Should the siege defence bonus in auto-resolve battles be decreased for the player?

[Totalgarbage]

Quite a few patches ago, the devs made an adjustment to how the enemy AI calculated the strength of player's troops by changing it to the AI perceiving the player's strength in half and then re-adjusted it to 3/4 or 4/5 of the party's strength iirc so that players are able to do siege defence battles more often. While this adjustment has been successful in letting the players experience siege defences, it also made auto-resolve for siege defences too broken for the player. I am suggesting that the siege defence bonus for the player's party in auto-resolve battles should be decreased by the same amount for the player to make it more fair while still giving the player a very noticeable advantage in siege defences. So if regular AI get a 300% boost in siege defences, the player should get 225% bonus instead if the AI calculates the player's party strength as 75% (instead, if they calculate it as 80%, give a 240% bonus, etc.).

 

[Piconi]

While this adjustment has been successful in letting the players experience siege defences, it also made auto-resolve for siege defences too broken for the player

Could you provide an example of a defensive siege that was broken by auto-resolve in your opinion (number of attackers vs defenders, army compositions...)?

 

[Totalgarbage]

Could you provide an example of a defensive siege that was broken by auto-resolve in your opinion (number of attackers vs defenders, army compositions...)?

Oh sorry, by broken I meant overpowered, as in auto-resolving siege defences can give the player a "free" win (though a lot of soldiers die of course) quite often because the AI is made to intentionally underestimate the player's party strength so that players are more likely to experience siege defences.

The enemy AI decides to siege the player after estimating only 75% or 80% of their actual party strength, and if we say that the bonus to units in auto-resolve is 300% for defenders (I'm not counting unit type bonuses/maluses to auto-resolve and assuming that both the attackers and defenders only have tier 3 troops and 0 in tactics for simplicity's sake) the enemy should realistically use 300 + ~30 troops to be on the safe side for every 100 soldier the player has to win an offensive siege battle (and the player obviously has the garrison & militia on their side so the numbers in reality would be 250 vs 750 + ~75). However, what actually happens is the enemy AI sees your 100 troops as 75, and therefore thinks that their 225 units would be sufficient to win this siege battle and gets stomped if the player actually auto-resolves.



@1.15, Strat demonstrates that he can win the "unwinnable" siege defence in 2 seconds by simply auto-resolving. This video is from a year ago and shows the state of things in 1.0 official release. Auto-resolve bonuses and maluses depending on unit types, terrain and weather have been added with 1.2, but the premise still remains. I unfortunately cannot record a video due to hardware problems, but you should be able to test this yourself in seconds.

 

[Piconi]

We did, in the meantime, give archers a considerable siege defense bonus in auto-resolving. Could it be that you also had a bit higher archer ratio? Players often do, myself included.

 

[Totalgarbage]

My argument isn't about the unit specific bonuses to auto-resolve. My argument is basically:

-TaleWorlds changed the AI's calculation of player's party strength when the player is in a castle or a town, so that the AI perceives the player's party strength as ~3/4 of its actual total strength, and as a result, the players are able to experience defensive siege battles more often (which is good!).

-However, this deliberate underestimation of the player's strength can often result in the AI attempting to siege the player with insufficient party strength of their own: If the AI normally needs to attack with a strength of 300% of the defenders' total strength to be equal to them in most siege offences (both parties having 0 in tactics and the exact same troop types and ratios), AI when attacking the player thinks that they need to have 225% of the player's party strength to win an offensive siege against them.

-Therefore, I'm arguing that the player's, and only the player's defensive auto-resolve bonus should be decreased to 225% from 300% (or whatever the ratios are) to make up for this discrepancy in the AI's deliberately faulty calculation of the player's party strength when inside castles and towns. This shouldn't change the likelihood of enemy lords sieging when you're in a defensive settlement, it should only make it more balanced for the player.

You might then respond by saying that this would effectively nerf the player's auto-resolve capabilities and make it unfair for the player if the enemy is more than 2.25x stronger than you, and while you would be right, you would have to remember that the player has a MUCH greater unit retention and survival rate than the AI, and therefore has more high tier units than an AI would, so it would in effect be a handicap given to the player to make up for the fact that the player actually has a brain whereas the AI doesn't.

 

[Piconi]

You might then respond by saying that this would effectively nerf the player's auto-resolve capabilities and make it unfair for the player if the enemy is more than 2.25x stronger than you

Exactly what came to my mind the very first time I read the OP.
That being said, your feedback will be relayed to the relevant developers.