Is Overbuff reliable?

Its as reliable as your profile as seen from the forums. Because thats exactly where the data is scraped from for every single stats site.

Until blizzard has a true API available for the game, none of them can be considered anywhere near 100% accurate.

Take all the data with a huge grain of salt. Like a 20% margin of error grain of salt for your own personal stats, and even more for the “pick rates / win rates” screens.

Wha… how did you do that??

I went to your profile to check that it was public then I copied your profile name and number into the search bar for Overbuff.

I’m not 100% sure on how exactly Overbuff’s search functions. It’s possible that when it doesn’t find a result it checks to see if a profile with that name and number exists and if so makes the result and my typing out the whole thing triggered it. Otherwise, it might just be a typo/database issue (From what I can tell you were first scrapped back in March of 2018 https://www.overbuff.com/players/pc/Portamento-11974/activity ).

Actually, Overbuff uses around 900,000 accounts compared to the 40,000,000 accounts that Overwatch has (whether they’re active or not, Idk). In this case, Overbuff falls short of the 10% requirement to be valid.

But it is still the closest thing we have and I would assume isn’t insanely far off actual statistics.

It’s honestly the only reliable sources that we have currently as of right now, however, it lost a lot of its creditability when private accounts was pushed to LIVE.

Oh, okay, I got it now. Thanks.

Not a plug, but Omnic Meta actually gets most of your stats correct.

Our population is approx. 300,000. So they only needed 400/300,000.

This is how statistics works, interestingly. Through formulas, they “generalize” conditions with apparently unrepresentative proportion.

I’m genuinely curious - how do you know this? What are you basing this claim off of? Could you give an example of what kind of data you find highly inaccurate and why?

There are really good statistical reasons to believe that Overbuff is accurate when you look at its data properly.

It’s reliable enough. I haven’t seen a single thing so far that would go against what is displayed there.

It’s reliable for silver through diamond, where the majority of the population plays. The closer you get to bronze and masters, the fewer people play,which means fewer potential accounts hooked up to Overbuff, but this all on a day to day basis.

Even if the exact percentage isn’t reliable, Overbuff can still show trends through the weeks and months, for any rank.

miss masteroverwatch it was great site too

It’s more reliable than an opinion of a player, that’s for certain.

1 Like

He’s never provided any kind of response other than “private profiles” to this argument, just FYI. I think he mostly just opposes it because people have used Overbuff data to argue in favor of the Torbjorn rework (which he opposes) and have used it in the past to argue in favor of another Mercy rework (which he also opposes)

1 Like

I found that even before the private profile was a thing, overbuff is kind of giving you a wrong impression of how you perform.

I know a lot of people has their main heroes ranked in top 1000 out of 300000 players, but most often than not they couldn’t perform against even diamond + players

That’s because the ranking system factors in playtime (or matches played, as it were), not just your raw stats. Basically, it tries to eliminate randomness in the system by weighting your raw stats by your matches played, meaning that someone who plays at god tier for only a couple of matches doesn’t outrank someone who has 1000 pretty solid matches on the same hero. It should try to normalize for playtime, and I think it does, but for accounts with small amounts of playtime on them, it’s tough to do.

It is not.
-Most ppl play on smurfs and those smurfs have locked info. Majority of the population has at least 1-2 smurfs
-90% of the lower elos have their profiles locked

So if you want to get pickrates and winrates for the throw picks like bastion, torb or overtunned heroes like tracer and genji, they will be pretty inaccurate.

1 Like

The funny thing is at least overbuff is based upon game data and not random BS from players that haven’t sampled large swaths of data themselves.

That is like saying the sea doesn’t exists cuz you can’t see it because of the mountain is blocking your view.

This is wrong btw, since Overbuff always has just used data of people who searched their own profile on the site. Private profiles changed very little.