oh hey they actually gave us numbers for once…time to crosscheck with overbuff (hardly ever get a chance to actually compare)
lets see…over past month…GM…
soldier - 3rd amongst dps (4th overall)
genji - 9th (16th overall)
make of it what you will…i was just curious since we can never really do this
Those numbers are worthless on their own. Sym before the rework had close to 70% Winrate.
This also says a lot about Overbuff and other sites that tries to measure OW data and why people should NOT trust them.
In Overbuff GM Win rate for Soldier: 76 is #4 and Genji is #16. 
.
I wonder what their play time stats look like. I can come out of spawn as soldier for 1 push, swap Hanzo and win the game, have a 100% soldier winrate and still not make a case about why he’s a good hero.
Granted that’s obviously exaggerated, but a lotta 76 players I see seem to swap off quickly.
zarya 60, Ashe/sombra at 58 (1,2 & 3)…time to roll out the nerfs!!! 
Bahahahahahahah, me watching people discredit numbers when the whole forum used stats to crucify brig.
Oh now they don’t matter.
Soldier is the dps learning hero he is supposed to be basic compared to ashe.
Until blizzard releases an official feature to check stats this is the best we can get
I can’t heal through stupid ~WildPants
Keep in mind the dev post only ranked them in terms of the damage category not overall when you do that Genji is a lot closer than 16 to his actual number additionally GM is also the smallest pool and such inaccuracies can be expected with how small the sample size is but the overall numbers and the specifics for the larger ranks should still be for the most part accurate
Ok, if people want to use wrong data, that actually leads to horrible wrong resolutions and correlation discovery… But all the nerfs/buffs posts based in these proven-wrong data cannot be considered more than trolling.
Yeah, they cherry picked two numbers without giving us the relevant context. On it’s own like that the numbers mean nothing. 
If they want to discuss balance based on numbers they should actually, you know, give us the numbers to look at. Or at the very least not talking win-rate without considering pick rate…
There only problem with this comment is that we don’t know if they’re talking about win rates from the past week, past month, from the time of most recent patch, if he pulled from GM alone (since he brings up balancing from the highest end), etc.
Overbuff only really has past week and month stats from the moment you’re looking at the numbers, which constantly change when they update daily. It doesn’t mean it is wildly inaccurate.
Um, they’re discussing why Overbuff is less accurate than Blizzard’s numbers, which is a known problem. Brigitte stats have been consistently stellar since her rework despite what all the people who miss being a raid boss say; there’s very little reason to think those stats aren’t roughly correct.
It was pretty consistently around 60% for sym 2.0 on overbuff. I don’t know if those numbers were off though in game I had a 70% winrate for sym I was gm with her. She just had really good stats for the few people that played her
Yes that makes sense for sure. But I still think Overbuff is overrated. They don’t have the complete data that Blizzard have. Private profile is something already big to discredit it. That’s why I’m skeptical when it comes to people using Overbuff numbers.
But did he used overbuff as his source?
Don’t they also collect data themselves?
If so, does the data from overbuff match their own data?
Since everyone and their grandma has a private profile overbuff is useless. Surprised that some people still think the data is anywhere near accurate or usable.
This is what happens when they refuse to give us official numbers. People want data and right now the only way to get it is third party sites.
If Blizzard didn’t hide their internal numbers people wouldn’t have to use less precises data from others
Depends on your timeframe and context; month vs week timeframe and overall vs in-role changes the math dramatically, not to mention the mountains of flaws in the app, such as Sym still being in the Support role and the separation of the Offense and Defense categories
overbuff can only gather info on public profiles , while Blizzard’s stats base on all profiles , thats why there is a certain different between the two
Also Overbuff winrate is weird.
It’s skewed by Attack biased, Defense biased, or Mirror match biased. Devs winrate probably isn’t skewed.