Anwig Salade WSACC test

[Xanthe]

Agreed. It is a small sample, but a ~7% hit rate increase does translate to ~15 accuracy. Thanks for testing.

[pchan]

what club did you use ?

[RyaWHM]

Oh, right, some helpful details:

• KC for 7% TP/swing
• 0% double/triple attack
• low haste (18% total) made it easier to separate melee-/ws-/mob-generated TP
• turned away before potion wore; paused after renewing before turning back

EDIT: typo!

[pchan]

Overall (combined)
818/2294 melee hits
= 35.658% melee acc

143/335 subsequent (non-1st) WS hits
= 42.678% WS acc

==========

Conclusions
... too small sample size, is what it looks like to me.


EDIT: fixed typos/colors; added "(2x)" for clarification that there were 6 crabs in total

Well ignoring the fact that you are mixing crabs of different levels, you get an error of +/- 5% (1.96*sqrt(0.43*(1-0.43))/sqrt(335))
on the wsacc of the last 5 hits, and +/-2% for melee hits. So it doesn't rule out anything. Also you could very well imagine that the accuracy of the WS hits (withouthead gear) is naturally superior or inferior to melee accuracy (penalties ?). So I think it would be better to compare WS acc with no head to WS acc with head.

[CDF]

If you pool all the data, yes, this is enough to show that the weapon skill accuracy +15 affects subsequent hits.

If you're anal and exclude the level 60 robber crab data just because of putatively lower AGI (I heard it's 39 as opposed to 42 for 61 and above), then it's borderline. In either case, you have "good enough" sample sizes anyway even if they're unequal.