"Tank Kill probabilities?" Topic

This is the sort of thing the Pentagon seems to always be into- did anyone produce a list of Pk results per shot for WW2 tanks?

I've googled, but can't find anything remotely close to what I want- straight chance per shot of kill (basically combining hit chance with penetration) I HAVE LOADS OF DATA, so I can go through and work it all out, or type out a spreadsheet, but if there is already a list I don't really want to spend too much time- As its an infantry game, and tanks shouldn't be a major part, I don't want to spend more time on them than on the infantry!

I did find this cool site in my research
wwiiequipment.com/pencalc

I've read that a weapon's effective range is supposed to mean that one shot out of two will hit the target, e.g. 50% success, so you can extrapolate from that.

You can use that as a basis for varying ranges, weapons performance, and target armor.

Most games seem to separate the chance for a hit from the penetration determination, since with multiple weapons and varying armor thicknesses, it can get a bit complicated.

Dumping the info into a spreadsheet should help save some time, since it will do the calculations for you.

I hope that helps.

I spent way too long on that professionally, with more modern weapons, developing Ph/Pk data for simulations and simulators. This is migraine-inducing stuff.

"As its an infantry game, and tanks shouldn't be a major part, I don't want to spend more time on them than on the infantry!"

Then don't. K.I.S.S. Really. It's a game, not operations research.

Allen

Allen- I appreciate that (KISS) but I want to in roughly the right area:- "Exactly how did your 2lb kill my King Tiger?". This exactly the reason for my post- I am trying to kiss it!

If I can find just a few significant examples, I can extropolate the rest from that, plus I have a feeling there will be a limited number of tanks in the game ("You have an Maus in your carry case? That's nice- it's staying there") and types ("I don't care if the model does have the Porsche turret, the armour stays the same.")

Exactly how did you're 2lb

If you have gamers in your group who are only 2 pounds in weight, then that is quite exceptional…

Griefbringer

As I remember, the 1970's rule set "Angriff" had what you're talking about. Percentage of kills vs weapon, range, and angle of attack.

Also, I suppose you could use the Squad Leader stats. They would probably also work pretty well. Didn't they have some charts in the back of the rules with all the info from the counters in a convenient spreadsheet format?

From John Salts WW2 penetration doc compilation:

PDF link

Page 39 has a table showing %kill probs for various weapons against various targets at various ranges.

Taken from WO291/171, OR report on effectiveness of British anti-tank guns, 1943.

Two documents I have addresses tank engagements, to include numbers engaged and results, but does NOT specifically address number of rounds fired. Still some good information you might be able to use. The first is:
"A Survey of Tank Warfare in Europe from D-Day to 12 August 1944, An Army Operational Group (AORG) memorandum no. C6. This concentrates on Commonwealth engagements.

The second is:
Data on World War II Tank Engagements Involving the US Third and Fourth Armored Divisions by David C. Hardison, Ballistic Research Laboratories, Aberdeen Proving Ground, June 1954.

I got both, as reprints, at a reasonable price from Merriam Press.

I cut my WW2 wargaming teeth on Angriff….great game for the time. I'd kill myself before I played it today! It really was of the "data is everything" school where the penetration capability of every gun was accounted for at every range -- even beyond maximum effective range. And then the penetration capability was compared to the armor thickness of each vehicle depending on which part of the vehicle was penetrated. If I remember right, it took 6.378 hours to compute each shot!

raylev- sounds like Panzer by Yaquinto. Hex based Eastern front 1:1 tank/counter. I keep meaning to dig it out of the garage- from memory approx 15 hit locations, the location number depending on angle of attack – 0', 30, 45, 60, 90' (then the same for rear) and whether shot falling level or rising. armour thickness for each computed- all on data cards about 5"x3"- I got it from a salesman at a former club for £1- it was demo version, but unpunched. This pre home computers (well in a big way- early 90's). Think the type face must be 7 or 8 point!

