Development Discussion: Speeding up battle calculator (and thus Hard AI)

[Trevan]

@RoiEX , @LaFayette , I've written up a proof of concept for the tree. It isn't really expectiminimax since it isn't trying to find the best path. It is just calculating the win/lose/tie probabilities but it follows the same general idea.

I've tested it with some 1940 Global units (that don't give any support) and the probabilities are inline with the battle calculator. I've also done a few small trees by hand to verify the calculations are working.

You can see the code at https://github.com/trevan/triplea/commit/1658df71e12e8fb13ee09d5e2f6a33d1dc91f6ea. It doesn't do every battle situation (such as AA, supports, two hits, etc). I wrote a test file that mocks out ProData so I can test out the implementation.

What do you think? Should I create a WIP PR so discussion can happen in github?

[RoiEX]

@Trevan I must say I'm impressed, the code definitely looks promising.

I don't really have any objections to create a PR, but I'd rather have a way to actually make use of this class. If it's just gathering dust in the deoths of the repository it would be an unfortunate waste of time.

I think it would be a good idea to compare the new "IBattleCalculator" implementation with the 2 (technically 3, but whatever) existing implementations to see how close the estimated results are. Data is key here.

After that we could consider replacing one of the implementations for the AI, or just add an experimental AI in order for people to test this.

Ideally we get to a point where the "real" simulation results are super close to the estimated results 99% of the time so we're able to use the lightweight code for the hard ai.
We could also try to set up some custom maps to let your implementation battle it out in simulated games against the current hard ai using different IBattleCalculator implementations each.

Not sure what @redrum thinks about this, but I think he'll agree that this definitely looks promising

[redrum]

@RoiEX Agree. At a minimum if we got a decent implementation that's very fast then could probably replace Fast AI's calc as it just uses a very simple strength formula (low bar here).

[Trevan]

@RoiEX, I was thinking the PR would be useful just for a place for discussion. But I'm fine with leaving it as is.

I decided to test a battle with 100 infantry attacking 100 infantry and it never finished so I need to look into that.

I've compared the estimates to the FastOddsEstimator and to the BattleCalculator in the UI (I think that is ConcurrentBattleCalculator). The results are basically the same as what ConcurrentBattleCalculator returns (if you use units that don't support each other).

After I've figured out the performance issue, I'll work on adding handling support units so that I can do better comparisons and then I'll post them here.

[Trevan]

I've improved the performance and added support handling.

I used TestMapGameData.GLOBAL1940 for the comparisons. I got numbers for ConcurrentBattleCalculator with 200 rounds and 2000 rounds. I also used the PerfTimer to calculate how long it took to run:

1 infantry, 1 armour vs 2 infantry
Hard AI 200 runs: 1984.9 ms, Win: 0.43, Lose: 0.425
Hard AI 2000 runs: 1703.1 ms, Win: 0.4931640625, Lose: 0.388671875
MiniMax: 39.8 ms, Win: 0.506233909545216, Lose: 0.38491296598075025

1 armour, 1 tactical_bomber vs 2 infantry, 1 fighter
Hard AI 200 runs: 2031.1 ms, Win: 0.05, Lose: 0.95
Hard AI 2000 runs: 1709.8 ms, Win: 0.072265625, Lose: 0.9228515625
MiniMax: 56.0 ms, Win: 0.1150459321571266, Lose: 0.7947330200775955

1 infantry, 1 artillery vs 2 infantry
Hard AI 200 runs: 2091.1 ms, Win: 0.425, Lose: 0.52
Hard AI 2000 runs: 1740.2 ms, Win: 0.4521484375, Lose: 0.4580078125
MiniMax: 39.1 ms, Win: 0.4574200859119973, Lose: 0.45742008591199734

1 destroyer, 1 cruiser vs 1 destroyer, 1 cruiser
Hard AI 200 runs: 1832.9 ms, Win: 0.39, Lose: 0.425
Hard AI 2000 runs: 1616.8 ms, Win: 0.4228515625, Lose: 0.427734375
MiniMax: 36.1 ms, Win: 0.4250031963247701, Lose: 0.4250031963247701

