The AI cheating is breaking certain maps

[David]

I never quite noticed it until I put a high level AI on my team to follow its movements quite how much they cheat. Resources are one thing, that's fine, most AIs get a head start at high difficulties to be able to compete with players. The AI has the typical cheater issues where they start very strong/near impossible, but if you just hold out they lose steam and fall behind very quickly, which is normal for most strategy games I am fine with that, it's okay.

What is special is just how much the AI cheats against neutral camps. Their battles are legit absurd, like insanity level cheating, mathematically impossible fights that should just not happen and that can cause some issues. For example the AI showing up at the players doorstep extremely early on smaller maps, or maps where there's a really short route that's blocked by extremely tough camps.
A lot of maps are designed with tough encounters blocking paths and special resources that should only become accessible late in the game, however the AI can completely ignore these design choices and go through unimpeded. This also causes issues with the difficulty settings becoming unreliable depending on the map you choose.

Here's my test:
I used 'Overwhelming' AI difficulty for this test - The 2nd highest setting, because I noticed a sharper than usual jump between Challenging and Overwhelming, I wanted to see why the difference between these two settings is so high.

I raised initial camp strength to 300% with maximum growth to see just how much the AI cheats. The nearest camp was 51-99 sappers with 11-25 archers on turn 3, guarding an amber production structure - Very high value target blocked intentionally by a decently tough fight that I cranked to be 3 times as tough as normal.

The AI cleared it on turn 3. Its wielder was level 1 with 45 rats 1 necromancer and 5 oathbound. The AI not only cleared this camp, but did so with only 4 rats and 1 oathbound lost. That is mathematically impossible.
Those are ranged troops capable of shooting at the rats and killing at least 20 in a single turn, the idea that the AI lost 4 rats in the entire battle is insanity.
I can't explain how it would be possible for a level 1 wielder to win that fight let alone win with so few losses.
So I'm assuming the AI got some ridiculously heavy buffs against neutral camps specifically or the fight simulation is just really REALLY broken.
I did another test where the same camp was a Barya camp instead or Arleon, with steam pipers, veteran pikeneers and veteran musketeers and suspiciously the AI lost exactly the same amount of units. 4 rats and 1 oathbound. It appears that whether it's archers&sappers or pikes&muskets, the result stays the same.

Now, I'm sure the simulations are simplified to save time for the player, and kudos to the game devs bc turn ends take nearly no time at all to load, but something's not quite working as it should if the AI can bulldozer tough encounters meant to stop players from progressing without an effort. It essentially breaks multiple maps completely, making the AI difficulty settings vastly unreliable depending on what map you're playing on. Some maps you might select a difficulty you're comfortable with and breeze through the map no problems, other maps you're going to get attacked on turn 10 by an army with a lvl 15 wielder and 15k army value, because the only thing standing between you and the AI was a highly difficult to defeat neutral camp that the AI can clear with no losses. This also has the side-effect of giving the AI insane amounts of experience points, since they can take MUCH tougher fights much earlier without losing any units.

I did some further tests and it looks like under normal settings the AI is just not losing any units in any of the neutral camp fights. They can clear pretty much anything with anything without losing units, which is just not a good way to balance the higher difficulty AI.

Recommendation:
1) Try to make fights between the AI and neutral camps at least reasonably withing what's mathematically possible.
2) Scaling difficulty - Rather than all bonuses being applied from turn 1, the cheats the AI gets should increase with time. This way the AI would no longer run out of steam after their one big early game attack and can maintain pace with the player. Also this way they wouldn't be doing All-in strategies anymore that attack ridiculously early and lose them the game if you defend them.

[Mali]

I assume the AI also cheats with vision. I noticed it states they get extra vision on higher difficulties, however even on Worthy(where there should be no advantages) sometimes the enemy heroes just b-lines straight for one of my weaker heroes even though we shouldn't be able to see each other due to fog.

Kind of forces the player to not use any heroes in the campaign that causes a game over upon losing them. The ai seems to love running past everything if it somehow knows Stoutheart is in range and weaker. On her fourth campaign map on Overwhelming I had to opt to not use her entirely and sit her in the first town, as most of my game overs were the AI detecting she was in range through fog.

[Wirtzak]

Originally posted by David:

I never quite noticed it until I put a high level AI on my team to follow its movements quite how much they cheat. Resources are one thing, that's fine, most AIs get a head start at high difficulties to be able to compete with players. The AI has the typical cheater issues where they start very strong/near impossible, but if you just hold out they lose steam and fall behind very quickly, which is normal for most strategy games I am fine with that, it's okay.

