a "point and hit" effect, and it falls into the sort of shooting that minimal training allows to be effective enough (for the same reason it's easy to train conscripts to use a WWI rifle).

I'm assuming that the larger the gun / higher the muzzle velocity, the longer the "flat trajectory" zone?

After this flat trajectory zone, you get into a "plunging fire" zone where you need significantly greater training, gun and shell quality, and fire control to hit b/c the space in which the shell lands is much smaller. Then of course the multitude of variables become much more important, everything from waves, ship stability, crew training, etc.

Does that sound right?

So in a game, we'd be looking at a series of "lethal flat trajectory zones" that increase with muzzle velocity (and perhaps the size of the gun) of each ships gun types. Behond that, the guns would be using "hail mary" shooting drasitcally affected by a wide variety of variables.

And I want to add that I'm talking about ironclads to pre-dreads, not WWI.

Where would one get gun data on the late ironclad / pre-dread era, especially the Spanish-American War and Russo-Japanese War?

Thanks for the info, all quite useful.

For weapon data, this site is invaluable:

navweaps.com

Also there are great gunnery charts in the Brassey naval annual series.

About plunging fire, donīt forget that ships of that era were not designed with plunging fire in mind, so if you can hit something it is easier to do some damage.

Dear HesseCassel, Et Al!

Yes, Sir, I'd say that in the simplest terms, you've got it!

Though, as I said, Plunging Fire (which is a great thing to deliver if and when you can!) was not a feature of the period prior to the 1920's/30's, and not for doctrinal reasons, but basic Physics.

I am aware you expressed interest in Pre-Dreadnoughts, but gunnery was an evolving science, and the differences between the eras was relevant to my points. Ironclad era gunnery can range from black powder and round shot, to rather more sophisticated shell shapes and early high explosives, but still much more primitive generally than the period of, say, 1895 to 1905.

And regarding where you can go to find all the data we as war gamers want, well… you can't find it except with "Cordite and Steel." There are sites, including the one thoughtfully provided by Texas Jack, and this one below that can provide a wealth of data, and some Danger Space info, but not for all navies/guns/ships as we have been running for years.

Link: link

To GildasFaciat and The Ditto Abominable Snowman: You might look at my rant and note the paragraphs that included the lines, "Leaving aside the obvious relevance of the presence or not of Director Control, gun crew training, combat experience, and their physical/moral condition, …," and, "Other variables that war games must ignore include atmospheric pressure, humdity, temperature, and more esoterica which must be assumed to be a "wash" for prediction purposes."

(Note to The Snowman--Huzzah to you, Sir, for looking things up, and not just storming ahead!)

I am very well aware of the many variables that go to effect the chances of a given gun (crew implied) actually striking a target, but the Physics are the most important factors, unless you assume the crew means to miss.

Back in the 70's when main frames were available only through Universities and Government agencies, our group of programers were able, over a couple of years, to run the gun data (max ranges, penetrations, and danger spaces) for practically every gun afloat from around 1890 to 1945. I still have one set of the 20 lbs + of print out from the last of the original runs of the program, but it is from runs made in the last couple of years or so that we have reconfigured that output so it is already in "game form." (Originally, the several hundred data points per gun had to be hand transcribed from print out to game tables.)

To Ogdenlulimus: You're damned right! As I said before, "Cordite and Steel" is coming back as soon as we can solve the practical matter of converting all the Ship Cards (one for every BB, CA, CL, and most DD's from 1890 to 1920) to new forms, among other concerns which have, fortunately, mostly been fixed already.

There is simply no other data set out there to work from as complete as what we "OCD number-crunchers" devoted ourselves to. No brag, just fact.

Albeit a fact I appreciate can't be proved to anyone until some evidence is provided to judge for yourselves.

Someone--all of you!--give me a specific Royal Navy main battery (1900-14) by size and caliber, the height of a target in five foot increments up to its highest main battery ("B" Turret), and any range out to 20,000 yards (the practical horizon), or the gun's max as best you have it. With those points, I can give a basic chance to hit simply by consulting the table for that gun/target/range.