1 destroyer, 1 cruiser, 1 battleship vs 1 destroyer, 2 cruiser
Hard AI 200 runs: 2019.2 ms, Win: 0.725, Lose: 0.18
Hard AI 2000 runs: 1787.6 ms, Win: 0.7939453125, Lose: 0.14453125
MiniMax: 39.0 ms, Win: 0.5358330747385637, Lose: 0.34400771480746317

1 infantry, 1 fighter vs 1 infantry, 1 aa gun
Hard AI 200 runs: 1844.4 ms, Win: 0.395, Lose: 0.605
Hard AI 2000 runs: 1727.8 ms, Win: 0.4306640625, Lose: 0.5625
MiniMax: 32.2 ms, Win: 0.6961033280443275, Lose: 0.2000276003717069

The numbers look pretty close except when AA guns or two hit battleships are used. I don't handle those.

[redrum]

@Trevan The speed is quite impressive. Do you think adding in additional unit properties like multi-hit units and AA rolls would be achievable? It might also be interesting to compare it to Fast AI as that uses a very simplistic formula but is very fast.

[Trevan]

@redrum here's the numbers rerun with FastAi's calculator included.

1 infantry, 1 armour vs 2 infantry
MiniMax: 93.5 ms, Win: 0.506233909545216, Lose: 0.38491296598075025
Hard AI 200 runs: 1787.9 ms, Win: 0.47, Lose: 0.435
Hard AI 2000 runs: 1471.6 ms, Win: 0.501953125, Lose: 0.388671875
FastOdds: 6.8 ms, Win: 0.5

1 armour, 1 tactical_bomber vs 2 infantry, 1 fighter
MiniMax: 2.2 ms, Win: 0.1150459321571266, Lose: 0.7947330200775955
Hard AI 200 runs: 1025.3 ms, Win: 0.085, Lose: 0.91
Hard AI 2000 runs: 1194.6 ms, Win: 0.06640625, Lose: 0.9287109375
FastOdds: 0.5 ms, Win: 0.34082222539412593

1 infantry, 1 artillery vs 2 infantry
MiniMax: 2.9 ms, Win: 0.4574200859119973, Lose: 0.45742008591199734
Hard AI 200 runs: 1038.0 ms, Win: 0.465, Lose: 0.405
Hard AI 2000 runs: 1385.9 ms, Win: 0.4521484375, Lose: 0.4619140625
FastOdds: 0.4 ms, Win: 0.5

1 cruiser, 1 destroyer vs 1 cruiser, 1 destroyer
MiniMax: 1.6 ms, Win: 0.4250031963247701, Lose: 0.4250031963247701
Hard AI 200 runs: 1047.1 ms, Win: 0.405, Lose: 0.46
Hard AI 2000 runs: 1140.1 ms, Win: 0.412109375, Lose: 0.4345703125
FastOdds: 0.7 ms, Win: 0.5

1 cruiser, 1 destroyer, 1 battleship vs 2 cruiser, 1 destroyer
MiniMax: 3.1 ms, Win: 0.5358330747385637, Lose: 0.34400771480746317
Hard AI 200 runs: 1032.0 ms, Win: 0.775, Lose: 0.165
Hard AI 2000 runs: 1141.3 ms, Win: 0.8125, Lose: 0.1328125
FastOdds: 1.6 ms, Win: 0.6591777746058741

1 infantry, 1 fighter vs 1 infantry, 1 aa gun
MiniMax: 1.3 ms, Win: 0.6961033280443275, Lose: 0.2000276003717069
Hard AI 200 runs: 1138.5 ms, Win: 0.38, Lose: 0.615
Hard AI 2000 runs: 1248.8 ms, Win: 0.4130859375, Lose: 0.583984375
FastOdds: 0.2 ms, Win: 0.7180577051181194

Fast Ai is really fast. Not sure how to compare the win percentage.

I have an idea on how to do multi-hit and AA. I'm going to attempt it. I also need to deal with units with multiple rolls. Is there a map in TestMapGameData that has units like that?

