Auto-translated
The idea might be mathematically correct, and even (if it's developed and the details are clearly defined), it wouldn't be difficult to program.
But the idea is fundamentally flawed because what is easy and simple for the computer will drive a regular player crazy, who just wants to relax after a hard day at work. This will result in a different timer for each battle. People need simpler solutions, not to sit and guess - how many seconds the timer calculated for each turn of the battle this time, twenty-seven or forty-three, and why. And God forbid, it miscalculated and instead of forty-three seconds, it calculated thirty-nine, and because of this, I didn't have time to make a move. :(
This may seem cumbersome, but the point is that for the Lords, there will be a minimum timer for each template; in my example, this is 5 minutes and no less + the time for each round + emergency time, and for the Fortress, it will always be a minimum of 30-40 minutes or you need 2 hours to assess the situation to make the first move... I don't know.
The coefficient can be simplified even further - for example, if there is one unit of level 7, the coefficient is not 1, but 1.5 (one unit of level 7 = 3-4 units of level 6), the coefficient will increase if there are 5+ units of level 7 in the army, and the maximum coefficient will be 2-2.5, the largest coefficient should be assigned to parameters from 1 to 3(?), since losing all artifacts = gg in 99.99% of cases.