Finally, since I'm evidently ranting anyway, I hope more folks will also begin to understand that just because a gun is a "Twelve Incher," it is not necessarily going to perform like any other "Twelve Incher." The difference in caliber alone (not just model and practical rate of fire) can make guns of the same size behave dramatically differently.

And Bigger is not always Better.

Stoke that in your boiler and steam on it!

TVAG

TVAG, while I appreciate your point about danger zones, I must also remind you that these zones would be included in any study of hit ratios in the few Pre-Dreadnought actions that occured.

Also, I believe the effect it would have in a war game is minimal. If ship X is three feet higher and four feet longer than ship y, the additional chances of scoring a hit canīt be much more than one percent or so (though it would increase with decreased range, indicating that distance is anything but irrelevant). Now this is just an off the cuff estimate, but if I am wrong could you please post a source that states otherwise? Of course, to some people even the increase of one percent is relevant, so maybe I, as a non-rivet counter, am missing the boat here, as it were.

I am also curious what data "we as wargamers want" is available only through your game. I have found all of what I think of as relevant information at either the website I provided (which, by the way, is also the link you provided), or in Brasseyīs, so I would really like to know what magic statistic I am missing out on! :)

I wish you the best of luck with your game, every new ruleset adds to our hobby.

Another factor is the quality, or even the existence of range finders, and the training of the men using them. Ranging error is usually (when you can find it) expressed in terms of percentage of range.

An untrained man using eyeball Mk.I has about a plus or minus 20% error. With traing that can be brought down to maybe 12-15%. That means that if the true range is 1000 yards, the untrained man's estimate will be anywhere from 800 to 1200 yards. If the trajectory is fairly flat, no big deal. If it isn't (say at 4000 yards, where the estimate is 3200-4800), the likelihood of the target actually being in the danger zone drops dramatically.

The biggest influence on the quality of ranging is the distance between the lenses – 1/4 foot in the case of the human eye. That's why range finders are classed as 'x foot'. Since the accuracy difference between two range finders is inverse to the square root of the ratio between them, this means that a trained man with an 8 foot range finder has an accuracy or about 2.4% of range (13.5% times the square root of 1/32). This translates to plus or minus 96 yards at 4000 yards and plus or minus 240 yards at 10,000.

rmaker: what rangefinders were used during the 1895-1905 period to the major belligerents?

TVAG: when you get your numbers in a gamer-friendly format, maybe WORD or EXCEL maybe you should sell the data itself, in addition to putting it into your game? I know I'd be interested in that kind of info.

Dear Texas Jack,

Not to challenge you, but if Danger Spaces were considered in "any study of hit ratios" in Pre-Dreadnought (or other) actions, would you please cite me one? Seriously, if you have one or more, I would very much appreciate copies for my own study.

You are, of course, correct that fractional differences in target height can be expected to only make fractional differences in results, but we found that five foot increments frequently double--and more--the chances to hit. Much of what occurs in the flight of a shell works in exponential fashion, not arithmetic, and the wide range of chances to hit are affected accordingly. In "Cordite And Steel," we recognize target heights from 5' to 50', masts and funnels not couonted. Indeed, the latter figure is just for completeness--until Aircraft Carriers come along, or the target is a Passenger Liner, that's an almost impossible height for a warship.

Of course I ultimately speak for myself when I say "What we wargamers want," but the computer work we did is the only source out there that calculates the chances of hitting (in percentiles) that requires no guessing, fudging, or just wishing real hard. If you have another, again I would be grateful to see it, if only to compare our results.

Mind, the rules supplement for the 1927 US Navy Wargames against Japan do provide some percentile chances of our 16" gun hitting a given target at a given range, and our figures are within 1 point of theirs. We found this data by the wildest chance through an Estate Sale (it's still marked "Classified"), so officially I don't have it. If you can share any similar information, it would only help.

And thanks for the "Good Luck" wish! Mind, "Cordite And Steel" came out in the 70's through TSR, as mentioned originally, and the re-release--which I hope will come this year!--will simply be the ultimate "2nd Edition" with all its changes in format.

And, HesseCassel, you may be onto something. We've debated whether to release everything on CD at once, or just a basic selection of gun and ship data to support the historical OB's that come with the game. Selling the data itself is a distinct possibility as everything in the gun collections by navy already in hand are already set up in an Excel (type, at least) Spread Sheet. This could be done more or less immediately if we thought there was a demand to justify the effort.