[Cernel]

@Trevan said in Development Discussion: Speeding up battle calculator (and thus Hard AI):

MiniMax: 93.5 ms, Win: 0.506233909545216, Lose: 0.38491296598075025
Hard AI 200 runs: 1787.9 ms, Win: 0.47, Lose: 0.435
Hard AI 2000 runs: 1471.6 ms, Win: 0.501953125, Lose: 0.388671875
FastOdds: 6.8 ms, Win: 0.5

How is 200 runs slower than 2000 runs?

[alkexr]

@Trevan Unfortunately this sort of algorithm has to be exponentially slow as the number of units increases, by its nature. I'm saying this without even looking at the code. With some cleverness it might be possible to bring this down to exponential with the number of unit types (not units), but will never be able to handle if, say, a hypothetical map had dozens of unit types, many of them targeting each other unpredictably with all sorts of AA attacks, and battles between stacks of 50 happening regularly. But the speed increase is extremely impressive, and even if the AI only used this to simulate smaller battles while keeping the battle calculator for battles that are too large for this algorithm, that's already a massive breakthrough in improving performance.

[Trevan]

@alkexr here's some numbers of larger groups:

10 infantry, 10 artillery vs same
MiniMax: 219.3 ms, Win: 0.497401100632783, Lose: 0.4974011006327829
Hard AI 200 runs: 2092.2 ms, Win: 0.475, Lose: 0.515
Hard AI 2000 runs: 2086.4 ms, Win: 0.4658203125, Lose: 0.5283203125
FastOdds: 8.9 ms, Win: 0.5, Lose: 0.0
20 infantry, 20 artillery vs same
MiniMax: 585.7 ms, Win: 0.49887594912628747, Lose: 0.49887594912628747
Hard AI 200 runs: 2538.8 ms, Win: 0.54, Lose: 0.46
Hard AI 2000 runs: 3255.3 ms, Win: 0.5029296875, Lose: 0.494140625
FastOdds: 14.7 ms, Win: 0.5, Lose: 0.0
30 infantry, 30 artillery vs same
MiniMax: 1244.4 ms, Win: 0.49930484393068525, Lose: 0.49930484393068525
Hard AI 200 runs: 2873.4 ms, Win: 0.5, Lose: 0.5
Hard AI 2000 runs: 3627.7 ms, Win: 0.501953125, Lose: 0.49609375
FastOdds: 15.5 ms, Win: 0.5, Lose: 0.0
40 infantry, 40 artillery vs same
MiniMax: 2032.1 ms, Win: 0.4995101769999315, Lose: 0.4995101769999315
Hard AI 200 runs: 2904.5 ms, Win: 0.505, Lose: 0.49
Hard AI 2000 runs: 4244.6 ms, Win: 0.51171875, Lose: 0.48828125
FastOdds: 16.9 ms, Win: 0.5, Lose: 0.0
50 infantry, 50 artillery vs same
MiniMax: 3366.8 ms, Win: 0.4996227075274397, Lose: 0.4996227075274401
Hard AI 200 runs: 3574.4 ms, Win: 0.49, Lose: 0.51
Hard AI 2000 runs: 4614.9 ms, Win: 0.4853515625, Lose: 0.5146484375
FastOdds: 16.8 ms, Win: 0.5, Lose: 0.0
60 infantry, 60 artillery vs same
MiniMax: 5306.7 ms, Win: 0.4996976675867421, Lose: 0.4996976675867422
Hard AI 200 runs: 3264.8 ms, Win: 0.54, Lose: 0.46
Hard AI 2000 runs: 5255.6 ms, Win: 0.4931640625, Lose: 0.5068359375
FastOdds: 22.5 ms, Win: 0.5, Lose: 0.0
10 infantry, 10 artillery, 10 armour, 10 fighter, 10 tactical_bomber, 10 bomber vs same
MiniMax: 12470.0 ms, Win: 0.8637525783579469, Lose: 0.13152813492843518
Hard AI 200 runs: 2565.5 ms, Win: 0.32, Lose: 0.68
Hard AI 2000 runs: 2721.8 ms, Win: 0.3125, Lose: 0.6875
FastOdds: 23.4 ms, Win: 0.6286565968532855, Lose: 0.0

Interesting that the battle with 30 infantry and artillery was a lot faster than the one with 10 of each land type. Both battle has a total of 60 units on both sides but the addition of multiple unit types slowed it down.

