If you look at the pick-rate of each hero for this month/week, it reflects the current meta perfectly. So why is it considered unreliable and not a useful tool when it is clearly showing current trends.
This is also evident whenever a rework for a hero is released, you see a massive spike for that heroes pick-rate as you would expect, showing that it is not completely unreliable.
because it doesnt show every single profile that played. Its not 100% accurate, but id only say its off by 2% give or take in terms of pickrates/winrates
That means it’s pretty frickin good then right??
yes, i find the overbuff not being reliable a laughing stock.
You don’t get it OP.
Overbuff is only unreliable when it goes against any narrative that I want to push. It magically becomes reliable when it’s applicable for whatever narrative I do want to push.
It’s only unreliable if you’re seeking perfection, otherwise there’s a large enough sample size for it to be a believable representation of everything.
The same thing that makes it unreliable is what makes it reflect the meta - selection bias.
Overbuff basically samples everyone who uses LFG instead of solo que in essence. So the Tank and Support rate on it is overrepresented to the actual pickrates of the 4-5 DPS comps common in Silver-Plat which have the most players.
I mean hey it has to be right it thinks I’m a top 10% rein player.
Wait a minute! It thinks I’m a bottom 50% [insert hero here]! It’s an unreliable piece of garbage.
Reeeeeee
This is how overbuff is used sadly. Although I agree with most of other people here that are saying it isn’t perfect but it’s good enough.
Overbuff being the only source we have makes it the only reliable source. Until there’s another way to see stats overbuff is basically law
That’s not how that works at all though? You can’t just make overbuff stats reflect what the most played comps is. That kind of happens naturally.
Stop creating new threads as a response, use the reply button.
It’s just a huge collection of Genji’s Mains reflecting of one another.
Are you kidding? That’s exactly how selection bias works. The group of people who opt in to public profiles overlaps extensively with those who follow the meta.
Question, since the meta is made by pro’s. And everyone seems to want to do what the pro’s do. If they jumped off a cliff would you do the same? Be human, be original, make the game fun not just a copy paste game.
The meta is only the meta because it’s the best team comp in current patch. People don’t just play hero’s because they’re at the top most played on overbuff lol.
Well, you can lead a horse to water, but you can’t make them drink.
It only doesn’t seem to matter when it doesn’t reflect what people want it to show… Like when I use it to statistically show things certain people don’t particularily enjoy.
It’s still accurate statistics that reflect very little bias.
Only people who have learned statistics in a classroom and never had to apply it IRL would throw out data like this, this is very very very useful information that can have great information mined out of it.
Because Overbuff doesn’t back up the current narrative I am trying to create.
its not realiable because people have profiles privated 
really tho, it just added a small margin of error
also it thinks im like a top tier qp dva