[jwallstone]

Thanks for the analysis, maltz. I have to disagree with you on one point though, but I agree with you on everything else. Your conclusion that adding 1 attack increases damage by 2% is a bit misleading. The 3.3% increase is always over base damage, and is always advertised as such, and in your example, it still holds true (+15 attack gives +49.5%, +16 attack gives 52.8%, a difference of 3.3%). If you expect attack to CONTINUALLY increase your damage by 3.3%, then you are expecting an EXPONENTIAL increase in power from attack. This would be easily unbalanced and might suggest a strategy of always picking attack to get exponential increases. This is definitely not what should be the case and is not what the designers wanted.

Note that the same holds true for leadership. It is NOT exponential. You get a set increase in army power, regardless of how much leadership you already have. According to your analysis, if you already have 10,000 leadership, increasing it by 1000 gives you a 10% increase in army power. However, after getting 11000, the next 1000 only helps you by ~9%. Therefore, according to this analysis, leadership is steadily devalued throughout the game.

I would argue that this is an incorrect way to think about it. You are MEANT to increase army power linearly, not exponentially, so that sort of thinking can lead to the wrong conclusion. Attack and leadership both scale linearly, the difference is that the same attack point chosen early has the same effect on the late game as one chosen later, but the leadership choice offered grows throughout the game.