Battle Test Report

victorhauser · Post by **victorhauser** » Fri Jun 30, 2000 10:49 pm

I just completed five battle tests where I had the computer play against itself. The conditions were the same for each battle.

Initial Conditions for Each Battle
June 44
Meeting Engagement
Player1 = German
Player2 = Soviet
2440 Battle Points for the Germans (100 points for the HQ and 2340 points for 20 Tiger Is in two companies of 10 Tigers each).
2500 Battle Points for the Soviets (100 points for the HQ and 2400 points for 20 T-34/85s in two companies of 10 T-34s each).
AI set to 0.
All Realism Settings turned ON except Country Training turned OFF.
Troop Quality set to 70 for both sides.
All other settings on default/100%
Computer Control for both sides.
Human Buy for both sides.
Computer Deploy for both sides.
Noon, No Major Water

Battle #1 Results
Soviet surrender after 11 turns
Visibility 18 hazy
6 Tigers destroyed
1 Tiger abandoned
2 Tigers immobilized
18 T-34/85s destroyed
2 T-34/85s abandoned

Battle #2 Results
Soviet surrender after 11 turns
Visibility 14 rain
5 Tigers destroyed
19 T-34/85s destroyed
1 T-34/85s abandoned

Battle #3 Results
Soviet surrender after 12 turns
Visibility 21 clear
5 Tigers destroyed
18 T-34/85s destroyed
2 T-34/85s abandoned

Battle #4 Results
Soviet surrender after 10 turns
Visibility 22 clear
1 Tiger destroyed
1 Tigers immobilized
18 T-34/85s destroyed
2 T-34/85s abandoned

Battle #5 Results
Soviet surrender after 21 turns
Visibility 22 clear
8 Tigers destroyed
1 Tiger abandoned
1 Tiger immobilized
17 T-34/85s destroyed
3 T-34/85s abandoned

Average Results of the Five Battles
Soviet surrender after 13.0 turns
Visibility 19.4 per game
5.0 Tigers destroyed per game
1.0 Tiger abandoned per game
0.8 Tigers immobilized per game
18.0 T-34/85s destroyed per game
2.0 T-34/85s abandoned per game
0.0 T-34/85s immobilized per game

Preliminary Tentative Conclusion
Tiger Is are approximately 2.94 times as effective as T-34/85s in "tank-to-tank" combat {20.0 (sum of Soviet average losses[destroyed plus abandoned] per game) divided by 6.8 (sum of German average losses [destroyed plus abandoned plus immobilized] per game) = 2.94}.

Seth · Post by **Seth** » Fri Jun 30, 2000 11:08 pm

Fascinating. Care to try out some other units that are roughly point-equivalent, but one seems to have a definite edge? The Tigers did even better than I expected. Maybe Panthers?

Drake666 · Post by **Drake666** » Fri Jun 30, 2000 11:10 pm

That dont prove anything. The computer dont have a clue how to use the equipment. You give two T-34/85s to a human player and but him against a humen player with one Tiger and the Kiger is going to be died in no time.

victorhauser · Post by **victorhauser** » Fri Jun 30, 2000 11:18 pm

Well then, Drake, why don't I give you 20 T-34/85s and I'll take 20 Tiger Is and we can test human vs. human, okay?

Larry Holt · Post by **Larry Holt** » Fri Jun 30, 2000 11:27 pm

Originally posted by Drake666:
That dont prove anything. The computer dont have a clue how to use the equipment. ...

While its true the AI doesn't know how to maneuver, etc. since it plays both sides equally badly, this issue cancels out. With these two tanks, there is no special tactical consideration that the AI could fail to employ. E.G. for max arty effectiveness, deploy it rearward and use its indirect fire capability instead of as a direct fire weapon, etc.

What it proves is that over time, all other things being equal, Tigers are nearly three times as effective as T-34's against each other.

