Connection ErrorBad RequestThis page went staleSession ExpiredSign in requiredForbiddenNot FoundToo Many RequestsServer ErrorBad GatewayService UnavailableGateway TimeoutHTTP Version Not Supported
Session Expired
Your session has expired. Please log in to continue.
Sign in required
You need an account to do that. Sign in or create one — it only takes a minute.
This page went stale
Your security token was rejected, usually because the page sat open too long or you signed in elsewhere. Nothing was saved. Reload the page and try again.
Request Failed
We encountered an issue while processing your request.
Please check your internet connection and try again.
Very interesting, tell me, was the initial idea behind the reaction to the wall of fire only to find a way around it? For example, if the same Moon attacks with a single unit and completely blocks it with a wall, making it impossible to approach without taking damage, I would defend and wait for the mana to run out instead of attacking. This is logical from the point of view of minimizing losses. If, for example, you have to attack directly, despite the increased damage, because otherwise you will take even more damage, then you can wait two turns. Or can't the AI think two moves ahead?
It's hard to say exactly what the original idea was, but I think the AI has poorly set priorities: the danger of the wall is considered on par with various indirect dangers (standing in a pile under a fireball or hydra, standing within range of an enemy stack), which are often ignored (I think, primarily with the goal of preventing weak AI stacks from running away from the player and instead being killed). At the same time, the damage from the wall of fire is the most direct, and it should be taken into account at least as much as the damage from a retaliatory strike. The AI cannot think two moves ahead (more precisely, in individual actions it may anticipate something, but in general - no), but it could treat the wall like a ditch: if it destroys a smaller percentage of HP than some fixed value, it should be ignored. There are, of course, differences, but in principle, the main mechanism for calculating the danger of fire walls has been created (the AI always assesses the damage it will take if it stands on a particular hex), but it does not always work. Dirty_Player
It seems to me that the AI's behavior is something that it shouldn't even try to do. It's designed so that it goes through walls - let it go, it runs gnomes after the serpent - let it run. The AI doesn't care. But with the berserker, it's a different story, it's a bug.
The situation is simply that the AI is a black box for players, so any of its shortcomings are easiest to attribute to "AI stupidity" (and in general, this is correct - they have no way of determining whether it is a bug or not), although in reality there are both obvious bugs and not so obvious ones, but still errors. With ordinary gameplay elements, everything is simpler - if it doesn't work as it seems it should, then it's a bug. In fact, the only thing I can say for sure is about the "developers' intention," and in this case, the AI with the wall of fire was probably intended to work in the same way as the berserker bug. This does not mean that if you fix one, you need to fix the other, it just means that one specific argument about the "developers' intention" is not correct here.