I don't think the OP random number generator fix actually applies any interesting math to the random numbers. 11d10 would create a distribution range of 11-110, minus 10 would pull it back to 1-100 - I'm pretty sure that just rolling dice more times doesn't create a more averaged distribution. But I could be wrong here, basic level math only really. I am a programmer however :) What about instead: (RAND() + RAND()) / 2. Or 3 times and divide by 3, or however much you want to pull it towards the center of the 1-100 range. I think 2 times would have a rather large impact for selected scripts, if the UDK scripts can be modified like that (I'm not knowledgeable about it). Good suggestion otherwise. Reminds me of Sid Meier's story when he was focus-testing a Civ game, he found out that people would get pissed whenever their 10 attack vs. 1 def would lose, so to fix he implemented automatic win for large attack/def differences. It's only fair really.
