Bug-like Battle AI behavior

[PDF]

Chazar said:

I disagree:
Consider a single militia man, gone berserk by spell, which was in his own squad, put on a rear flank with "hold and attack rearmost". Now the battle is almost over, the wounded enemy masses are fleeing into a friendly province just to heal, regroup and attack again, but your heavy infantry which has advanced to the other end of the battlefield (since it was ordered to attack archers) could easily dispatch those fleeing units. But instead, they let those fleeing units unharmed and chase after that single militia man, which they wont even reach in time!

I would rather have my infantry finish of those fleeing units, or at least I would want some of them to chase the militia man, but certainly not all of them! Different circumstances should really matter (including distance as you have already pointed out), hence my suggestion below!

Bah, you're taking an extreme example to show that dealing with a real threat, but a very feeble one, can be sub-optimal. Yes it is !
To be more precise I agree that the AI should only dispatch forces proportionate to a threat, not necessarily all the army vs "the" main threat. Else everyone would gang up against 1 unit/group..
But anyway I'd still prefer that to seeing a whole army destroyed by *not* dealing with a very potent threat !

[Chazar]

PDF: I am not disagreeing with you, all I say is that it might be pretty difficult to make the AI behave in the way you desire...

[Pocus]

a 'potency' (threat in this case) parameter have not to just take into account what is the current status of the unit (routing, or being a badass Niefel berseked), it can also use a memory of what the unit did, by counting the gold/gem value of units killed by this threat. For example, as soon as the Niefel start to kill Seraphs, his threat value would go through the roof, and he would be targetted again by the seraphs?
just my two eurocents.

[PDF]

I checked the battle - indeed the Niefel had a Ring of Tamed Lightning, that I didn't notice at first, so that he was not targetted by Bolts makes sense...
Apologies to have implied the AI was so stupid has to not cast Bolts at him out of sheer stupidity .
Well, it's not clear why the ghosts wolves/ph warriors didn't go at him - maybe even they did try to but he was flying, making it difficult to reach. Overall my army was mainly powerless against him, the only sensible action of the seraphs should have been to leave .
No real "bug" indeed...

[The Panther]

I believe that your own army will never rout if the enemy has already completely routed. This happens regardless of how hopeless it may be against a berserked SC that stayed to fight after his side previously declared that the battle was 'over' by retreating. Your Seraphs will then all fight to the death because the AI had previously decided you had the battle won and were in the mop-up phase, regardless of the actual truth of the situation. I also believe that this never changes until the 50 turn limit is reached, even if your own army starts dying like crazy and ought to then rout on its own.

This seems to be yet another indication that the AI is not programmed to handle SCs at all.

[deccan]

The Panther said:
I believe that your own army will never rout if the enemy has already completely routed. This happens regardless of how hopeless it may be against a berserked SC that stayed to fight after his side previously declared that the battle was 'over' by retreating.

I agree. This is also very apparent with armies that have immortal commanders fighting in friendly dominion. Also, in the situation you described, even if all of their commanders are killed and they are left leaderless, all troops will still stay on the battlefield and be slaughtered to the Last man.

[Cainehill]

Pretty sure the armies can / will rout even after the opponent has routed, as I had someone routed and the battle won when _my_ Caelumians routed (due to spell damage or some such, I believe). So I lost the province, even though the enemy routed first, since the flyers could insta-retreat off the field.

[The Panther]

I have seen this before also, where both armies esentially rout in the same turn, or maybe the following turn or two due to poison damage. But I have never seen an army rout several turns later after a full-out retreat by one side. Immortals and berserked units seem to always count as fully retreated in this, even though they stay to fight to the end.

Were there several turns between the enemy routing and your army routing?

[Cainehill]

Eh - it couldn't have been _many_ turns, or the other side would have managed to retreat. But there definately weren't any berserkers or immortals involved, simply routed troops that were slow enough the other side managed to turn tail as well.