[alkexr]

@Trevan Hmm, right, you can actually do this in cubic (?) time, so long as the order of losses is linear, because the number of casualties already sustained uniquely determines the composition of the remaining army. But I still don't see how you do it in less than (number of units)^(number of target lists * constant) time, where target lists include normal attacks, AA attack types, canNotBeTargetedBy and related mechanics, etc. Because, without assuming something about the order in which units die, technically there are a million different army compositions strictly smaller than "10 of each land type", as opposed to 60 if you assume a linear order of losses.

[Trevan]

@alkexr yes, my algorithm is assuming the order of losses is linear. I believe the time is roughly what you said. Each target group will have its own order of loss.

[Trevan]

AA guns were a lot harder than I expected but I think I got the majority of it working. I did a comparison on the TestMapGameData.TWW data because it had units that could do aa firing with different target groups. I ran a battle between an infantry, tank, and fighter vs an anti-tank gun, mobile artillery, and anti-air gun.

MiniMax: 197.7 ms, Win: 0.5520100874509435, Lose: 0.40334641519178716
Hard AI 200 runs: 2025.6 ms, Win: 0.57, Lose: 0.42
Hard AI 2000 runs: 3222.8 ms, Win: 0.5517578125, Lose: 0.412109375
FastOdds: 7.9 ms, Win: 0.7703534497836491, Lose: 0.0

Now I'll work on multi-hit units.

[Swampy]

Since I play against the AI quite a bit.. would love this project to succeed as Hard AI is ridiculously slow on large maps/lots of units. Would like to contribute.. but have very little coding skill unfortunately. If you need testers or other basic help... let me know.

[Trevan]

@redrum @RoiEX
I think I have a good majority of the battle logic done. I know I'm missing bombardment, some details in sub first strike (order of strikes and destroyers presence), transport handling, amphibious battles, and paratroopers. There's probably other complexities that I missed as well. I've also worked on speeding it up. Here's a bunch of comparisons:

30 infantry, 30 artillery, 30 armour, 30 fighter, 30 tactical_bomber, 30 bomber vs same
MiniMax: 21218.5 ms, Win: 0.9779225005914358, Lose: 0.021637464729431084
BattleCalculator: 1841.4 ms, Win: 0.95, Lose: 0.045
Hard AI 200 runs: 2893.2 ms, Win: 0.955, Lose: 0.045
Hard AI 2000 runs: 7900.4 ms, Win: 0.970703125, Lose: 0.029296875
FastOdds: 29.4 ms, Win: 0.6517051232809282, Lose: 0.0

10 infantry, 10 artillery, 10 armour, 10 fighter, 10 tactical_bomber, 10 bomber vs same
MiniMax: 1342.4 ms, Win: 0.8637259513264075, Lose: 0.131552773090954
BattleCalculator: 459.2 ms, Win: 0.895, Lose: 0.105
Hard AI 200 runs: 1280.3 ms, Win: 0.87, Lose: 0.12
Hard AI 2000 runs: 2414.4 ms, Win: 0.8486328125, Lose: 0.1455078125
FastOdds: 7.4 ms, Win: 0.6286565968532855, Lose: 0.0

1 infantry, 1 armour vs 2 infantry
MiniMax: 1.4 ms, Win: 0.506233909545216, Lose: 0.38491296598075025
BattleCalculator: 74.3 ms, Win: 0.505, Lose: 0.365
Hard AI 200 runs: 987.0 ms, Win: 0.495, Lose: 0.39
Hard AI 2000 runs: 1000.7 ms, Win: 0.501953125, Lose: 0.3974609375
FastOdds: 0.5 ms, Win: 0.5, Lose: 0.0

1 armour, 1 tactical_bomber vs 2 infantry, 1 fighter
MiniMax: 1.2 ms, Win: 0.11504593215712659, Lose: 0.7947330200775955
BattleCalculator: 99.8 ms, Win: 0.115, Lose: 0.775
Hard AI 200 runs: 1156.5 ms, Win: 0.175, Lose: 0.75
Hard AI 2000 runs: 1244.3 ms, Win: 0.1201171875, Lose: 0.7958984375
FastOdds: 0.2 ms, Win: 0.34082222539412593, Lose: 0.0

