Every claim I could find that was actually backed by data supports this statement. It was a marginal difference though.

example: https://www.bluegartr.com/threads/10554 ... vs-Evasion

question: any of that testing on nasomi?

We can't legitimately parse here, which is the crowning roadblock to publicly accepted min/max numbers, without a load of theorycraft being tossed in to be contrary. Alot of people have made custom parsers, which is honestly elementary to write, and make it only read. I don't use that stuff myself, mainly cause I don't play to achieve some sense of personal accomplishment. I play because I like having a free/safe/fun hobby, that I enjoy with the people I enjoy playing it with. Parsing data and such would be a different hobby, I just "try" and if I notice results, I prefer the results that I notice personally, that apply to my own builds and jobs. Not really a recommendation for stat over skill, just what I would rather have.

_________________
!!! Surprise !!!
~not a noob~

Some of the hp latents ya gotta remember trigger from base hp not modhp. If the frenzy wallet works I'm sure was can fix any broken latent.

My methods weren't super precise, and I no longer have the data I wrote down, but I did some testing. Had 2 evasion skill merits, decided to see how effective evasion skill was. Tested this on lesser colibri since they hit fast and I could get lots of data quickly. First tested 200 hits wearing my normal colibri fightin' set. Maxed out my merits (another 6 points of skill) and did it again. Then I switched out my fowling/merman's earrings for novia/dodge earrings giving me a 14 point eva boost and did it again. No buffs minus shadows, no food. The 6 skill points gave me a 14% boost in evades, and the extra 14 eva only raised that to 16% (did the math to make sure I wasn't capped before anyone asks). So yeah, I'm pretty sure evasion skill wins, at least at 75.

Sounds like bad methodology. Lack of control for colibri level will result in wild fluctuations in your evasion from test to test. Also the sample size is too small considering the theoretical difference between skill and eva aren't huge. Simple test: someone with 200+ eva skill equip boxer's mantle, /checkparam, then equip rasetsu tekko and /checkparam. Could also try equipping Boxer's and Eva torque and see what the eva numbers are.

Edit: Did a quick test... Boxer's Mantle gives +9 eva, OHat gives +10. Use eva over an equal amount of eva skill.

Like I said. It was just a quick and dirty. But I think there's more that comes out of evasion skill than just your total eva number.