Profile of Grumpy

Hello everyone,

I couldn’t find an answer on this but, if this has already been discussed, please let me know.

I’m curious what drives the decision making for the AI. I’m aware they “cheat” with resources and understand their general build orders, but I have no clue what the priorities are for AI and so on. I’ve noticed the computers have not adapted to the latest patches well, so I’m thinking there was some stuff hard coded in there which didn’t translate well to the current game state.

Did anyone ever crack the code?

I can help with this. I love cracking codes.

It depends on what type of game you're playing. If it is just you and a bunch of computers then this is what happens:

Soviet, OKW, OST, and USF computers play somewhat normally. It normally seems like Soviet>OKW>OST>USF, but there aren't huge differences.

UKF is somewhat different. It will normally start by building emplacements in locations that it can't defend. Then after it loses everything, it will try to turtle by stacking everything it has left on top of your stuff. This makes it easier for the enemy AI by not requiring it to decide which targets to arty. This may seem disappointing at first, but it was carefully modeled by Relic after watching the behavior of actual UKF players in game. It's kinda your fault for including them even in a comp stomp.

Hope this helps.....

I messaged the creator of the COH2Stats site and got a clarification. My numbers should've been:

Week Start 15-Jun 22-Jun 29-Jun 6-Jul 13-Jul
Games Played 134 111 99 70 69
Axis Win 75 64 65 46 45
Axis Win % 56.0% 57.7% 65.7% 65.7% 65.2%
OST Player % 43.7% 40.0% 41.2% 44.6% 39.5%
OKW Player % 56.3% 60.0% 58.8% 55.4% 60.5%

Axis Win % STDEV 0.048146573

There are 483 games since the patch with all players in the top 200. The win rate is pretty stable around 65% Axis. The standard deviation of win rate is 0.048 so the average is a little over 3 sigma above 50%. That means there is about a 0.3% chance that this game mode is balanced and these are just abnormal results.

I'd need to be a lot better at multivariate analysis to say if the Ost versus OKW percentage actually has an effect on the outcome or is simply a measure of popularity.

(edit - here are the 2v2 numbers, which are also relatively stabile)
Week Start 15-Jun 22-Jun 29-Jun 6-Jul 13-Jul Stdev
Games Played 619 498 407 363 325
Axis Win 300 228 196 165 153
Axis Win % 48.5% 45.8% 48.2% 45.5% 47.1% 0.013565168

These numbers say the opposite, not quite to the same degree, but in this case enough to say it's a little unbalanced towards Allies. I do think that it has a lot to do with how the units and maps interact. The 4v4 maps aren't nearly wide enough, so most engagements are going straight out from your base with little room for flanking or strategic positioning. Units are a lot more clumped together so a unit like a walking stuka, that isn't good in a small game mode, becomes great in a 4v4. There's nothing worse than having a random teammate get beat in their lane in a 4v4 and have them decide to stack their units on top of mine.

I don't think it's very surprising that smaller games tilt towards axis. A lot of times it feels like I get pushed around by Penal blobs until I get the Sturmtiger out.

Every time I think to myself: well, I need an MG, how bad can the Maxim be. And it turns out that this MG can be really really fucking bad. Basically every time I build that pos, it turns out to be a grave mistake. Even floating MP would help you more than this crap. It just shows that the balance team has no idea what they are doing, taking away every bit of uniqueness (Maxim with fast setup times but small arc) and making these units just unbearably worse (or buffing other units to levels previously unknown to mankind...)

The maxim doesn't do any better in 3v3's. I started getting a lot better results by doing either 4 con starts or 3cons and second engie. The setup time was nerfed to match the MG42. The arc is about 90 versus 120? for either of the Axis MG's. It might not sound like a huge difference, but it often takes 2 maxims to cover what 1 MG42 covers. A MG42 will suppress more or less instantly but the maxim can have units walk through the arc and often not suppress.

Overall, the balance doesn't seem that bad unless you're playing in the Sturmtiger league (4v4), just don't get maxims and outblob with Cons or Penals.

Yeah but they've nerfed its penetration since then. And the barrage was nerfed a tiny bit too I think

That said I think free barrage would be mistake. The cost isn't that much and Soviets don't have heavy muni drain to begin with

I don't want free barrage because I don't want to play against it. Buffing the damage to 160 would help it against mediums but wouldn't affect heavies that much. I think the penetration nerf was done specifically to detune it against heavies and it would still be an effective nerf even if the damage was buffed to 160.

A small cost decrease to the barrage would be helpful. I play a lot of Soviet and think it has a heavy muni drain.

120*1.167=140
640/120=5.3≈6 penetrating shots to kill.
640/140=4.5≈5 penetrating shots to kill

If the game was played on a spreadsheet, that would matter. Single AT sources rarely kill anything. The SU76 often ends up paired with a ZIS or a SU85. They usually hit and pen P4's. The alpha damage is 280 or 300, but it will take 3 more penetrating hits to kill a medium in either case.