Matrix Games Forums

So I'm not an expert by any means, but in every attack I remember, odds of 2.0 or better force a retreat, and odds of 1.99 or less get a held result. I'm on turn 6 and attack Odessa with the help of pioneers and heavy artillery. First dedicated attack gets 2.42 to 1, but it's a held result. All the divisions have enough MP for another dedicated attack. This time it's 6.83 to 1. Another held... Is there a different criteria for forts or cities to force a retreat/surrender?

Yes if you are on v1.12 (not in v1.11) - you need to get the fort below a certain level before it actually will retreat.

You should check the 1.12 update list for the exact mechanics, but roughly summarized is this:
Fort 4 and above prevents retreat if city or urban hex.
Small ports and light urban add 1 effective level for this calculation, large ports (>12 I think, including Odessa) and heavy urban by 2.

The "fort level" used in this calculation is the sum of the previous factors on top of the post-combat fort level. However, the tricky part is that this does not include the bonus fort level removal, which is equal to sqrt(CV) * total (engineer + artillery) value.

Here, you see that you have 209 engineer + artillery value. This is enough to remove 0.02 * 2 of the level 4 fort, taking it down to 4.04. The bonus fort reduction is sqrt(6.83) * 209, so 546 total points. 300 is used to take from 4.04 to 3.98, then the remainder 246 is used to reduce from 3.98 to 3.72. If you watch the battle on relatively slow mode, it'll pause to show you the intermediary pre-bonus fort value of 4.04, but you can't see this afterwards.

I lost a couple of light urban cities to direct assault versus tyronec.

My own impression was its easier to retreat these troops than before the change.

Urban should be very hard to lower forts and i think game is giving too much credit to engineer in reducing fort level.