[kukimuki]

Thanks a lot for the link to the real time battles discussion.
________
Disclaimer: I have nothing against randoms. Maybe all the suggestion below was caused by a few funny-looking exceptions I didn't know of.

Got the impression that sometimes morale slightly over 10 is something that makes the troops concentrate more on the dilemma of 'to rout or not to rout' than the battle. But sometimes they fight with unexpected bravery. Which makes me think that there might be too high chance of getting the extreme results with the routing formula. Maybe there could be some inertia in the function (like more consecutive bad events have more serious impact on routing).

Also, havn't had time to check it again lately, but don't units still have normal morale when you pick them up in the neighboring provinces after routing? Was a surprise at first, somehow I naturally expected that fleeing from battle would ruin morale.

So it makes me think that the morale stat you see in unit info in battle is some base morale of the unit, not the actual morale in battle. Maybe it would be nice to see more components of morale and when morale drops below blah-blah, the unit would be likely to flee. And 'likely' would be somewhat more predictable than atm, coupled by more predictable 'unlikely'. Meaning more chance to get the average result, less chance to get the extremes. Or if the unit is secretive about it's real morale response to demoralizing battle events, maybe there could be some comprehensive estimation based on battle events.

I hope the routing formula used atm is something different from 'given enough time any army will rout no matter what because of the encoded probability'.