Anyway, if I wasn't buried alive--in a nice sort of way--here at TVAG, I'd try to get this "iifetime achievment" back out to where it could do some good!

TVAG

TVAG, thanks for the clarifications. I understand now it is the data you compiled rather than a mere statistic that you were referring to.

What I meant by danger zones being included in any study of hit ratios is simply that a hit is a hit, and that is what counts. It is unimportant as to what could have or would have or should have been a hit if only the target were ten feet higher, longer or a lighter shade of gray.

However, the idea that different ships have different chances of being hit are quite interesting, though perhaps a bit on the tedious side for my taste. Still, intriguing all the same! In my rules I give monitors and semi-submersibles a lower chance of being hit, but there is no penalty for an exceptionally tall ship, for example. I may be a customer for your data in the future. :)

However, wouldnīt be a bit time consuming if you had to look at a different hit ratio based upon the individual ship as well as what type of gun was firing on her? Is your game able to solve this quickly without breaking the flow of play and bogging folks down in a sea of statistics? You donīt have to tell me if itīs a secret :)

TVAG

Your methodology seems to make assumptions about the variables involved in your calculations – i.e. that they are precise and devoid of any error distribution. I may be wrong in this, so please correct me if you have included the potential for error in your calculations.

As an example let us assume that the elevation may have an error of up to +/-10 minutes of arc, distributed normally. Your danger zone remains as the error free calculations describe it BUT with a probability that its centre may vary.

Do we know what that error distribution was ? Can we be sure that it is negligible ? If not then your calculations ignoring the error will produce a false value for hit probability.

Now consider all the other variables used in your calculations – for how many can you say the error function has a negligible effect ? Is such data actually available ? I very much doubt that it is for the guns used on pre-dreadnought warships.

Once technology passes beyond that available when the pre-dreads were built, and the science of gunnery improves along with the optics for rangefinding and the mechanics of gun mount design THEN some of those error functions MAY become less significant, even negligible, but I do not think they could be ignored on ships built before 1906.

Don't get me wrong here, I'm not suggesting that this is just another layer of calculation to add to your model, I'm trying to show that a pure calculation model cannot be truly effective in predicting hit probability for pre-dread armament because there are too many unknown quantities.

To Texas Jack & Gildas Faciat,

Texas Jack!: The data is tabularized per gun such that one simply reads down for the target height, and over for the range to find the base chance to hit with one shot, and the chances for any additional hits up to the maximum number fired by that gun that turn at that range and target.

Thus, only one consultation of the table is necessary for any one set of shots.

Gildas!--I do indeed see and respect your point, but I think you are in honest error on assuming that a "Pre-Dreadnought" gun is somehow less predictable than a later piece. It's effectiveness will likely be less, but that's not the fault of the gun.

It's the improvements in fire control that are crucial, as well as changes in shell shape and other technicalities over time.

However, the weight of shell, its shape, muzzle velocity, height above the water, and angle of fire (to reach a certain range) for Pre-Dreadnought guns are in themselves absolute. They are what they are, and do not change, in the presence of Dreadnought era guns. Indeed, the 12"/45 on "Dreadnought" herself is the same gun as on the previous Pre-Dreadnought class of British BB's, and neither would have Director Control upon launch.

The change between "Pre-" and "Dreadnought" is not in the guns themselves (though larger sizes come later on), but the change to all "big gun" batteries, and later--most tellingly--Director Control. There was no sudden Transubstantiation when "Dreadnought" was built. The Spirit didn't leave the "Pre's" and move to the "Dread's." Only the potential effectiveness of fire changed.

In other words, if a game says "Pre's" work under a different set of rules just because a Dreadnought is in the game, taffy is being distributed.

A Dreadnought will almost certainly defeat a "Pre" in any otherwise even-up fight, but not because of some magic rule. It will likely win by being able to put more "big guns" on target, and, on the advent of Director Control, with greater accuracy.

