The Spirit Realm

more discussion about it
http://www.eqoutrider.net/forum/viewtop ... sc&start=0

If you have a close look at mojo's numbers, theres some quite striking errors even up in the 35/36/37 delay area, so you might be surprised at what the situation is at standard 2hander delays.
If anyone has a knight twink with no double atk aa they could continue mojo's test up into the higher delays, just need a bard willing to group with and go afk to provide the different overhaste values. I guess what will interest the paladins most of all is the situation at 40 delay (and possibly compared to 37 to have another sk vs. pal dps argument :p)

Based on the eqoutriders data I looked at our Epic (40 delay) with different hastes (using common overhaste values from bard, OMM, quarm etc...):

at 100% ratio is 102/20.3125 = 5.02
at 105% ratio is 102/19.53125= 5.22 (4% ratio improvement over 100% haste from 5% overhaste = good)
at 108% ratio is 102/19.53125= 5.22 (same as 5% overhaste)
at 110% ratio is 102/18.75 = 5.44 (8.3% in ratio)
at 125% ratio is 102/17.96875= 5.68 (13% ratio improvement over 100% haste = good : 2.25/2 is 12.5%)
at 70% (Valor of Marr+45% haste item) ratio is 102/23.4375 = 4.352 (87% of the 100% haste ratio)

To compute your effective ratios for your epics just add your damage augs to the 102. The root of the issue is that it seems the delay increments that are available are in multiples of 0.78125 and that the delays are rounded.

Overall it makes sense for us to get an overhaste besides that the 8% ones do not seem worth it if you already have a 5% one. We are less sensitive than other melees because we use relatively slow weapons, but we are still sensitive.

[Edited a few errors]

I used standard rounding here, and from what the rangers and the monks seemed to say it wasnt a ceiling function. I'll update the numbers with the 3 different rounding options.