Possible DLC bug

[Taxman66]

I admit I don't currently own SC ACW or the DLC.

I have been watching the Franco Prussian War scenario on Youtube between OCB and GwtC.

I have noticed that the (land obviously) combat odds are often inaccurate after the first attack on a unit.
The first attack (against full strength defenders) is always accurate within the +/- 1 random of the predicted combat odds.

The second attack (typically against a defender now at 4-6 strength after the first attack) then becomes inaccurate in regards to damage suffered by the attacker.
For example, I have seen the combat predictor state 5:4 (with a 10 strength attacker against a 4 strength defender (after having suffered 6 damage on the first attack) and the actual results is the defender eliminated (accurate) while the attacker takes 2 or 3 damage (inaccurate).

For the record, I more than occasionally see this in WiE naval combat with less than full strength defending surface ship. Again I recall this is when defending ship is suffering a 2nd (or later) attack in the same turn. Don't recall (though I could be wrong) if it happens against a damaged defending ship on the first attack.

[Hubert Cater]

Thanks Taxman, in case this is not a big ask, could you point me to the video and the time stamp so I could take a look?

[Taxman66]

Here's one example, I'm not going to hunt others but they exist if you need them. Look at the heavy fighting around Metz, and I'm sure it happens for both sides.

Gaming with the Colonel's Blood and Iron Episode 35
https://www.youtube.com/watch?v=LWqnjcF-Lkg
First Attack on defender at 2:38 mark. Odds show 6:6, results are 6:6
Second Attack on same defender (str 4) at 2:53 mark. Odds show 5:4, results are 2:4 - defender destroyed.

[Hubert Cater]

Thanks and it looks the French unit that was attacked, was set to retreat, but never did because it was destroyed.

For background, when a unit is set to retreat, the attacker's losses are halved, and in this case 5 reduced to 2.5 and then rounded down to 2.

I'll look into adding a 'shattered' message to a unit that was set to retreat but subsequently destroyed so hopefully this will be clearer for cases like this.