Martin- Thanks for the link. I owe Mr Salt a beer!

Are you sure you have the right "Angriff"? I have I think all editions and tank fire was fairly simple. For example take a German 75/L43 or 48. Measure range and look on gun chart for that gun. You have 2 numbers, a percentage and amount of armor penetrated at that range. You then roll two dice, one of them colored. A simple table tells you the number you need to roll for a hit at that percentage. The color die tells you where you hit. If you get a hit you merely look at the armor where the colored die says you hit and compare to penetration. It either does or doesnt. We would extract the guns and armor on a card for just the vehicles/guns in our scenarios and greatly sped things up. Only takes a few seconds to do.

Marc….yep, the same game.

Your perception of the rules vs. mine reminds me of the joke about how often a husband and wife have sex. The wife says, "we have it all the time, at least once a week." The husband says, "we hardly ever have sex, only once a week."

:-)

Just reading your simplified description scared me again.

"I owe Mr Salt a beer!"

I'll get him one when I see him in a couple of months.

But then I will owe you one… Mind you never said I was actually going to buy him one.

:)

Last Hussar,

I'll let you off the beer as long as you promise to stop calling me "Mister" ;-)

When considering the hit-and-kill probabilities from WO 291/171, bear in mind especially (as well as all the other caveats listed) that they are for shots where the gunner has adjusted the MPI onto the target. They are therefore certain to be over-estimates for first shots where the range is only estimated (typical range-estimation errors are 20% of range).

I'm not sure I really followed your explanation of your intended resolution mechanism in this thread's cousin, and I don't want to induce any Angriff angst in you, but I would suggest that a suitable way to proceed would be to have one roll to find out if a shot hits -- an anti-tank gunner can reasonably be expected to see if he is hitting or not -- and then delay the pen/kill roll. This should make life easier for a lot of things.

First, roll to hit, using your favourite hit-determining mechanism. If it's a miss, no markers, no bother.

For each hit, compare the gun's penetration with the target's armour at the applicable range. If the projectile value equals the plate value, place one marker; then, for every additional 10% (say) of the plate value the projectile has, place another marker.

When the time comes to resolve the effect of hits, roll 1d6 for each accumulated marker. Any sixes and the target is killed; you might also rule that if there is a second six, the target burns (making the kill obvious to the enemy). Further refinements might be to limit the maximum number of markers that can be placed as a result of a single hit, say 6 for HEAT, 8 for APCR, APCNR or APDS, and 10 for other natures.

This scheme gives all markers the same value, so there's no need to remember what classes of proectile have hit the target.

All the best,

John.

To John Salt, Esquire.

I'm still on the fence re whether to have a hit roll (at time of firing), followed by a delayed resolution damage, or just combine into one action as delayed resolution. Either way it will be a 6 followed by a 4+ (or vice versa) – this gives an 8% chance to hit per dice.

at the moment it is all delayed resolution
Gun Value-Armour (G-A)= number of markers. Resolved on Tank activation.

The twist here is if G-A>A then the markers above 'A' are resolved immediately

So Firefly (G=17) vs PzIV A=6
Fire markers 11.
11 is more than A by 5.
So
5 Resolved at time of shooting (reroll 6s, 4+ kills)
if not destroyed then:
Place the remaining 6 markers on the PzIV- resolved at tank activation

However Churchill – Gun 11
Vs PzIV (A=6)
5 hit markers. 5 is not in excess of Armour so all resolved at tank activation.

What I don't know is if the tank doesn't go obviously dead how obvious is it you have hit? If you nearly always know you have hit, but not sure how badly I will go with a Hit now/penetration later system. If I want to show an immediate 'Brew Up' then it will be 4+ to hit, roll hits on tank activation 6+ kills, but all 6+ hits roll immediately
OR
5+/5+ system, gain with 6+ on the hit roll rerolloing immediately.

Hope what I am trying to do makes sense- values may be off at the moment, but on the 'test bed guns' it seems to match other sources.