1 infantry, 1 artillery vs 2 infantry
MiniMax: 1.0 ms, Win: 0.4574200859119973, Lose: 0.45742008591199734
BattleCalculator: 84.5 ms, Win: 0.49, Lose: 0.44
Hard AI 200 runs: 1095.0 ms, Win: 0.51, Lose: 0.41
Hard AI 2000 runs: 1231.1 ms, Win: 0.4619140625, Lose: 0.443359375
FastOdds: 0.2 ms, Win: 0.5, Lose: 0.0

1 cruiser, 1 destroyer vs 1 cruiser, 1 destroyer
MiniMax: 2.2 ms, Win: 0.4250031963247701, Lose: 0.4250031963247701
BattleCalculator: 86.0 ms, Win: 0.395, Lose: 0.44
Hard AI 200 runs: 1076.7 ms, Win: 0.405, Lose: 0.465
Hard AI 2000 runs: 1213.1 ms, Win: 0.4306640625, Lose: 0.4189453125
FastOdds: 0.4 ms, Win: 0.5, Lose: 0.0

1 cruiser, 1 destroyer, 1 battleship vs 2 cruiser, 1 destroyer
MiniMax: 3.9 ms, Win: 0.7898869382202083, Lose: 0.14239504638191994
BattleCalculator: 122.6 ms, Win: 0.76, Lose: 0.125
Hard AI 200 runs: 1076.1 ms, Win: 0.79, Lose: 0.15
Hard AI 2000 runs: 1188.0 ms, Win: 0.79296875, Lose: 0.1474609375
FastOdds: 0.7 ms, Win: 0.6591777746058741, Lose: 0.0

1 infantry, 1 fighter vs 1 infantry, 1 aa gun
MiniMax: 11.1 ms, Win: 0.5600126123377998, Lose: 0.35180777879589964
BattleCalculator: 189.6 ms, Win: 0.585, Lose: 0.315
Hard AI 200 runs: 1127.4 ms, Win: 0.595, Lose: 0.335
Hard AI 2000 runs: 1410.4 ms, Win: 0.5869140625, Lose: 0.3203125
FastOdds: 0.3 ms, Win: 0.7180577051181194, Lose: 0.0

All of the win percentages are pretty darn close. The 30 of each unit type battle was slower than the original calculator (21 seconds vs 3 seconds) but the 10 of each unit type battle was about the same (1.34 seconds vs 1.28 seconds). The rest of the battles I did were small so it was a lot faster.

What would you like me to do now? Can I make this available to others (such @Swampy) so they can test it out?

[RoiEX]

I'm not 100% sure what @redrum's opinion on this is, but I'd really like to see this as an experimental Ai in the game.
But for now hide it behind the test flag?

[Alexei Svitkine]

Perhaps it could be added as a checkbox in the battle calculator for now as well?

Otherwise, I think the two things it needs is a heuristic for when it should be used (e.g. if too many units or when there are unsupported cases, it should do the usual battle calculator instead) and some kind of way to find cases where it disagrees with battle calculator. For example, maybe run some percentage of simulations using both and send error reports when the results disagree too much? Or perhaps we can have an experimental AI that always runs both so we can use it to validate it and find cases where it's incorrect?

[LaFayette]

An AI variant IMO would be an excellent step. The idea of computing the actual odds vs the current monte carlo approximation is a very good one.

The combat/battles code suffers from complexity and is already somewhat time-optimized. It would be very easy to underestimate the level of effort to replace the battle calc (simply having something that mostly works is just not quite enough, it would need to be cleanly coded, not be duplicative, well tested, complete, etc...)

[Trevan]

@RoiEX I'm not aware of the test flag. How do I use it?

I can make a sibling of FastAI that uses this calculator. I also like the idea of having it automatically compare the actual odds vs the new odds. @Alexei-Svitkine when you say "send error reports", how would I do that? Is there existing classes in the code that I can just use to send the reports?

[LaFayette]

    if(ClientSetting.showBetaFeatures.isSet()) {
        // add experimental feature
    }