I look at winrate first and pickrate second
If a player doesn’t win alot with the heroes they pick…
They are unlikely to play again.
4 Likes
Reaper: This is my curse 🥲 
1 Like
Between the 2 that should be the more important one…
But if you go by that you’d get the impression that the game is actually pretty balanced…and we can’t have that
Instead of “people pick some characters more than other despite similar performance rates” it’s “people pick a few characters more because they’re better…but we can’t corroborate that with any numbers…here are the win rates that don’t really prove any of what I just said though”
It’s aggravating really
2 Likes
I just looking at the ones near the bottom
And notice each has issues still with their role mechanics.
That Blizz still has never fixed.
1 Like
I mean spammy spammy do be going spammy spammy
Don’t forget the OP is actually not using raw data but “normalized” data for pick rates to give the appearance of a particular hero overperforming. Similarly, not doing the same for win rates while even putting a line at 50% to give the appearance that is the average and said hero is overperforming. It’s deceptive practice to try to push a certain narrative.
4 Likes
I dont think any defense is needed.
Someone has to be the “king” of the hitscan category and that dude will have a stupid pickrate. If you nerf Ashe and Widow hard enough, Mcree becomes the popular kid.
If you nerf Mcree, Soldier will become the popular kid and so on.
Thats what it is really.
PS: Yeah, no FR. For those who think I am defending Mcree, I am not. But believing a hitscan is not going to be ALWAYS on the Top3 in DPS is being delusional.
OP should make it clear that, from the Grand Master icon in the top right, this represents only GM rank, so top 1% of players.
There’s a range of opinions on the merits of balancing a game with 60m players based on such an elite (and small) data pool.
2 Likes
I like how Hog has one of the lowest win-rates in the tank category and people still complain about his “one shot” and how he’s “OP”.
Ok, I’ll try:
Why is his winrate so low compared to many other’s?
I know, it’s not a defense. It’s genuinely a question.
You can’t Normalize winrate like you can pickrate, normalizing pickrate makes sense as you have 3 roles who each have 2 places in a team, while not having the same amount of heroes to pick from, which means that the “raw” data is just going to show you that damage heroes are being barely picked, as you have more choice to pick from, yes, it is not perfect, but it is the best you can really do with the data given.
Winrate on the other hand has nothing to do with the hero pool, it’s merely the percentage of wins a certain hero has gotten, you can’t normalize something if there is no difference in the scale or if the scale is not relevant to the subject, the line at 50% is just to make it easier to follow and easier to compare visually and it is the theoretical average, but because things like lost data (because of private profiles) and the fact that a GM can play against a player who is master the average is more towards 53%-52% in GM.
There is no narrative that is being pushed, this is just data, you may like it or you may hate it, but that doesn’t make it invalid nor deceptive, I provide a much easier to visualize and compare data so people could come to their own conclusion about the state of the game, nothing here is skewed or biased against someone, it’s the same way I have done it for months, the meta has changed a couple of times, the data has changed, yet my way of calculating and presenting the data hasn’t.
8 Likes
Because of his pickrate… There’s likely to be one mccree on each side, so one of them loses and one of them wins. They get a 50% winrate on average.
5 Likes
How many players do you estimate your analysis represents?
The data is provided from Overbuff, it’s unknown how many players this represents as it is not something that is out for the public, but this is based on every player who is in GM and has a public profile, the data was collected over a month from GM.
Hero Statistics - Overbuff - Overwatch 2 Statistics
I know – I’m wondering how large a sample you estimate that is.
Back last August, Hog reached pick rates that suggested he was in every GM game, and yet was able to get a 55% or higher win rate; presumably this is because GMs don’t just play against other GMs, so it’s not inevitable that for one GM Hog to win, one GM Hog must lose. Plus the possibility of one tank switching off Hog and then the other tank choosing him.
Brigitte
Torbjorn
Reaper
These are the only three with less than 50% winrate, and all have less than 1.3% pickrate (just rounding 1.31 down)
Reaper’s numbers are prob so low bc he doesn’t have many good targets to farm in current meta. Torb…well idk whats going on with him.
But this should be definitive proof Brig needs help. Buffs or reworks or nerf-reverts.
People can’t just keep going 
“Lalala I don’t see anything Brig is still really strong in GM reee”
Brig is 2nd lowest picked in GM and master, and LOWEST in Diamond, Plat, Gold, Silver, and Bronze. And as far as winrate, she’s lowest in GM, Master, Diamond
2 Likes
We can’t know for sure…
At the end of the October last year Blizzard claimed that Overwatch had 10M active players during the month, we can guess that the number has went down since then, so lets say 7M last month, it is estimated that 1% of the playerbase is GM, which means around 70,000 GM players had played the game during the month, some may have not played comp , so that leaves us with around 50,000, then lets say that 40% has a public profile, my guess is that it is around 20,000 players but it is probably less as I have given very rough estimates, but there are a lot of games which are being played each day so even if you have a small sample size you still get respectable data.
1 Like
0.61% and 49.59% for Reaper… god, he needs help.
1 Like
Sad Torb noises…