What it does not prove is any other speculation as to equipment effectiveness. For example, this doesn't take combined arms into consideration. E.G. While the WWII US Army had tanks that were less capable than the Germans, doctrinally the US Army intended not to use them as antitank vehicles, this role was supposed to be carried out by tank destroyers. (I know, I know that this was doctrine and it failed in the field lots of times the US had to rely upon tanks in this role - I just can't think of a better example off the top of my head) So, a battle between US tanks and German tanks would show that the German tanks were x times more effective but would not be doctrinally accurate as it did not test under doctrinal conditions. (however it might be historically accurate given the US' departure from doctrine). None-the-less my argument shows that a test is only valid for the conditions tested. To become generally valid, you must include all factors in varing degrees in a stastically valid proportion to that which is found in real life. As the number of varibles increases, the number of test cases goes up geometrically. Given Victor's one varible, the type of tank, five test runs seems valid.

------------------
An old soldier but not yet a faded one.
OK, maybe just a bit faded.

Larry Holt · Post by **Larry Holt** » Fri Jun 30, 2000 11:30 pm

Originally posted by victorhauser:
...
Preliminary Tentative Conclusion
Tiger Is are approximately 2.94 times as effective as T-34/85s in "tank-to-tank" combat ...

Then I would expect that over several tests, 60 T-34's and 20 Tigers would fight to nearly equal results (draws or roughly equal numbers of wins). If I can install the patches I'll try it this weekend.

------------------
An old soldier but not yet a faded one.
OK, maybe just a bit faded.

Drake666 · Post by **Drake666** » Fri Jun 30, 2000 11:31 pm

I did not say they were even. I think you should get three T-34/85s for every 2 tigers would be around right.

As for the game send it over. Random map meeting engagement, with each side with their 20 tanks and no other equipment, sounds like a test.

Seth · Post by **Seth** » Fri Jun 30, 2000 11:58 pm

Actually, point for point, the Tiger is 3.016591251885 times more effective

Seth · Post by **Seth** » Sat Jul 01, 2000 12:07 am

Serious question now. Doesn't averaging out the three conditions (destroyed, immobilized, abandoned) before adding them up queer the formula? If I add them all before averaging, then I get 6.2 KO'ed Tigers per game, yielding a base 3.225806451613:1 effectiveness, and 3.308519437552:1 adjusted for point-to-point.

Paul Vebber · Post by **Paul Vebber** » Sat Jul 01, 2000 12:43 am

I think that such demonstrations can give an insight into potential problem areas, which this seems to indicate teh T-34/85 (or Tiger relatively speaking) but don't think that it calls for making the T-34/85 3 times cheaper.

The AI vs AI tends to reward the side better ina "train wreck" where two forces smash headlong, as someone pointed out, it tends to minimize factors of speed, non-frontal armor, or lomg range hitting advantages.

I will make sure the OOB group takes a good hard look at point costs as I do agree that they need significant tweaking, but I think it will take some time to come up with god numbers.

Thanks Victor! If nothing else its an interesting test!

Seth · Post by **Seth** » Sat Jul 01, 2000 12:50 am

There's definitely a tweak required here, but the Tiger would end up costing nearly 400 points using these figures. Person v. Person, or Person v. AI tests would have different results, but people are far too variable to really give an idea of what's going on. I am inclined to say that the cost of the Tiger should go up a bit, and the T-34 should drop a bit. How much, I have no idea. Maybe I'll try some tests this looooooong weekend.

[This message has been edited by Seth (edited 06-30-2000).]

victorhauser · Post by **victorhauser** » Sat Jul 01, 2000 12:56 am

Originally posted by Drake666:
I did not say they were even. I think you should get three T-34/85s for every 2 tigers would be around right.

By Jove, I think he's got it!

My point in all this is to create some dialog on how much a unit should cost. Drake is the first person I've read in this forum (I admit I haven't read all the posts) to state that 3 T-34/85s should cost as much as 2 Tiger Is. The current OOB has a price of 117 for the Tiger I and 120 points for the T-34/85. But Drake believes that the point cost for the T-34/85 should be around 79, and not 120.

What do the rest of you think?
Is the Tiger I priced correctly? If yes, then is a price of 79 a "reasonable" price for a T-34/85?

I'd like to see some numbers from you guys. So far, all I've seen are generalizations. Let's see some estimated prices from you guys and then we can plug those numbers into our OOB editors and run some more tests.

