http://i.imgur.com/XeYpMAu.png http://champion.gg/champion/Aatrox/Top
Low sample sizes, but 38% winrate when you don't change from the old build and 50% winrate when you do.
Really makes you think.
Siegfriedx26 votesThe sample size is so small it's not even considered a sample. The practical use of that data is non existant
vOmvbrawfU2 votesExcept it is considered a sample.
Best Kog NA1 votesExcept it isn't a proper one.
vOmvbrawfU2 votesOk? How does that dispute my claim at all?
Best Kog NA1 votesLow sample = Unusable data = claim w/o statistical backing
vOmvbrawfU2 votesA) Low sample is still a sample. Which is what I said. You're saying it's a low sample, but then saying it's not a sample. Makes no sense at all. B) What power calculation did you use to determine if the sample size was adequate or not? None. You're using your opinion on it. Your opinion has no statistical backing :)
Keep trying if you'd like to.
Best Kog NA1 votesChampion.gg's sample size when OP was posting was 8000 (rounded to make math easier) games in platinum+ ranked games. Assuming that that is a valid sample size to extrapolate to Aatrox's true playrate / winrate we can move on to the builds. Using that sample as the population we have a population size of 8000. The amount of games played with that build is 51. That's less than a 10% confidence interval which requires 100 samples to be counted as a sample that can be extrapolated.
vOmvbrawfU1 votes[{quoted}](name=Best Kog NA,realm=NA,application-id=3ErqAdtq,discussion-id=JhEJA3LX,comment-id=0003000000000000000000000000,timestamp=2017-03-15T23:15:50.610+0000)
That's less than a 10% confidence interval which requires 100 samples to be counted as a sample that can be extrapolated.
Bro, that is not even close to the definition of a confidence interval lmfao. You literally have no knowledge of statistics whatosoever. Which is fine, just try to come across like you do.
Best Kog NA1 votesI'm just throwing words around that I got from Wikepedia. But I guarantee that you can't extrapolate to a whole population based on the results of one person playing multiple games with the same build.
vOmvbrawfU1 votesIn statistics, it takes smaller and smaller sample sizes to determine a statistical difference between events the greater the difference. There is no way to tell without running power calculations. There's over 300 million people in the United States. A recent study testing apixaban in severe renal disease patients had a total of 146 patients in it. And that study is probably going to lead to a change in clinical practice and allow physicians to start prescribing apixaban over warfarin in renal impairment.
Point being, off the top of your head, you literally have no idea whether or not the sample is able to be applied to the population.
Not to mention, this all has literally nothing to do with my original comment. He said a small sample is not even considered a sample. But, in fact, a small sample is without a doubt, a sample. Which leads me to re-ask my original question to you: Ok? How does that dispute my claim at all?
Sasogwa6 votesThe old build? Nobody builded that before.
nice table4 votesbuildered*
Meep Man1 votesbuilt*
The Letter H2 votesBuiltere'ditadive'
XmortoxX6 voteslol wtf is that first build.
I have never seen a single aatrox player build that
Papa Andrei6 votesIsn;t that the same site that says 3 rageblades and Botrk is best build for Varus also?
Spartan117103 votesi refuse to beleive only 250 games of aatrox have been played. That site is bs
Failog2 votesConsidering I have played literally nearly 10000 ranked matches in the last 2 years, it is BS.
Jorencice1 votesDude... Stop. Thats to many games. Do something else...
Its like an average of 8 hours a day...
warpenguin5559 votesikr. thats way too high
Quepha1 votesi refuse to beleive only 250 games of aatrox have been played. That site is bs
champion.gg takes a certain sample size of games at the start of the patch and runs an analysis on those games and displays their findings for the remainder of the patch.
also, the most popular and highest winrate builds do not make up EVERY SINGLE BUILD played by Aatroxes, there are many other builds that don't bubble to the very top in those categories.
Colgate Minuette1 votesi refuse to beleive only 250 games of aatrox have been played. That site is bs
It's only taking samples from ranked games platinum and up. I can kinda see how in NA only 250~ish people in platinum or higher chose to risk playing Aatrox in ranked after his rework.
PermabannedGaren3 votesChampion.gg just had a big update. The builds haven't quite caught up yet, and aren't accurate.
Troll for Trump2 votesWhat kind of weed are they smoking for 200 ppl to be building crit chance on Aatrox..
you don't change from the old build
Excluding the GA, the bottom one WAS the old build. That most played one is completely random.
Automated Riven1 votesyeah this is pretty much whats up.
SanKakU1 votesBoth of those builds look bad for both new and old Aatrox.
Verxint2 votesI've only played a few games of him, been too busy to spam him, but I've been going the second build except replace BorK with FroMal. It seems to work pretty well, just be sure to get your passive started in a fight.
My point is that usually people complaining that a champion's update made said champion worse than before are trying to play it exactly like pre-update and then are surprised when it doesn't work.
Failog2 votesFalse argument. Real aatrox mains seldom went glass cannon. Our builds were always fairly tanky. You have no idea what are you even talking about.
Apocalypsic1 votesFalse argument. Real aatrox mains seldom went glass cannon. Our builds were always fairly tanky. You have no idea what are you even talking about.
Exactly. I'm no main but I had enough experience on him to not be that stupid. That first build isn't even a viable Aatrox build.
My build always looked like Botrk, with Deaths Dance, and then full defensive itemization after that.
OurLestrade1 votesThey changed his skill costs to flat health instead of percent health for a reason, they're encouraging you to build some bulk rather than going full glass cannon.
Failog1 votes
Already had a 60% win rate with this build on the old Aatrox. CONSISTENTLY OVER HUNDREDS OF MATCHES.
Herim1 votes
Yay or nay?
Legacy Hydralpha1 votesThat's good as a bruiser/tank build, but I'd recommend trying to put 
in there. I think new Aatrox can use two builds effectively:
Bruiser/Tank, with items such as
Attack Speed/Carry, with items such as 
Preferably, I'd run either 
(or )(or ) or and a situational item.
Troll for Trump1 votesActually, with all his abilities being physical(besides his ult), Cleaver's slightly better. But it's probably better to get something that'd benefit his dueling potential, or improve his sustain/tankiness.
LadyRenly1 votesbork on aatrox lmfaooo
Only Play Darius1 votesWhy were people building him straight glass to begin with....
Sanctimonious1 votes imo (I might have the wrong hydra, but this is what i've been seeing in plat / diamond. It works pretty good.)
Borbland1 votesWeird, when I played old Aatrox, I always played the second build naturally and it worked pretty welll.
I cant imagine you used to build Aatrox somewhat like Tryndamere with pure AD.
Statchecky1 votesPersonally my go to build on old Aatrox was along these lines:
But itemization aside, the big difference between old and new Aatrox is that outside of blood rush or whatever its being called now you don't have a passive and this hurts his early game which was already not great.