What is special is just how much the AI cheats against neutral camps. Their battles are legit absurd, like insanity level cheating, mathematically impossible fights that should just not happen and that can cause some issues. For example the AI showing up at the players doorstep extremely early on smaller maps, or maps where there's a really short route that's blocked by extremely tough camps.
A lot of maps are designed with tough encounters blocking paths and special resources that should only become accessible late in the game, however the AI can completely ignore these design choices and go through unimpeded. This also causes issues with the difficulty settings becoming unreliable depending on the map you choose.

Here's my test:
I used 'Overwhelming' AI difficulty for this test - The 2nd highest setting, because I noticed a sharper than usual jump between Challenging and Overwhelming, I wanted to see why the difference between these two settings is so high.

I raised initial camp strength to 300% with maximum growth to see just how much the AI cheats. The nearest camp was 51-99 sappers with 11-25 archers on turn 3, guarding an amber production structure - Very high value target blocked intentionally by a decently tough fight that I cranked to be 3 times as tough as normal.

The AI cleared it on turn 3. Its wielder was level 1 with 45 rats 1 necromancer and 5 oathbound. The AI not only cleared this camp, but did so with only 4 rats and 1 oathbound lost. That is mathematically impossible.
Those are ranged troops capable of shooting at the rats and killing at least 20 in a single turn, the idea that the AI lost 4 rats in the entire battle is insanity.
I can't explain how it would be possible for a level 1 wielder to win that fight let alone win with so few losses.
So I'm assuming the AI got some ridiculously heavy buffs against neutral camps specifically or the fight simulation is just really REALLY broken.
Now, I'm sure the simulations are simplified to save time for the player, and kudos to the game devs bc turn ends take nearly no time at all to load, but something's not quite working as it should if the AI can bulldozer tough encounters meant to stop players from progressing without an effort. It essentially breaks multiple maps completely, making the AI difficulty settings vastly unreliable depending on what map you're playing on. Some maps you might select a difficulty you're comfortable with and breeze through the map no problems, other maps you're going to get attacked on turn 10 by an army with a lvl 15 wielder and 15k army value, because the only thing standing between you and the AI was a highly difficult to defeat neutral camp that the AI can clear with no losses. This also has the side-effect of giving the AI insane amounts of experience points, since they can take MUCH tougher fights much earlier without losing any units.

Hi! The AI's battle against neutrals is not resolved with the same quickbattles as your own battles. It would be too demanding and make you wait for the AI during it's turns much longer. That's why a different quicker method is used to resolve these types of battles. It should still give you similar outcomes to real battles, but what you describe seems like a true issue with this method.

Thanks for letting us know. Remember what the map was called that you played? Next time this happens to you, please report through the game's built in bug reporter (pres b+b+b). This lets us track down bugs much more efficiently.

[David]

Originally posted by Albin (Lavapotion):

Hi! The AI's battle against neutrals is not resolved with the same quickbattles as your own battles. It would be too demanding and make you wait for the AI during it's turns much longer. That's why a different quicker method is used to resolve these types of battles. It should still give you similar outcomes to real battles, but what you describe seems like a true issue with this method.

Thanks for letting us know. Remember what the map was called that you played? Next time this happens to you, please report through the game's built in bug reporter (pres b+b+b). This lets us track down bugs much more efficiently.

Thank you for your quite frankly ridiculously swift response! I did not think this was a bug but rather an intentional feature, which is why I did not wish to use the bug reporter.

I figured the simulation was different, because it is much much quicker and very smooth, again kudos to the devs that worked on it because it's very fast, but yes these issues appear on many of the smaller maps, the tests I've done were on 'Calm before the storm', but I've also noticed the same problems on 'Divided'. Both are small maps with quick paths between the two main bases blocked by tough neutral camps.

I'd also highly recommend at least trying out 'scaling difficulty' options because they seem very promising from other games. It makes the difficulty curve so much smoother to have the AI gain their cheats gradually every 5 or so turns rather than all at once at the start.

[Wirtzak]

Originally posted by Mali:

I assume the AI also cheats with vision. I noticed it states they get extra vision on higher difficulties, however even on Worthy(where there should be no advantages) sometimes the enemy heroes just b-lines straight for one of my weaker heroes even though we shouldn't be able to see each other due to fog.

Kind of forces the player to not use any heroes in the campaign that causes a game over upon losing them. The ai seems to love running past everything if it somehow knows Stoutheart is in range and weaker. On her fourth campaign map on Overwhelming I had to opt to not use her entirely and sit her in the first town, as most of my game overs were the AI detecting she was in range through fog.