Finally, while a difference of a foot here and an inch there presumably causes some change in any calculation, the question really becomes, is that change of statistical significance? By using Gun Platform and Target Values in 5 foot increments, for example, there is plenty of room for smaller variables to average out. Or, to put it another way, "A difference that makes no difference, is no difference."

In the end, the hardest thing to quantify to everyone's satisfaction (HAH!) are those "non-physical" factors that influence the inherent accuracy of the guns themselves. As I referenced before, crew quality/training/experience/and physical/moral condition all can influence outcomes. And certainly the presence and differences between Director Control systems would be crucial.

Indeed, most modifiers would be "negative" since the absolutes of gun performance are mostly degraded by human contributions. Then again, a good system of fire control can bring the guns back up towards their pristine states of performance.

We expect to provide our take on these modifiers as multipliers to the previously determined chances to hit, so that (if used) Hit Calculation would be a two step process--finding the basic odds to hit (and the potential number of hits), and then the modifiers to get final odds to hit.

For example, our Old Friend, the Graf von Spee's East Asia Squadron held the Hochseeflotte's Shooting prize for accuracy. This almost certainly made an enormous difference in the fate of Craddock's fleet off Coronel, the crews of which were largely reservists, with old rangefinders and lack of experience.

On the other hand, at the Falklands, Spee's turn in the barrel was pretty much inevitable for gun range reasons alone, but the Director Control that could be thrown at him must only have accellerated the outcome.

Both cases strongly suggest that these crew and technical factors should not be ignored anymore than the actual, base, and predictable performance of the guns themselves. Danger Space, penetration, and other gun data would be known to a warship's Captain and Gunnery Officers. They would be able to ID their enemy and know whether to stand out or move in to fight with an advantage, or at least reduce that of the enemy.

A game that just gives arbitrary chances to hit, and differentiates vessels only by their number of "points" takes all that away from the players, and makes only a bone-rolling game, rather than letting a player's knowledge and expertise of the guns and vessels give him an edge. Or the lack thereof at a terrible disadvantage!

TVAG

Of course, all those "intangibles" and human factors were taken care of in the original "Cordite & Steel" by having the players estimate the range to their target. That is probably my favorite thing about the rules, it makes "you" those factors usually handled by modifiers and dice. Unfortunately, it's hard to find the room to play the game that way, so we have to make do with modifiers and dice.

Gildas!--I do indeed see and respect your point, but I think you are in honest error on assuming that a "Pre-Dreadnought" gun is somehow less predictable than a later piece. It's effectiveness will likely be less, but that's not the fault of the gun.

TVAG : Many pre-dreads had guns that were a great deal less sophisticated in both design and construction than your example. Many ships still carried guns built in the late 1880's, not all countries could match the speed of development of the major naval powers and had to manage with what was available.

By constantly referring to WW1 examples you clearly believe that the changes in gun technology from 1890 to 1913 had only minimal effect on the potential (if not actual) accuracy of the guns. I do not believe that to be even an approximation to the reality.

The predictability of ANY system (naval guns included) is a function of the predictability of ALL aspects of that system. I still contend that the uncertainty of the factors involved in your calculations (muzzle velocity, elevation, angle of training etc) is enough to make prediction of flight based upon fixed values of limited use in calculating hit probabilities.

The variation in range (some 100's of yards at ranges up to 5000y) seen in naval guns test fired in the early 1890's (at Sheerness I think) from fixed land mountings shows that there is at least some evidence for my point.

I just want to say thanks, for what is a very interesting discussion about gunnery, with details and considerations I knew little or nothing about.

TVAG I can't speak for the market place. But it seems to me that a download from the many available vendors should lower the price to get it to market, and then the results will have to speak for themselves. Obvious gaming interest, but possibility of historians and others, plus a few Navy techies who might be interested.

Just a thought.

I'm primarily interested in gunnery for gaming, so can't speak to much else. But this is one of the best threads I've ever had at TMP.

To HesseCassel: Well, I think all of us on this thread are flattered by your kind words, and if others have become aware of things they didn't know, so much the better.

And regarding the encouragement from you and others to do so, my colleague, Chris Ferree, and I have been goaded to leap ahead with the the project of bringing "Cordite And Steel" back, though in a way I hadn't considered until I put my uninvited oar into your thread.

