Private profiles might he common but it’s strange how all throughout ow2 they haven’t missed the mark even once.
Is it just the sheer amount of people who do have open profiles allows us to see trends?
It shows zarya meta, it even shows when they nerfed sigma and rein for “Over performing”.
It even showed genji having a good pick and win rate like blizzard said… So far everything has lined up exactly with what blizzard has said to us when they actually do give us data.
You have to actually mock sign in to allow them access to your data.
So the sheer amount of people is what allows it to be so accurate then…
Ya and they still get data wrong and if you never sign in they pull random data
But its not accurate when data is wrong its not accurate you would need a way to pull all the incorrect data out and only keep correct data its a tainted data straight up
Thats why I said I think it has plenty of data to see “Trends” say Reinhardt is popular even if like 60% people have private pfps the other 40% can see that theres a trend towards Reinhardts pick rate.
This leads to instancous where blizzard says “rein and sigma were overperforming” thi it’s not complete data overbuff actually does correlate to exactly what blizzard themselves have said.
If we get into “Specifics” that’s probably when overbuff stops being accurate, but when you stay with “trends” well that’s usually all the information you need especially when just the sheer velocity and amount of that trend basically show everything you’d need to know.
Do I need to know “Exactly” how many people are winning with and are playing Reinhardt? To the exact % Or do I simply need to see whats trending?
Have you ever heard of the law of large numbers?
In probability theory, the law of large numbers (LLN ) is a theorem that describes the result of performing the same experiment a large number of times. According to the law, the average of the results obtained from a large number of trials should be close to the expected value and tends to become closer to the expected value as more trials are performed.
Law of large numbers - Wikipedia
THERES ACTUALLY a therome for this!
Meaning that their sample size is even smaller than I thought it was, and even less worth taking seriously…
Ever taken a course on this kind of stuff?
You’d be surprised how small of a sample size you need to get a very accurate overall image with such sampling methods.
Not really, no. But the thing is, it’s not exactly the same as other things would be. The stats are gonna be less accurate than things not involving OW would be, because people who main certain heroes are more or less likely to hide their profile. That’s just one of the things that you have to account for with this.
For example, most Tracer mains have their account private, because if we don’t the enemy team will play Torb right off the bat if they check our profile. So her pickrate and winrate looks WAY lower than it is on there.
This is going to be true for… almost all data ever. Other than the precise telemetry than devs collect.
Even scientific research papers suffer from this flaw, which is very often why a study will present a “fact” and then another study will disagree later on.
You’re almost never going to have 100% accurate data. You just have to do your best to eliminate biases and then hope that what’s left gives you a large enough sample to draw conclusions about the population.
I don’t know if Overbuff is accurate. Ayanga is obviously claiming it is, but I’ve seen some of his other claims. All in all, though, it’s plausible that he’s right.
If every player - regardless of rank, role, heroes and casualness - had an equal chance of having a private profile and there was still a large sample size of public profiles left in the pool then there’s no reason why Overbuff wouldn’t still be accurste despite the existence of private profiles.
Look just because I said we should bring 5 man rez back that 1 time dosen’t mean my takes are bad you big meanie!
Sorry but- Ya, not really
Look- OB isn’t perfect. It’s even admittedly had bugs (which the guys behind it have acknowledged, usually with a warning on the site), but for most of its history OB has been impressively accurate.
There have been numerous times when the devs have released stats, and OB has pretty much always been right in line with it. 100% perfectly? No. But it’s been exceedingly rare that it’s ever been far off, either.
We’ve never seen the devs go “ya, this is the most picked hero” only for overbuff to be showing “nah, we only see like half that”- instead we typically see that hero on OB dominating the pick stats.
The same has been true for win rates.
You cant take something from overbuff and be like “this is 100% exactly what the actual stat is!”, but at the same time it’s been an impressively powerful tool to gauging trends.
I’ve seen you mention t500 is dominated by tracer? Go check OB right now and see what DPS is top of the pick charts lately for GM.
Again, You only need a very small % of people to share their information to get a very accurate picture- the things you’re insisting might throw it off, are so trivial that they really don’t. And again- we’ve seen this time and time again for literal years.
Yes i agree it does happen in all data but in that data when i work with it at my job i can curate and pull out the bad data. I cant do that with this plus i dont have all the data i dont know whats the persent of lucio games with reins to wins i blwould love that data but i dont even think blizzard uses that data based on patch notes.
Yes its why when people use overbuff as fact i cant take then seriously.
Yes and no depends on what your testing for but in the basic sense yes.
It has alot of incorrect data for example my data on it is incorrect same with multiple of my friends.
Private profiles arnt the problem here they add to it but they arnt the main problem the main problem is incorrect data and lots of it.
Refresh the page and don’t go to the console section that’s just a placeholder.
they dont have access to data on profiles that are set to private and as far as i know online data for peoples profiles is no longer available when ow2 came out
So yeaaahh remember that time period when overbuff was down… Thats what they were fixing.
Correct me if I’m wrong, but OW1 didn’t have private profiles, correct?
Overbuff is good at picking up general trends, you aren’t wrong about that. It’s not good at showing actual problems, or anything actually important. And that’s why I don’t take it seriously when people use it in debates.