Talk:Spring 2014 Event/@comment-60.50.242.46-20140501144645/@comment-24430017-20140502055655

If we are just looking for effect on armor, we don't need confidence intervals. Since we already know how the combat mechanics work, we can look at the range of damage to see if there are any differences.

For evasion, we can look at individual hit rates instead of accumulated damage. We can then compare to different formations to see if there is a statistically significant difference.

Considering how many variables, it would be easiest to test using pvp on a throwaway account. This would eliminate many of the problems like resources, fatigue, repairs. You could keep running sorties without stop. The difficulties would be finding a good opponent to test on (hopefully known stats/maxed and doesn't change fleet).