PS
You have 'thanks' on the introduction page. The rules currently stand at 50+ pages when printed, which includes blank pages after the title page, contents, army lists, embryonic vehicle lists/rules, and a page of counters! Remember I started these to try and get a set of rules for my WH40k son! (there is still a bit of 40K in there, like DNA from a extinct ancestor species!)

I would think that your chances of knowing whether you hit the target would depend a lot on the range. Up close you'd probably be able to tell; farther away, maybe not. (Sorry to complicate your life.)

I rather like your system and may steal it, if I may, for my long-delayed company-stand rules. OK?

Don- another take on delayed resolution, inspired by the VLB rules (as discussed on the ACW boards)

link

I know its ACW, but the idea is you roll to see how many moves before you resolve- given the scale of your game bases may have to be in contact to fight.

given the scale of your game bases may have to be in contact to fight.

Infantry, yes. Artillery can fire further, of course, but if not observed it won't be very effective. Tank/AT a few inches, terrain admitting (which it often won't).

The Last Hussar wrote:

To John Salt, Esquire.

Aaaarrrgggh!

What I don't know is if the tank doesn't go obviously dead how obvious is it you have hit?

I would have thought that for tank shooting you pretty much have to be able to tell the hits from the misses to make it worth bothering. You have a high-mag sight, a tracer in the base of the shot, and a target that does not spontaneously leap into bushes or ditches the way infantry do, so it ought to be much easier than firing at them. If you can't sense the shot, you can't correct for range, and if you can't do that, you are going to have a pretty low P(hit) at any range. Of course, there is the problem of obscuration with big and blasty guns like the 17-pdr; if you fancied it, you could probably come up with some rules showing how high-muzzle-blast guns have a harder time sensing hits.

You have 'thanks' on the introduction page.

Oooh, ta.

All the best,

John.

LH: In addition to my company-stand micro-armour rules, I'm now working up a scenario for our online 20mm group, so I need a good method of resolving armor-penetration for that. Also accuracy. So keep thinking out loud. (For these games we use simultaneous movement, adjudicated by a GM.)

Report of our last game (my first with them) here:
TMP link

Chris, who ran that game, has not written his rules down yet.

Oh…Yeah!

ARRRRGGGGHHHHH

Been Bugged

Wrote huge post

Went to edit it and that appeared! I could edit it back, but I'm leaving it so you know.

Again.

I am using ACC and PEN. Both need a 5+, so the probability table looks like this
Dice….Prob
1 0.34
2 0.56
3 0.71
4 0.81
5 0.87
6 0.92
7 0.95
8 0.96
9 0.98

My research shows that many guns were 90%+ to hit up to 800m
I have 2 ranges at 1cm=10m*- up to 1m ant over 1m. 1m still puts most of an average table in short range of a tank gun.

For hits you throw the number of dice in the Acc column. ANY 5+ hits. Lose ACC for moving (you and them) and cover.

If hit scored then the delayed resolution Fire Markers are placed- Pen – Armour (see above for immediate resolution rules). On resolution any 5+ Kills. Hit no Kill is a button up.

*Amended as I typed this post from 1:500! Did some more maths!

My research shows that many guns were 90%+ to hit up to 800m.

Unfortunately, many accuracy studies are pretty much flawed. They either use the dispersion of the gun or do their test on a range that where the gunner is intimately familiar with the ranges. Studies show that the primary determination of inaccuracy is range estimation error,

You have a high-mag sight, a tracer in the base of the shot, and a target that does not spontaneously leap into bushes or ditches the way infantry do, so it ought to be much easier than firing at them. If you can't sense the shot, you can't correct for range, and if you can't do that, you are going to have a pretty low P(hit) at any range.

I read in a book on Panzerjagers that the Germans used captured Polish 75mm AP shells in their captured 75mm French guns to deal with the T-34 in 1941. Each gun burned through an average of 20 rounds per tank as they could not aim very well as the Polish 75mm didn't have a tracer.

