Auto-translated
Well, I've completed my experiments with griffins.
The most significant result is a bug that I think can be considered an exploit (although it's mainly applicable against the AI):
* Griffins belonging to the hero on the right, i.e., the defender, correctly attack creatures below them.
* Griffins belonging to the hero on the left, in all the experiments conducted, attack themselves and do not hit the creatures below them.
"Creatures below the griffins" refers to creatures that moved during the griffins' absence and were hit by the aerial strike in such a way that any of the cells they were standing on fell under the griffins' ability's area of effect.
The attached replays are named "right griffin or left griffin.sav."
A special gift for Hottabych: in one case, the griffins on the right managed to attack in such a way that the demon punished them twice with a spell. To achieve this:
* The demon attacks the knight.
* A spell is cast on the griffins of the hero on the right (i.e., the knight).
* A creature belonging to the demon is charmed.
* The charmed creature is monitored.
* The knight's griffins use the aerial strike on the cell where they are standing.
* The charmed creature moves to that cell.
* The knight's griffins attack, dealing damage to the charmed creature, and the demon makes two consecutive attacks on the griffins.
* The attached replays are named "for Hottabych.sav."
* In battle (not in the recording), there will be a lag in the animation, as if an invisible creature is making the first attack, but according to the log and damage, it was the demon.
* It doesn't depend on whether the knight's griffins are charmed or not.
* It won't work when attacking the knight's creatures.
A similar effect with the knight's monitoring is achieved as follows:
* The knight attacks the demon.
* The demon's griffins are charmed.
* A spell is cast on them.
* The knight monitors a creature belonging to him, as the charm has already been used.
* The demon's griffins attack, dealing damage to the monitored creature, and the knight makes two consecutive attacks on the griffins.
* The attached replays are named "for Hattabych 2.sav," although it seems to be corrupted... the monitoring is visible on two creatures...
* In battle (not in the recording), there will also be a lag in the animation.
* It doesn't matter if the demon has charmed the creature or not.
The file "tests of griffins 2.txt" contains a description of the experiments.
*The experiments with double charm were unsuccessful – at most, the color of the effect changes from red to green. I couldn't charm two creatures with one hero.
**The experiments involving the barbarian hero were conducted once for each situation, without any combinations.
P.S. Sorry for the inconsistent style of describing the experiments
