AI unbeatable (at least in Moscow)?

[DDT]

Are Cossacks in Moscow invincible?

Let me begin by saying I love Shenandoah's Crisis in Command series. I have Bulge and Moscow and will definitely by Alamein, and really enjoy them. Well, mostly.

I have been in Moscow for 5 turns, and I've had it surrounded the whole time. That shouldn't matter because Cossacks are unaffected by supply.

Inside is a 3-pip Cossack. As a 3-pip defender in Moscow, he has a 30% to hit for each of his 3 pips (base 30 + 10 for Moscow defense - 10 for attacking armor. There is no river bonus). Defensively, he is entitled to 2 loss reductions per combat based on the terrain and his size. Supply and isolation are impossible because he is a Cossack.

I attack with one 4-pip armor, with a 40% to hit. I also have 2 elite armor, each with 4 pips, each with a 50% chance to hit (elite armor).

So defender has 3 chances to hit, each with a 30% to hit.
Attacker has 4 at 40% and 8 at 50%. The first two hits will be negated by terrain.

What are the odds of ending up with an undamaged Cossack versus 4 lost elite armor points and 1 regular armor point at the end of those turns? I'm no statistician but that seems ludicrous. What are the odds I would have taken Moscow in two turns? How about one? Any statisticians out there? I'd like to know because I am unlikely to get more offensive strength in Moscow in any future game.

I tracked this data because I felt I was getting way too many unlikely combat results against the AI, and I didn't have it in me to track all of the battles in my game. I was the attacker each time. (Shenandoah, any chance you could give me more time to check out pre-battle information when the AI attacks)?

[s_nkarp]

Those results are unlikely, but all the people who got normal results....aren't writing. Set up the same situation again (easiest in face-to-face), If you get similarly odd results please do send feedback in game (which gives us the data needed to really see what happened). Otherwise we spend a lot of time chasing bugs that are just outliers you'll get in any game with random resolution. It's important that the combat system is right, but weird things do happen -- rarely!

[ijontichy]

I did some of the statistics for your situation, DDT. The chance of your 1 regular armour and 2 elite Axis armour units (each with 4 pips strength) getting a result of 2 hits or less (in other words, doing no damage to the Soviet unit) is about 3.2%. Flipping a coin 12 times and getting 2 heads or less would be a similar result (there I got about 1.8%). But your option is a bit more likely because 4 of the coins are weighted 60:40 towards tails. But only a bit.

So the chances of this happening 4 or 5 times in a row are about as likely as you winning the lottery. Multiply 0.032 by itself a few times and see what you get. So either there's a bug or we both haven't understood the rules properly. The rules are a bit tricky to understand, you can definitely misinterpret or miss something if you're not careful enough. But I've read what you wrote, and it seems your interpretation of the rules are correct.

My only suggestion is to upgrade to 2.0.1 if you haven't done so already. I found that the statistical results seemed fairer in this new version. My evidence for this is a bit sketchy, relying on a small number of tests and general feel of the game.

[Ryvin]

Has Moscow defensive work??? I do not think so... miss something?

[s_nkarp]

There's no 10% bonus for Moscow - just the 2 pips for the city + 2 for forest.