A program for determining luck/morality in Heroes 3!

It has been noticed that when repeating certain battles, round after round, everything repeats exactly – the morale and luck rolls, the damage dealt, the spells cast by the spellcasting units, etc.
This was observed in the griffin conservatory. However, in the battle with the titans, the djinns cast different spells already in the 1st round.
I came to the conclusion that in internal structures such as conservatories, crypts, naga banks, utopias, etc., the battle map (let's call the entire battle process, step by step, a map) remains unchanged when repeating the same actions.

Here, many people talk about generating morale using delays. If this is still news to someone, I will say: there is an authoritative opinion that morale (or luck?) can be influenced by performing certain actions in battle after a specific time interval. For example, you have 2 stacks of snipers in your army, 1 + 10 units each. So, sometimes during a battle, morale is rolled for the 10 snipers, and sometimes it isn't. The question is: what could this depend on. So, as I said, there is an authoritative opinion from more than one player that the morale roll in such cases depends on the time delay between the turn passing to the 10 snipers and the moment of attack (i.e., the shot). But it is still unclear what these delays depend on and what they are equal to in each specific case.

My opinion is this: in battles in internal structures such as the griffin conservatory, the situation described above is impossible. That is, if repeating the same steps causes morale to be rolled for a sniper, it will always be rolled for him when repeating these same steps, regardless of any unclear delays or anything else.

However, regarding battles in open areas (for example, the battle with the titans), the assumption about delays undoubtedly has some basis. I will say right away that, to date, I have not used this hypothesis at all when participating in tournaments. In particular, I conducted the battle with the titans with the same assumption as for internal structures. Sometimes it somehow works (i.e., everything repeats from battle to battle if you repeat the same steps), and sometimes there are several options. But the question is quite interesting, and I would like to study and discuss it. My personal opinion regarding whether using and selecting delays is a fair method: although I have not used it (yet), I believe that there is nothing wrong with it.