And while Larry correctly states that all units need to be priced according to some "combined arms" context, we have to start somewhere. And since SPWAW is primarily a tank-vs-tank game, then I see no problem with starting there.

Why not make a list of what you consider to be the most effective down to the least effective tanks and then compare your list to the SPWAW price list (no peeking at the price list before you make your own list

).

This is important work and our feedback is of value to the SPWAW staff. The better the price list, then the more fair our tournaments and head-to-head games will be.

P.S. Due to the application of Lanchester's Laws, if something is roughly 3 times as effective as another thing in battle, then the actual numbers employed should be roughly the square root of that ratio. Thus, if a Tiger I is deemed to be 3 times as effective as a T-34/85, then the number of T-34/85s employed should be roughly the square root of 3 (which is approximately 1.7)times the number of Tiger Is. So an "even" fight should result if 34 T-34/85s (1.7 * 20) fight 20 Tiger Is using the conditions I described at the start of this topic.

WW2'er · Post by **WW2'er** » Sat Jul 01, 2000 2:16 am

Hey! I just want to say thanks Victor for bringing this up. I hope many will reply. It's obvious that Matrix WILL listen to the suggestions. I'm not as much of a numbers guru as others who have replied. I'm just the average gamer having a blast with this game. Yet I realize that these prices do have a large effect on campaigns and PBEM games.

So from the "average" SPWAW player, here's hoping a lot of you voice your opinions and the game is made even better because of it!

WW2'er

Originally posted by victorhauser:

This is important work and our feedback is of value to the SPWAW staff. The better the price list, then the more fair our tournaments and head-to-head games will be.

Drake666 · Post by **Drake666** » Sat Jul 01, 2000 2:36 am

Victorhauser I dont have time to rate every unit but I will rate the major powers on what country has the best chance of winning right now with the unit prices the way they are.

1939 to 44
Soviets
UK
Germany
USA

1944 and 45
USA
Soviets
UK
Germany

As you can see the Germans rate low for all of the war. I think the German mobile AT are the worse in the late war. With the way their prices are now I never buy German ATs late in the war.

Charles22 · Post by **Charles22** » Sat Jul 01, 2000 6:00 am

Victor: You, you, you, Russophile you. (Now using my Bomber Harris voice) "You have sowed the wind, and now, you shall reap the whirlwind."

Hmmm, let's see here. I did a little test of my own, Using your own system the following results were achieved:

Initial Conditions for The One Battle (One is enough to make my point here)
July 41
Meeting Engagement
Player1 = German
Player2 = Soviet
Premise of battle? KV-1E is too cheap. This battle features, for that time and some time afterwards, the best of the USSR vs. the best of Germany, the PZIIIH vs. the KV-1E. Again, 20 vs. 20. The PZIIIH costs 54, while the KV-1e costs 82.
AI set to 0.
All Realism Settings turned ON except Country Training turned OFF.
Troop Quality set to 70 for both sides.
All other settings on default/100%
Computer Control for both sides.
Human Buy for both sides.
Computer Deploy for both sides.
Noon, No Major Water

Battle #1 Results
German surrender after 15? turns
Visibility (didn't catch it, but was at least 10)
13 PZIIIH's destroyed
6 PZIIIH's abandoned
3 KV-1E's with one damage point apiece

So what should the KV-1E cost here, 1,000 points? The point isn't effectiveness, that's just plain cookie-cutter territory, and it slices both ways. So let's see, if the KV-1E costs 1,000, and Tiger can double it in a battle, does that make the Tiger a 2,000 point unit? And if the T34/85 can outperform the KV-1E, against the PZIIIH, does that make it a 1,000+ unit also? And if the 3-to-1 ratio on the Tiger holds, does that make it some 3,000+ points? This is ridiculous.

[This message has been edited by Charles22 (edited 06-30-2000).]

Wild Bill · Post by **Wild Bill** » Sat Jul 01, 2000 7:17 am

Question: Maybe I missed something here? Were these scenarios done in mixed terrain? Or in open ground? Were the experience and morale values the same for both sides. Were they all equidistant? Etc, etc.

A Tiger tank with 80 experience and a 75 morale factor will most likely cream 4 T-34/85 with a 50-60 experience and a 50-55 morale factor.