Mobius- in the kidnapped post (?post napped) I did actually comment I realise that these are under range conditions- unfortuately in the retype it got forgotten- what with having to retype. I have minuses for movement and cover- with the d6 system as outlined the % fall away pretty quick.

(As I type this I am transfering some DV of the "Festival of History" to the external disk- a 25pdr is being deployed).

LH: What determines how many dice you throw?

I agree with Mobius that there's more to accuracy than the consistency with which the gun delivers its shot! Range estimation and other gunner skills (especially if leading a moving target); wind/weather effects; quality of optics; probably other things that don't spring to mind just this minute.

I'm mostly interested in how you determine penetrations/kills.

Don

Each tank 2 gun stats Accuracy and Penetration. ACC is rated from 1-7. any (currently) 6 you throw is a hit- just one no matter how many. This gives a maximum chance of 72% (I keep tinkering with the system). If the Range tests give 80%+ they get 7 dice (72%). Under 72% I round to the actual dice chance values, so 25% Range I give 2d, (31%), if I improve a gun it is to make sure guns that have noticably different stats are different in the game.

Cover and movement reduces the number of dice 72% – 67 – 60 – 52 – 42 – 31 – 17% (ie 17% is need a 6 on 1 die- the formula is 1-(.83^d) to give the probability out of 1- basically count possible chance of NOT getting a 6 and invert, x100 for %) Note the way the degradation of shot speeds up- I like this.

Not all tanks start at 7, especially at longer (1000m+) range- then losses for movement, cover etc. Weather can be scenario specified.

I then use the PEN-Armour to determine hits. After a lot of maths to compare conversion from thickness to kill chance, I've gone for a straight approx 10mm=1 Armour with leeway- 18mm is still a 1.

PEN is the same conversion+4. Thus a tank with quoted penetration 50mm is PEN of 9.

+4 gives 52% chance that a gun with penetration of (say) 50mm will kill a tank with armour of 50mm. (9-5=4 Fire markers). Obviously this reduces as the difference betwee Gun and Armour lessend. Thus penetration of -40mm over armour has 0. At this I allow a 50/50 (ie 4+ on a further 1d) chance to place 1 marker. At -1 no markers (though considering needing 5+ at -1 and 6 at -2 on further die roll)

I allow 'under penetrated' kills because the armour value is an average for the front or side, so it could hit a weaker spot or (say) track. also this is the Crew running, minor mechanical fault that needs a 15 minute fix but NOT in the middle of a battle (hose clip came off!) etc.

Pz IV H. Up to 100cm Acc 7, Pen 13, Over 100cm Acc 4 and Pen 13. Front armour 8, side armour 3.
(the stats I use show the the 7.5cm gun power tails off only after about 1,500m)

This fires at the front of a Firefly (Short 7/16, Long 7/16, Armour F7 S4) at a range of 120cm, which is behind bocage, poking through thus ACC-2. The German player throws 2 dice (Acc4 – 2 cover), giving a 31% chance of getting a 6 on either/both dice. If he hits he inflicts 6 Fire markers, 13-7. These are thrown when the Firefly next activates. Any '6' kills. (67% chance). If the Firefly had been in the open at less than 100 cm then the hit chance would be 72% (ie at least one 6 on 7 dice, Kill chance 67%=> 72%x67% gives a 48% chance to kill.

The Firefly fires back, having survived. Lets assume the PzIV is also in bocage.
Acc at 100cm+ =7. -2 (cover) =5. Chance of a 6= 60%

Pen 16- Armour 8 = 8 Hit markers 77% kill chance

Total kill chance 43%

Note if the Firefly had hit the side then the armour would be 3. Pen 16-3 =13.
The Brit playe can roll 10 hits immediately, because 13 is 10 more than the armour value. If these don't kill (Pk=84%), then the other 3 are thrown when the Panzer activates. This allows for obvious brew-ups for proportionately high penetration hits.

