Inquest

[Medasin]

<p>Does inquest proc off melee auto attacks?</p>

[Medasin]

<p>NVM, finally had it proc off ann auto attack, must have the worst luck ever. 400 auto attacks and not a single proc.</p><p>It does seem to me that it doesn't proc its stated 10% though, closer to 5%, out of a total of ~20000 attacks that should be counted, it only proced 1032 times, far less then the ~2000 that it should be at 10%.</p>

[Avirodar]

<p>Just making sure you were not factoring in double attack hits?</p>

[Medasin]

<p>So procs only happen for the first double attack hit?</p><p>Edit: Well, whether or not they don't, I did a reparse, another 10000 attacks, removing double attack hits, Only managed 600 Inquest procs.  I seriously think something is horribly wrong.  I am no statistician but 6% on 10000 seems to be statistically significant, as being broken.  BTW, this was nothing but autoattacks on a training dummy, making sure power was empty before starting the parse.  Sorry didn't think to take a screenie of the parse to post.</p>

[Lord Hackenslash]

<p>What weapon were you using?</p><p>The proc % might be normalized so if you use a delay less than 3 it might affect it's proc rate negatively. likewise a delay above 3 might make it appear to proc more.</p>

[Avirodar]

<p>My time is limited, so I can not do an extensive test.  However, this is what I done over a 4 minute parse on a training dummy.Total melee swings : 246    (includes DA+Flurries)Default attack swings : 100Double attack swings : 97Flurry attack swings : 49    (triggered 17 times for 49 swings)My hit ratio was 88% as a whole, but of my 100 default auto attacks (non-DA/Flurry), 85 hit.... So regarding attacks that can proc Inquest, the count is 85.This quick 4 minute parse is not definitive proof (a 30+ minute parse is more desirable), but with a 5 second delay weapon, and a haste mod of 156, from 85 successful auto attacks I got the proc 8 times, which is close to a 9.5% trigger rate. From my initial glance, a variance of half a percent is well within any margin influenced by chance and luck. The result I got, was actually much closer to the stated proc chance than I was expecting. I would not be suprised if I ran some more tests, to have some results come back at 7%, and other parses come back saying 13% (without any proc mod effects), due to what I consider to be normal ranges for variance.</p>