I'll be posting a rather oddball request for help from the naval wargaming community shortly which, if responded to well enough, could shave at least a year off the return time for the rules proper. But in the meantime, with just a small amount of work, TVAG will begin offering all this gun data--the volume of it will knock your socks off--to the public at large for use with game design or general information. This should begin in a matter of DAYS as the first release will be all the guns of the Royal Navy from c1890 through WW II. More details will follow as I post my call for participants to finish the last big job.

To Gildas: Well, I've just about run out of ways to tell you that there is nothing "unknown" or "unknowable" about early modern naval artillery. What part of having shell weights, shapes, muzzle velocities, target and gun platform heights, and the range to a target don't you seem to get?

I don't understand why you seem to think these either don't exist, or are somehow "unreliable," particularly between guns of the 1880's and later of different navies.

An initial velocity of 1250 feet per second in a gun from 1880, is the same initial velocity of 1250 fps in 1950. Is a Russian initial velocity of 1250 fps inferior to the 1250 fps of a French gun?

There's plenty of room for discussion on so many things, including the best ways to apply the data, but have you never seen any naval works from the the Pre-Dreadnought era? Don't you have access to any of the "Janes Fighting Ships" reprints. Brassey's? Libraries? Have you never seen lists of muzzle velocities, shell weights, etc, in books from the 1890's--and even before? Do you think engineers were so ignorant they couldn't measure these things accurately? Did Aliens build the Pyramids because we humans are just too stupid?

Yes, I do refer repeatedly to the Dreadnought era, but because you seem to argue that there is some amorphous difference between it and the Pre-Dreadnought period. I've already explained the real differences, and they do not occur at the muzzles.

How else would you have me explain it?

"Some people come to the well of knowledge to drink, others just to gargle."--Carl Sagan

TVAG

TVAG

Quite clearly I'm not getting over my point. I never said they were 'unknown' or 'unknowable', you are putting words in my mouth that I did not use. You do not seem to comprehend the concept of error in measurement when dealing with real (rather than idealised or theoretical) systems.

Take muzzle velocity as an example.

You are assuming that every shot fired has EXACTLY the same MV and there is NEVER any variation (or atleast that any variation is negligible). This is simply not the case. Values given in gun specifications cannot be taken to be that precise, there is ALWAYS some variation between shots. It is this 'uncertainty' that I am referring to and the spread of possible MV for a particular gun/charge/shell is the 'error distribution' for that combination.

Gun & shell construction improved considerably between 1880 and 1906 and the error distribution of MV would be greater for earlier guns because they were not as well made. This means that the range of MV from shot-to-shot would be greater. It is this higher variance of MV which makes the earlier guns less predictable in their performance.

I find you tone rather condescending and it is clear to me that you haven't researched some of the technology as thoroughly as you think you have. I suggest a look at some of the volumes of the various professional journals and see the reports of those engineers attempting to measure MV accurately. You will see that they found the task a good deal more difficult than you seem to imagine.

If you were actually an engineer (as I was earlier in my career) then you would realise that you need to be neither 'stupid' nor 'ignorant' to be unable to solve a problem or measure something with high accuracy.

As to sources, I own most of the volumes you list, and a good deal more besides, but I am not foolish enough to simply take the values given as any more than the guidelines for comparison that they published for. The figures are mostly taken from published sources from manufacturers or, where that was not available, by comparison with guns of similar manufacture. They are not exclusively taken from rigorous naval testing and so must be considered as a good 'norm' rather than an absolute, unimpeachable value.

1. With a little digging it is posssible to derive the historical hit percentage for a number of WW I and WW II naval actions. Try Campbell's (sic) book on Jutland. Very good data is also available for the Solomon's actions. Cape Matapan and related actions in the Med are also documented.

2. Don't get confused between actual hits and salvo straddles. Actual hit percentages were sickenly low with non-radar technology.

3. If I remember correctly: Queen Elizabeth class, 1940, stationary target, 10,000 yards range, roughly 100% straddles, 33% actual hits. Jutland, main guns, all ships, 16,000 +/- yards, 3% actual hits.