The AI does not use the same fog of war as the player. They still have their own "fog" based on their exploration but it works differently, so you are right about the conditions not being exactly the same as for a human.

So the AI can be considered cheating, but to the AI, I am sure we are cheating ;). We are like trainable AIs with limitless potential to develop strategy. Jokes aside, we would want to make it truly equal in all ways, but the part with AI's vision has not been our priority. All I can say is, I hope this too has time and resources to be improved at some point.

[David]

Originally posted by Mali:

I assume the AI also cheats with vision. I noticed it states they get extra vision on higher difficulties, however even on Worthy(where there should be no advantages) sometimes the enemy heroes just b-lines straight for one of my weaker heroes even though we shouldn't be able to see each other due to fog.

I've noticed this too, but that seems pretty standard for most strategy games. It's very hard to code fog of war for the AI without breaking it completely.
Honestly I don't mind it that much. If you feel really annoyed by it, remember that this goes both ways, if you know they will attack your weak wielder, bait their army in and eliminate them with another wielder once they take the bait.

[Suicidal Maniac]

Add it to the list of why this game sux.

[Lakota]

The OP AI is the reason I don't play the game anymore

[LexGeneralis]

Originally posted by David:

I figured the simulation was different, because it is much much quicker and very smooth, again kudos to the devs that worked on it because it's very fast, but yes these issues appear on many of the smaller maps, the tests I've done were on 'Calm before the storm', but I've also noticed the same problems on 'Divided'. Both are small maps with quick paths between the two main bases blocked by tough neutral camps.

It actually happens on every map. I was thinking there's something off and was really curious about the mechanics behind it. Also I love to play with 300% mobs, but noticed a high difference between normal and worthy AI. At worthy, those fight begin to be absurd while normal AI can't beat them at all for a very long time.

If the wielder level would be higher with some tier three magic - ok, yes, doable. But you're absoulutey right that this can't be the case at the very low stages.

[Slowpork]

Yeah, digital bastards cheat like there is no tomorrow.

Fog of War? What's that?

Clearing neutrals and then rushing your main base the moment you step out? You can bet your ass they will do it. And then they will do it again.

Sniping your weaker heroes? Yup.

And the cherry on top, extreme levels of luck when it comes to Skill choice. I played Loth mission 4 on fair, it took me 3 ♥♥♥♥♥♥♥ armies to kill one army of Marsh Bulwark. Why? The bastard had nearly max melee/ranged/spell resistances. Good ♥♥♥♥♥♥♥ luck killing 3-4 such Wielders.

Basically, it's all about who can outcheese/outsmart the other "player". Rushing might work on smaller maps but on bigger ones I find sniping their cities, while 2nd Wielder hunkers down on your main base works great.

[Jean-Maurice Nya]

Still, AI is nothing compared to HoMM IV very difficult AI =^.^=

[['w']]

Thank you, saved me some money

[LexGeneralis]

Loth 4 is a story on its own, but the game itself is totally fine. However, as it is a strategy game, you need to think, yeah.

[Clown Reemus]

Originally posted by Jean-Maurice Nya:

Still, AI is nothing compared to HoMM IV very difficult AI =^.^=

I won't hide I am personally offended by any complaint against HoMM IV, but what you say is not anyhow an excuse for this game here featuring a lazy AI design, with zero personal attempts to make the AI both challenging and enjoyable to play against. They took the very primitive

IF > AI
THEN > GIVE MONEY

and called it a day. Provide me one reason why I should support this design philosophy today, after 30 years of solid strategy tradition with both ups and downs (downs being this design philosophy, obviously).

[bshock]

I don't enjoy these type AI's because it's more of a linear puzzle game than it is a strategy game. Heavily proactive and realizing what exactly the AI does every single game, often times each campaign mission having a near exact set of steps to defeat. It doesn't really lend itself to any sort of replay value.

[Ail]

Originally posted by Albin (Lavapotion):

So the AI can be considered cheating, but to the AI, I am sure we are cheating ;). We are like trainable AIs with limitless potential to develop strategy. Jokes aside, we would want to make it truly equal in all ways, but the part with AI's vision has not been our priority. All I can say is, I hope this too has time and resources to be improved at some point.

If, at any point you want some free help with that, I'm willing to work on the AI in my free-time.
I'm quite experienced in doing stuff like that. My latest project was an alternate AI for OpenXCom, which I also started out having to cheat but later replaced the cheating with memorizing where enemies were last spotted and sophisticated guesses where they could have gone in the meantime.

Here's my github of this project for reference: https://github.com/Xilmi/OpenXcom