Victor, I am not knocking or making light of your tests. Not at all.

The point I am trying to make is that not test is really that accurate unless you make all things totally equal.

And could that ever be in this game? No, I don't think so. There are too many variables (different probabilities) that affect the game.

Sometimes we forget this. But they do play a role. A big one. Sometimes they make all the difference.

And think about it. Would you want it any other way? Would you want a cut and dried slide rule game where results are almost guaranteed? Of course not!

The whole purpose in battles in wargaming is NOT having any guarantees, of having some fog of war, some uncertainty. I call them tension-builders.

Haven't you ever faced an impossible situation in SPWAW and come through? The skill of your unit plus a roll of the die changed what seemed to be a deadly fact.

I think tests should be run. But please remember that a lot of factors should be considered when so doing.

Me? Give me excitement! Give me uncertainty! Let me overcome! Let me have a unit that I can pin a medal on when the smoke clears.
Give me some action that causes me to make more noise than the game does

That is my kind of wargame!

Wild Bill

------------------
In Arduis Fidelis
Wild Bill Wilder
Coordinator, Scenario Design
Matrix Games

victorhauser · Post by **victorhauser** » Sat Jul 01, 2000 8:46 am

There are lies, damned lies, and then there are statistics.

Charles, are you saying that all discussions about pricing are meaningless because we can never get the prices "correct" in the first place due to variables and conditions beyond our control?

Are you saying that the current SPWAW prices are just fine and we are wasting forum bandwidth with these discussions?

Are you saying that prices don't matter?

Are you saying something else?

It occurs to me that you think I favor one particular nation over another. It is not in my best interest to crusade a cause that results in skewed prices for any one nation since I do play human-vs-human games and I'd like to participate in SPWAW tournaments. And since I can't know in advance which nation I'll be playing in those cases, then it's important to me that all prices be as fair as possible.

It also occurs to me that since you do not play human-vs-human games, then the relative prices of units is not nearly as important to you since in a campaign format a player can "adjust" the situations and battles according to whatever prices are listed.

P.S. Wild Bill, I established all parameters as equal and posted them in the first post of this topic when I ran my test battles.

Drake666 · Post by **Drake666** » Sat Jul 01, 2000 9:52 am

Charles22, all victorhauser is trying to do is get some talk going about what unit prices need changing and they do need changing badly. I dont agree much with his tests becouse it was computer vs computer and what we really need to look at is what units people are buying the most of in each year and for what country.

Take the T-43 for example.

T-43 available in Feb 42 at a price of 72
vs the
T-34 M43 available in Oct 42 at a price of 72

Now that is just crazy becouse the T-43 has close to twice the armour of the T-34 M43 and they are the same price. Now who would buy a T-34 M43 as it is now in a Email game.

Tombstone · Post by **Tombstone** » Sat Jul 01, 2000 10:22 am

We also need balance, and some kind of indication of this balance in the game.

We're losing sight of the fact that either Tiger's are underpriced or T-34/85's are overpriced. I also think that the 85's should cost 66% of a Tiger. At 120 I think tiger's are a bit underpriced considering the point value sceme of the rest of the game.
Practically speaking though, when I use T-34/85's I like their speed, how much infantry they can carry, and their rate of fire. I don't really use their armor much at all. It's possible through careful use faster units' mobility in dense terrain to defeat heavier armored and armed units. Under some cicumstances I would rather take 20 85's over 20 Tigers, but the simple fact is that, at least in the SP system, Tigers are more powerful then T-34/85's. I think too much of this thread is focusing around Vic's attempts at a scientific method to his problem(which I applaud), and not enough on the problem itself. We all know how it feels to pit Tiger vs. lighter/faster and vice versa. When you have the luxury of heavy armor and a deadly main-gun you let them come to you and move from good vantage point to good vantage point. When on the ugly end of the stick you really gotta be careful. I dont understand why Victor has to even run a test like that.

Tomo

BA Evans · Post by **BA Evans** » Sat Jul 01, 2000 11:25 am

Hi Victorhauser,

Does it matter which country is listed first? By this I mean, does the first country always get to move first? It might help if each side moved first in 1/2 of the battles.

Just a suggestion,
BA Evans