You could mod the Research action in EaMain/EaActions.lua. Look for:
Code:
TestTarget[GameInfoTypes.EA_ACTION_RESEARCH] = function()
and change equation for g_int1.
I've been thinking as a general balance issue that all of the "yield actions" need reduction. Either that or all of the GP mods need reduction. I'll probably do one of these in v8.
So we are talking about these two lines:
1. g_int1 = floor(g_mod * scienceModifier / 200 + 0.5)
2. g_int1 = floor(g_mod / 4) + 1
Sorry, I cannot see the connection between great persons level of scholarship and the beaker they add per turn.
Also, I agree with the 'all gp mods' need reducing (those warrior can add a huge amount of xp per turn when levelled up), but I do need to remember that in my 6000 turn games these 'problems' are much more pronounce then with normal games where man GP are dying in eighty turns, making the above almost irrelevant to them. Perhaps immortal/living need a different mechanics???? Or better yet, a scaled mechanic for all, based on the law of diminishing returns.
That is new, early learning comes in great chunks (think new born baby), while new late learning comes in much small increments, build upon the earlier learning (think mature age student trying to go back to university).
Now to some degree the experience needed for the next level of scholarship/training etc does already scale, but the exponential scale of it is out of wack with the benefits it provides (seeing that it is a shared mechanic with normal unit levelling--no surprise there). So IMHO, either the rate of level gain should be lower, or the bonus each level provides should be. And in either case, to be perfect (again IMHO) lower here actual means scaled better.