Maybe she's not as garbage as you think. Maybe the buffs were completely unnecessary. And don't give me this "low play rate" crap. Because there are other champs with low play rates (about the same as Akali numbers wise), but there win rates are very different.
Faie4 votesPeople have to stop using Challenger winrates to talk about champion balance. The number of Challenger games played is so low that the sample size is next-to useless.
Marshbouy4 votesYorick was played 1 game in challenger. That's a 100% winrate. he needs to be gutted.
G4LLOWSC4L1BR4TR2 votesAt the time of posting this, Urgot has a 63% winrate in challenger. Does he need to be gutted? Is he brokenly OP? Or perhaps is it more likely that small sample size can skew things wildly, both up or down?
Mr Carbon1 votesYou have to look at the sample size guy.... She was only played 36 times all of last month....
Cloud2731 votesIrelia was played only 37 times (same as Akali), but she only has a 51% win rate. Why is this? Because Irelia is clearly more balanced than Akali.
ZephyrDrake1 votesIrelia was played only 37 times (same as Akali), but she only has a 51% win rate. Why is this? Because Irelia is clearly more balanced than Akali.
or because one champion had a better streak of matches than the other. There are times where there are champions who have close to 100% winrate in challenger yet they are nowhere near that strong, people playing said champions just either are that good with the champion that they can overcome their weaknesses or just got a few lucky matches in which they got easy wins
AHeroNamedHawke2 votesIrelia was played only 37 times (same as Akali), but she only has a 51% win rate. Why is this? Because Irelia is clearly more balanced than Akali.
It's like an extremely tiny sample size will lead to wildy different results
Are you always this stupid or is it a special occasion
Cloud2731 votes[{quoted}](name=Charles Barklius,realm=NA,application-id=3ErqAdtq,discussion-id=9AeXo7Zl,comment-id=00010001,timestamp=2016-12-06T23:47:12.209+0000)
It's like an extremely tiny sample size will lead to wildy different results
Are you always this stupid or is it a special occasion
Sample size is completely irrelevant. You can have 2 champions with nearly identical play rates and their win rates can be very different. Clearly sample size is not an indication of champion health but rather an indication of popularity.
AHeroNamedHawke2 votesSample size is completely irrelevant. You can have 2 champions with nearly identical play rates and their play rates can be very different. Clearly sample size is not an indication of champion health but rather an indication of popularity.
You didn't answer my question
You clearly have no understanding of statistics, I just want to make sure I'm not making fun of someone that's actually handicapped
Cloud2731 votes[{quoted}](name=Charles Barklius,realm=NA,application-id=3ErqAdtq,discussion-id=9AeXo7Zl,comment-id=0001000100000000,timestamp=2016-12-06T23:52:17.837+0000)
You didn't answer my question
You clearly have no understanding of statistics, I just want to make sure I'm not making fun of someone that's actually handicapped
I understand perfectly. And my sample size is more than sufficient. Is it possible for you to have a conversation without flaming or are you always this rude?
APrinnyDood1 votes37 games is absolutely not sufficient for any kind of judgements about this game. You need a sample size of around 1-2K at least to get any kind of data that means shit.
There's a reason no one uses challenger winrates for anything. Hell, two patches ago Vayne had a 26% winrate in challenger. Should we be buffing her?
Neko Caitlyn1 votesYes plz 
AHeroNamedHawke1 votesI understand perfectly. And my sample size is more than sufficient. Is it possible for you to have a conversation without flaming or are you always this rude?
I'm retiring this year, know why? I found a super rare coin that lands on heads every single time I flip it. How do I know this? Well, let me explain.
I took a regular quarter and my super quarter, and a flipped them both twice. The regular quarter landed on tails once, and my super quarter landed on heads twice. From this I can deduce that my super quarter will land on heads literally every time it's flipped, and I will now start my gambling career around it.
^ That is what you just told me, with a straight face. Are you fucking with me dude?
The Sword Saint2 votes[{quoted}](name=Charles Barklius,realm=NA,application-id=3ErqAdtq,discussion-id=9AeXo7Zl,comment-id=000100010000000000000001,timestamp=2016-12-07T00:24:08.980+0000)
I'm retiring this year, know why? I found a super rare coin that lands on heads every single time I flip it. How do I know this? Well, let me explain.
