The State of Assassination

[Deleted]

OK, this has jackshit to do with what I said. What I said, and explained very well actualy, is that Sub and Outlaw rigth now does not have a big enought sample size to come to any conclusions in regards to its actual performance.

You are plain wrong though. 90,95,99% DO have big enough sample size. The possible error due to difference in simple sizes is marginal.

[Cyanix]

I do want to point out something tho; You should not look at boss damage on fights like Chrono and see yourself X% above the next guy and think to yourself "This is why they bring me, this is my niche". Plenty of classes would have around that % (or more) if they were told to sit on the boss, look at starbro dmg to boss logs for example. Your RL lets you stay on boss for the various fights not because you are so amazing at it, but because you are so absolutely bad at anything else from a dps standpoint. This is what you should be concerned about. If Sin is going to be the freaking worst by miles at constant target swaps and quick burst on priority adds then it needs to be equally that much better at sitting on the boss imo.

This, exactly.

Needs to be uploaded on Reddit, because devs dont give a **** about what's happening on the Bliz forums.

[DakonBlackblade]

0

You are plain wrong though. 90,95,99% DO have big enough sample size. The possible error due to difference in simple sizes is marginal.

You ppl rely don't know statistics, its hard to discuss. The 90% + brackets are a factor of the whole data and it is based on that, the whole data for Krosus for example (which is one of the bosses with the largest amount of logs) is not even 1000 parsees for Sub and Uutlaw (its not even 600 for Sub actualy), that is nowhere near a large enought sample size when you want to compare it with Sin wich has over 10k entries for that boss. On Algur Sub has less than 200 entries. And from that point its all rather bad actualy. If we go for something like Guldan, which is a boss with very low number of kills for whatever spec, we see some very wierd crap going on across the board.

And stop using the 90%+ as a measura of anything, 90% of the ppl playing the spec isn't doing those numbers, those brackets are only usefull for evaluating maximum potential damage, you have to look at somehting more on the middleground to try and eliminate luck factor from the analysis, at least a bit.

[Deleted]

You ppl rely don't know statistics, its hard to discuss. The 90% + brackets are a factor of the whole data and it is based on that, the whole data for Krosus for example (which is one of the bosses with the largest amount of logs) is not even 1000 parsees for Sub and Uutlaw (its not even 600 for Sub actualy),

I ll make it simpler for you: so what you are saying is that if i have 10 000 wolves and 500 bears i can't compare say average meat consumption by percentile of bears vs wolves because there are not enough bears in the sample size? Are you serious?

And stop using the 90%+ as a measure of anything, 90% of the ppl playing the spec isn't doing those numbers, those brackets are only usefull for evaluating maximum potential damage

Let's pretend you meant 10%. Anyway... No, its not 10% of the people who do those numbers it's the top 10% of parses that have this number as an average.
So if outlaw rogue sample size is 1000, the average for top 10% is X dps
If assa rogue sample size is 10 000 the average for top 10% is Y dps

I could agree that top 1% (99% percentile) might be pure outliers but certainly not top 10%.
Let's also not forget that people killing mythic Krosus are not really random sh%ters — those are all decent players, so it's not like numbers are skewed by bad players.

It is what it is — the average for top 10% is X. And X for Assa is below other rogue specs. Because with best gear and in hands of best players other specs perform better.