A tiger II is Gun Close 7/18, Long 7/17, Armour F15 S8!

It's starting to make sense to me -- which is scary!

Come, join the dark side.

I eventually dropped the 'all in one roll' when I couldn't find a way around the inaccurate powerful gun vs the accurate weak one- They may have the same Pk, but external factors would affect them differently.

I keep playing with the actual number needed. I think I will finish with "6's" on both. I've finished the German list (not guns and A/T weapons) and most of the British.

My biggest Tank v Tank rules decision now is the ACC mods. most tanks move 20cm per move (slow ones 15cm, Fast 30cm)- I have to decide how many cm per -1 to give for firer movement. To reduce book keeping I may well go for Target moved and is
Slow -1
Normal -2
Fast -3
so no need to mark how far it moved. The assumption is that even if they go the same distance, Fast tanks do it quicker

Fast tanks do it quicker

Something to put on a bumper sticker.

Lateral movement (relative to the shooter) would be harder to track than closing/opening movement, would it not? Assuming that they can't move far enough fast enough to really affect your range estimate all that much.

I've tried to do that in a space ship rule set, and its a bitch to write effectively and keep movement simple.

If the target is moving directly toward or away from the shooter, can it move fast enough to really make a difference in the range? If not, it doesn't matter that it's moving.

So only worry about lateral movement.

The assumption is that even if they go the same distance, Fast tanks do it quicker

I may have missed it but if your time scale is the same per every turn then moving the same distance means they are going the same speed.

Any movement of a target will cause problems even if its moving directly in front and away. Totally flat ground is rare. Lots of the streets would even not be totally flat. If a target is cutting across your LOS from left to right or right to left. The speed would be connected to the distance of the target. You may look a bit strange to people but stand near a road and track the cars like targets with your arm. Do the same from a much greater distance from the road and you'll see what I mean.

As much as I like detail trying to figure in various speeds would be hard. A fast target my not always be traveling fast.

You know another great way to work through some gunnery stuff like this would be a game like Panzer Commander on PC. Even though it may not be ultra real the concepts here would be illustrated on the game when you in the gunners postion. Any tank game would help IM not familiar with all the new ones though.

if your time scale is the same per every turn then moving the same distance means they are going the same speed.

Yes, strictly speaking, but I am thinking of approaching it as the amount of time actually spent travelling being shorter. It is easy to wargame and see time as discrete packets, where as in reality you don't move, let the opponent react, move, let the opponent react etc. Or perhaps the ability to zig zag is in this

I know its not perfect. This part of the game is still at the rough outline stage.

Supergrover has a good point. You might make it so that lateral movement makes a bigger difference at closer ranges than at distant ones.

As for streets/roads/etc. not being flat, a change in level is a different question, and would logically cause accuracy problems; but movement forward and backward, per se, might OR MIGHT NOT cause a change in level, but is not much of a problem in and of itself.

If your going to track movement in some respects but not all? Why do it all? Change in level to me would indicate a hill. That's not what I am talking about. The road out my door I can only track a target moving down it for a about a block then lose it. The slight changes in ground can cause an issue. Especially in older tanks. Wargamers Digest did some crude experiments with lenses they also would illustrate it. If its moving no matter how its going to be harder to hit.

BTW percentage chance to hit and checking penetration for thickness was the system for any wargame I ever played. With enduring years of convoluted schemes to simplify this Ive never seen anything quicker, easier and clearer, then a game with a couple of charts one for penetration and one for thickness. I hope more games move to this system.

Supergrover- that is essentially what I have- ?D6 to hit, any '6' hits, then a penetration score- PEN-Armour- roll that many dice, any 6 kills.

Given the cross country speed of WW2 tanks was so slow I may well go with just one modifier for movement, though possibly with a minimum to claim it. Most tanks move up to 20cm, slow ones at 10cm, fast at 30cm. I may say only a tanks that have moved over 10cm can claim, though this would require a marker. Mind you, last night's play test of the basic game we used tiddly-winks to mare the delayed fire scores, and my opponent made some of my units look like they were on a roulette table!