4. I've fired 1940's style guns, telescopic sights, at dumb targets. It isn't as easy as you'd think.

Prior to smart munitions naval gunfire was very inaccurate. Sorry if this is depressing.

And that is 1940s technology, pre-dread gunnery was not nearly as sophisticated.

Dear Thomas,

Not "depressing" at all! The fact is, barring point blank fire (or very judciously applied radars), naval gunfire was highly inaccurate on a shot-by-shot basis in the periods of WW II and before that gamers are interested in. Chances to hit are most commonly in single digit percentiles, and primarily in the low end of that--when not virtually "zero."

Of course, to reference your example, a 3% hit rate is the average of all shots fired, and that's not how a miniatures game designer can rate the many different guns engaged at various ranges. Most actual shots/salvoes will have less than 3% chances to hit for all guns at the same and longer ranges, while shots from closer ranges can begin to go up dramatically. They can all be expressed with an average chance to hit, but this number doesn't apply from moment to moment and turn-to-turn.

Of course you never suggested such a thing, but I make the point to warn the game designers who would.

If I were designing a board game of a more strategic nature, an average of hits in a battle might be the basis for a combat results table, but not for a ship-to-ship miniatures game.

To Gildas: I apologize and am sorry you feel I was condescending. I didn't get into this thread to offend anyone. I confess to being frustrated that you seem to be arguing in a direction that makes me think you've not really heard what I'm trying to share.

But you do speak almost as if science is not enough. Your point about muzzle velocities varying from shot-to-shot is valid. Clearly, there are enough variables of internal ballistics to produce a range of "final" initial velocities.
But would any variance produce a substantially higher initial velocity than listed? Do you think the variance could be as much as 50% either way? 25%? 10%? 5%?

Clearly it's not the larger percentage, and is much closer to (or even below) the lowest figure.

So, again, a difference that makes no difference….

Indeed, can you name me ANY "absolute, unimpeachable" value in any engineering application? When they figure crash speed tests for safety issues, do they always achieve EXACTLY the same speed/impact per test? No. Can they still derive rules and predictable results from them? Yes!

It's not about "absolutes"--even though I used that word myself. It's just that if you can find different muzzle velocities for the exact same gun, shell, etc, you will not find "significant" differences (say, 5% or more). Where we have found a few that seemed at a wider variance, we averaged them and moved on.

If you wish to think that any muzzle velocity figure is itself an average, so much the better. It means we can speak of calculating base accuracies and recognize that the results we get have already had a range of variables included--and that the final result is still usable and useful.

In short, I like your use of the word, "norm," because if that's what we have, then "Game on!"

When we use published data, we take it (generally) on face value, for what is the alternative? In that sense, the figures are "absolute"--we have no others. Can we take a listed velocity and say we "feel" it's not right, so we'll just add or subtract some? Are we to say that the variations out past the fifth decimal point preclude any useful results? Are we just to give up and not try because we don't know EVERYTHING from the start?

We are optimists. We believe we can work with what we have to produce data of value. Is it perfect? We can't know, but I don't believe it. But are the calculations that design an aircraft also "perfect," and do I still fly (sometimes reluctantly!) in one anyway?

The proof--as ever--is in the pudding. If you should acquire any of the gun data we spent years (it's been over 35) calculating by computer, and find any portion of it at odds with your own data, we want to know. Your data may be better than ours, but we'd want to know how you arrived at your figures before challenging our own. Is that too much to ask?

And for any who somehow still doubt it, "Damn Straight, I am proud!" of what has been accomplished by the likes of the late Eric Just who first laboriously studied and adapted Bliss' "Tables of Exterior Ballistics" for this project, and proved it could be applied to the computer age for gaming and historical insight. And the work done by all the other contributors who have continued his work, and the countless hours over years to find the odd detail or missing data so that the rules could be written, transcribed, and rewritten based on the gun data, etc.

The release of this information will begin ASAP to be used as the reader thinks best.

This thread has, inevitably, fallen off the front page, and perhaps that's just as well. I now have plenty to do, and we all have better use of our time. I will drop back to see if anyone has something to add, but, like the thread author, HesseCassel, I'm grateful for the chance to rant, and for the energizing that has come out of it.

TVAG