Thanks guys
I've done a bit more testing on this today and mostly found it needs more testing and that trawling through the auto result stats quite boring. Anyway here are some stats for 500 dodge which I am guessing is 50% dodge.
All tests were against the same group of spiders against one peasant with no armour,a club and a massive amount of HP. I only counted the attacks made by the spiders against the peasant as some of them had defense.
Test Run 1:
Hits: 39
Misses: 37
Test Run 2:
Hits: 46
Misses: 50
Test Run 3:
Hits: 45
Misses: 66
Test Run 4:
Hits: 55
Misses: 55
As you can see three of the results show that the outcome is about 50/50 however test 3 is totally out of whack with this. Which just goes to show that a small selection of results isn't enough to get a proper representation. However I am willing to bet that test 3 is an aberration and that the 500 dodge is actually 50% dodge. The next step I guess is checking to make sure that the dodge chance increases in a linear fashion, which it more than likely will.