All this talk of Angriff had me dig out the rules. Tank-on-tank resolution really isn't that bad. It's a pain if you have to flip through the book. If you make a data card for the vehicles, it's not that bad at all.

You can speed things up by using percentage dice and rolling a d6 for hit location.

I'm thinking you could go two steps further and round the percentages to the nearest 5% and develop a 1 to 10 hit location system. Then, you use the 1s die to determine the hit location.

If you want to see hairy combat resolution, check out the Phoenix Command Mechanized rules…

Question: what kind of resources have the in-depth information on armour thickness that you see used in games? I mean, how do you know exactly how much armour was on the side of a turret versus the rear? Are there single reference books with this information (i.e. Jane's, etc.) or do you have to look at tank-specific information?

I've seen an ebook has data on pretty much all ww2 tanks. Loads of locations, thicknesses and angles. Dunno how accurate. I wasn't terribly interested because I don't think huge detail is the way to go in rules design.

As much as I loved the rules, Angriff's stats were silly. My edition, from 1972, only factored in slope on a couple of vehicles, most notably not the T-34 which made the front very easily vulnerable to most German guns.
--
Tim

How else do you think the Germans won in 1941? ;-)

Of course! and the bad metallurgy screwed the super cats over in late war!

I sense a doctoral dissertation…
--
Tim

agoodall wrote:

Question: what kind of resources have the in-depth information on armour thickness that you see used in games? I mean, how do you know exactly how much armour was on the side of a turret versus the rear? Are there single reference books with this information (i.e. Jane's, etc.) or do you have to look at tank-specific information?

For German armour, Chamberlain, Doyle and Jentz' "Encyclopedia of German Tanks of World War Two".

For Soviet armour, Zaloga and Grandsen's "Soviet Tanks and Combat Vehicles of World War Two".

For British armour, the relevant Bovington booklets.

For US Armour, Hoffschmidt and Tantum's "US Military Vehicles of World War II".

I don't know any comparably good sources on the Italians and French, but quite a few of their vehicles get mentioned in Chamberlain, Doyle and Jentz because captured examples were used in German service.

Hoffschmidt and Tantum's "Second World War Combat Weapons: Vol 2, Japanese" does not give such good detail on Japanese armour thicknesses, but since they are usually pathetically thin this does not much matter.

There is no published source I know that is consistently helpful about the hardness, yield strength or ultimate strength of the plate used, unfortunately.

All the best,

John.

You might make it so that lateral movement makes a bigger difference at closer ranges than at distant ones

Just going through the topic and noticed this stuff.

Don, you are quite right that lateral movement (across the sight picture) is harder to hit (and sooooooo incredibly frustrating) than movement head on, but the latter is still much harder to do and apply corrections to fire than for a stationary target. Much, much harder at closer range of course, but moving targets are still a bear at 1500 to 2000 meters.
--
Tim

Doyle & Jentz' "Panzertracts" series are probably about as definitive a set of works on German armour (& the various Beutepanzers, French, Russian etc) as you are likely to find & benefit from 30 years of additional research over their original Encyclopedia. They don't have a great focus upon armour thickness per se, but do cover it in the volumes that I possess. Hunnicutts various tracts on US tanks are also highly regarded.

Thanks, John and Huw! That's good stuff.

How do you guys feel about the accuracy of the old "tank Charts" rules. I enjoyed playing with them, but never could make the infantry work smoothly enough and they never became important in the games we played. But I came to be comfortable with the data. I later checked into some ballistics data and figured out how to convert armor thickness based on angle and it seemed like the author had done his homework right.
I myself am just getting back into wargames and am thinking about what rules to use. I like 1:1 tanks and larger units for infantry. Perhaps platoon. Would like to find a way to make infantry a very enjoyable part of the game.