I took a regular quarter and my super quarter, and a flipped them both twice. The regular quarter landed on tails once, and my super quarter landed on heads twice. From this I can deduce that my super quarter will land on heads literally every time it's flipped, and I will now start my gambling career around it.
^ That is what you just told me, with a straight face. Are you fucking with me dude?
What a hilariously accurate example of why sample sizes are important.
Sightless663 votesSample size is completely irrelevant.
You misunderstand the problem here. People aren't using sample size as a measure of champion health. People are using sample size as a way to say that the winrate stats aren't accurate, and they're absolutely correct.
A very basic rule of statistics is that whenever you take a very small sample of a population, you are opening up your results to a huge amount of random variance. If I'm only looking at 37 games, it's impossible to know if the results are due to champion strength, or just luck. It's very likely that in a large percentage of Irelia's 37 games, there just happened to be a feeder playing, while Akali's games didn't have any. If they had 1,000 games played, I couldn't make that claim because it would so improbable as to be redicious, but over 37 games, I absolutely can claim that this, or any other random factor, produced the results.
When you have more samples, random luck stops being such a large factor, because the probability of luck being a definitive factor in the results gets closer and closer to 0 as you increase the size of the sample.
Hell, if we were so inclined, we could mathematically show you how unreliable these results are.
Cloud2731 votesYou misunderstand the problem here. People aren't using sample size as a measure of champion health. People are using sample size as a way to say that the winrate stats aren't accurate, and they're absolutely correct.
A very basic rule of statistics is that whenever you take a very small sample of a population, you are opening up your results to a huge amount of random variance. If I'm only looking at 37 games, it's impossible to know if the results are due to champion strength, or just luck. It's very likely that in a large percentage of Irelia's 37 games, there just happened to be a feeder playing, while Akali's games didn't have any. If they had 1,000 games played, I couldn't make that claim because it would so improbable as to be redicious, but over 37 games, I absolutely can claim that this, or any other random factor, produced the results.
When you have more samples, random luck stops being such a large factor, because the probability of luck being a definitive factor in the results gets closer and closer to 0 as you increase the size of the sample.
Hell, if we were so inclined, we could mathematically show you how unreliable these results are.
I understand what you're saying, but if you look at the numbers for challenger there isn't a single champ with over 1000 games over the past month. And despite these low numbers, Riot is still able to assess champion health. Leblanc has only 258 games over the past month, but she has a 64% win rate. But clearly that was sufficient enough to warrant nerfs, or they would've just left her as is.
Sightless662 votesbut if you look at the numbers for challenger there isn't a single champ with over 1000 games over the past month
Exactly, and that's why we NEVER use challenger winrates. We can't start ignoring rules governing the interpretation of statistics just because we couldn't get a good enough sample. When confronted with sample sizes that are too low, our response should be to question the data, not to accept it anyway.
And despite these low numbers, Riot is still able to assess champion health
You are incorrectly assuming that Riot only uses challenger games to determine balance. They've explicitly said they use platinum+ data as the basis for most of their data-focused balancing (they also use player feedback and high-level player input to make these determinations). That's why champion.gg and op.gg and other sites began using plat+ data. If Riot didn't do that, they wouldn't have enough data to make decisions on.
Other people have already pointed out plenty of examples of champions who clearly don't have their balance represented well by these statistics. That should serve as a good proof of how terrible this data is.
Martensitic1 votesTFW you realize that this statistic is worthless, because its based on a tiiiiiiiny little n of total games played.
EndlessSorcerer1 votesAlas , poor Xerath with his 23% win-rate over 13 games in Challenger.
Clearly, he is in desperate need of buffs.
Nameless Voice1 votesYou think that's bad? Ivern has a 100% winrate for people who have over 125 games on him.
Astôlfo1 votesoh my fucking god
not this shitposting again
Kr1sys1 votesWhy is a player using small sample size data from a tier he/she will never get within a few tiers of playing to try and make a point about a champ that they're not capable of playing
BladeSteel1 votesziggs has a 70% winrate in challenger we better nerf him.