Combat inconsistency (especially on extreme) / potential bug

[WeaverofBrokenThreads]

The first time I lost a heavy battledress soldier to a slugthrower (2 final att score versus 1200 final hp) I thought 'Man, this guy used his whole life's reserve of luck to land that kill shot.'

After the third or fourth one, I stopped really caring. I got used to it. I sort of got a 'feel', a sort of intuition, for how well my troops will do even without knowing the exact science behind their performance. But then I started playing extreme, and the enemy obliterates my regular troops with a measly +20% mod. 'That can't be right' I thought. Then they obliterated my tanks as well.

I think there is something very wrong with the combat resolution code that merits taking a look at it. I think it is related to the AI specifically, as these things never seem to happen to me on attack or defense.

Conversely, even if I launch 2k attack score attacks on 100 final hp, the results are far from expected.

I can't post a picture because the forum hates me, but just imagine looking at the individual reports screen and it says: 67.3 vs 1964. It is a KILL result.

I understand there might be hidden mechanics, like critical hits or something, but I have not seen a single reference to this in any of the bonuses. The enemy did not have a champion, apoc paladin or any sort of unit feat that turns hits into kills - and to begin with, the idea that they got a hit in the first place is ludicrous. This happens way too often.

If there are hidden mechanics at play, can I please get an elaboration on how they work?

And on the other point, I feel like the AI gets much higher bonuses than the 20% that is reported. I cannot explain why, but the difference between a regular game and an extreme difficulty game are like night and day, even when fighting against similar numbers and composition of troops - and that just cannot be accounted for with a +20% mod. They are far deadlier. I don't think they fight harder, as in, ignoring more casualties, but I think the +20% mod to health is also quite a bit higher.

[Saros]

Here is the image weavre was talking about.

[WeaverofBrokenThreads]

Thank you!

[Saros]

Looking at this again where is the caliber modifier? The rifle militia is presumably using small arms against a Battledress trooper so should suffer a huge penalty?

[WeaverofBrokenThreads]

Doesn't appear to work currently. I can confirm that based on the individual reports, calibre matrix is not take into account at all against vehicles or personal armor.

[zgrssd]

ORIGINAL: Saros



Here is the image weavre was talking about.

I can actually tell you the odds now!
51 vs 1344 (before modifiers):
69940 possible combinations
1326 Attacker win combinations.
1.89% Attacker Hit chance.

I can not do fractions that precisely for after modifiers, but 68 vs 1964:
135585 Combinations
2346 attacker wins
1.7% Attacker hit chance

Almost 2% is actually a lot better then I thought. And my code is doing this the hard way, so I can exclude discrepancies from shortcuts.

[zgrssd]

2000 vs 100:
20301 Combinations
15150 Attacker wins
74.62%
Even with 1 order of magnitude, still only 75% hit chance.