Posts from Turn order.

Hello.
We are reading the "Guide for Heroes of Might and Magic IV" and trying to figure out the most pressing question for us.
We created a test battle (screenshot attached). The order in which our units move is clear. But it becomes ridiculous when the computer's turn begins. With a higher speed for the wolves, the fairies move first, and in subsequent attempts, the wolves move first, and so on. Or, for example, the morale drops for a wolf, but the fairy moves first.
What's the secret?

As stated in the preface, I didn't test everything myself – I accepted a lot on faith, unless it raised serious doubts. Unlike the third game, where each unit is clearly visible, including all its actions and capabilities, it's much harder to control in the fourth game (except perhaps by setting the enemy's speed to the minimum in battle, which I haven't tried). Especially when a general melee begins. I can't answer this question right away – I need to take a look, and perhaps clarify something. But I'm not surprised that the AI can act outside of its programmed algorithm. In the third game, for example, it repeatedly fled the battlefield on its own turn.

I tested the proposed option. In the first round, the turn order was correct (I played with two players, without AI). But starting from the second round, the turn order began to change slightly, according to an unclear algorithm (I didn't engage in battles, only maneuvers). For the Squires, the turn didn't start from 1st (although not from 6th or 7th), but more often from 3rd; the Wolf-Faeries had fewer glitches. Sometimes it was like this: the 6th unit was supposed to go, but the 3rd unit went first (it went first in the previous round), then the 6th unit with low morale, and then, as it should be, according to speed – Wolf-Faeries, Squires. After waiting, the order was reversed according to speed – Squires, Wolf-Faeries, and within each speed group, the order was the same as it was before the wait in that round. It seems that this violation does not depend on whether it is a human or AI player and is random. Perhaps it depends on how far a unit has moved from its starting position, but it is unlikely. And it is difficult to test this on a diagonally wavy field. This is like a distraction tactic for the AI – it is not certain that it will fall for it. It will only fall for it if it can immediately destroy the distraction. More often, it will pass by the distraction to attack the main unit. I have noticed similar violations of the turn order before, but I didn't recheck it and didn't pay much attention to it – I just moved and attacked. Sometimes, due to such a violation (not due to low morale), plans were disrupted, there was no one to counterattack, losses increased, and I had to find a way out of the situation, and in case of critical losses, I had to look for another tactical solution.

Which version of the game are you using?

I play (and test) on the War of the Whirlwind map. It unlocks almost all versions of the maps.

So, you simply created a battle in the editor that looks like the one in the screenshot above, and the turn order is disrupted?

Yes. But as I mentioned before, while playing different maps, I noticed that after the first round, the turn order could be disrupted – but I attributed this partly to my own inattentiveness, until I tested it more thoroughly now.

Yes. But as I mentioned before, while playing different maps, I noticed that after the first round, the turn order could be disrupted – but I attributed this partly to my own inattentiveness, until I tested it now.

Everyone knows that the main task of the AI is to gracefully lose to the player, but this doesn't always happen because it's impossible to program absolutely all behavioral nuances! And by the way, this bug (or maybe it's a feature) was partially fixed in Equilibris

It is absolutely impossible to program all the behavioral nuances of the creatures!

That is indeed true! I mentioned this somewhere before, and I will repeat it – it feels like the AI is also learning. When replaying a battle, it sometimes seems that the AI has taken its mistakes into account and is trying not to repeat them (just like a human). Of course, this can be attributed to randomness or a change in the AI's tactics in response to a change in the player's tactics, but sometimes the AI, when replayed, not only changes its tactics but plays smarter. I didn't notice this in the third game, but it is present in the fourth.:smile22:

That's absolutely true! I've mentioned this before, and I'll repeat it – it feels like the AI is also learning. When replaying a battle, it sometimes seems like the AI has learned from its mistakes and is trying not to repeat them (just like a human). Of course, this could be attributed to randomness or a change in the AI's tactics in response to a change in the player's tactics, but sometimes the AI doesn't just change its tactics when replaying; it plays smarter. I didn't notice this in the third game, but it's present in the fourth.:smile22:

It's very possible that such a mechanism is used in the fourth game; by the way, in many games, the AI does exactly that, i.e., it actually saves all previously played games (attempts)!

I remembered an interesting point regarding tactics against Chaos Dragons. On the ag.ru website, there were many descriptions of such a battle tactic against Chaos Dragons, which utilized their ability to destroy their own kin with fire breath. The descriptions were quite detailed and didn't raise any doubts, so I added them to the "23. Chaos Monsters in Battle" section of the Handbook. Although I didn't use these tactics much, I noticed that they were practically impossible to implement. I decided to test them and saw the following. Perhaps they could be used in the first version of the game, but they didn't work in "Warlords." First, the Chaos Dragons wouldn't position themselves in a way that allowed my unit to wedge itself between them and attack, causing them to hit their own kin with their fire. Second, the units suggested by players were not suitable for these purposes because their speed had to be HIGHER than the speed of the Chaos Dragons; otherwise, the Chaos Dragons would decide when and who to attack, not the other way around. This is not to mention the morale drop in the Chaos Dragons.
From this, I concluded that such tactics are of little real use, as they might work only by chance. Moreover, some players decided to move this from the realm of chance to the realm of skill and added a bit of fantasy for effect – an Ifrit will not move faster than a Chaos Dragon.