Is the win rate for a character determined if you were playing the character when the Victory screen comes up? Is it if you played the character during the round? What if I die immediately before the end of the game, change characters, but don’t make it out of spawn?
As far as I know it’s based on time played. I’m not 100% sure on the exact formula, but I imagine its something like if I was on a hero for 5 minutes in a winning game and 10 minutes in a losing game then my win rate on that hero would be 33%. because 5 winning minutes / 15 total minutes is 33%.
EDIT: I realize that what I said above means that longer games count more towards win rate so to prevent that I could also see them doing it as a percentage of the game rather than in terms of total play time because a 10 long game should have the same scale as a 20 minute long game,
Thanks - having a fraction of a win (or loss) for each character seems consistent with the other ways in which they track data. I still wish they would present a normalized set of numbers (i.e., if they are going to provide benchmarks based on a per 10 minute average, the stats for that round should be based on a per 10 minute average to provide an apples to apples comparison).
Games won are weighted by% of time played that game.
So in a 30 min game if you played 76 for 20 mins and Cree for 10 mins
76 would get weighted 0.67 of a win and Cree 0.33
If this was there only game they would still be at 100% winrate cause the formula is number of games won over number of games played
So 76 will be 0.67/0.67 which equals 1 which is 100%
If you then had a 10 min game you lost and only played 76 he’d stay at 0.67 games won and go to 1.67 games played
Meaning his winrate would be 0.67/1.67 = 0.4 = 40% winrate.
The tldr version wins and lose credit is based off percentage of time played in a map. The formula for winrate is percentage of games won over percentage of games played. Every game is worth 100% so length of games don’t affect it.
1 Like
So what this actually tells you is that only one-tricks really can have a true idea of their actual win rate with a hero, or people who pick and then don’t swap unless under extreme circumstances.
On the bright side that means my Pharah win rate is accurate so I like that 
Pretty much. It’s also why niche heroes have high winrates like sym cause when it works you stay on the hero for the rest of the game, when it doesn’t they switch off in the first 10% of the game.
Winrate as a metric for skill or how good you are at a hero is flawed anyway. SR essentially is your winrate relative to your skill.
Sym example shows that but there are also bad game heroes where you play that hero to carry with little to no support. So if I’m switching to hanzo in a losing game and have a 30% winerate on him it’s not to say I’m bad at him cause I maybe picking him in games where I have a 10% chance to win making that 30% really good.
Once you understand this you understand how silly people are that browse through profiles to insult people are. Then you understand why private profiles aren’t to protect people from toxicity but because people are too dumb to understand data and context.
At the beginning yes, but once you accumulate a lot of data from your game it evens the variance out. Do note that winrates and your stats are representatives of your performance in the games you’ve played, not your actual performance. That is because sometimes you lose because of your team and your performance in high intense games is heavily reliant on your teammates’ and opponents’ performance. That effects your stats heavily too
You know…
Thank you for saying this because I actually have a low winrate with Hanzo (35 percent this season) and historically this has always confused me because the rest of the numbers just don’t add up. I mean I have a 3.0 K:D ratio with him, average 25+ elims, 13k dmg/10 min, 43 percent accuracy, 12 percent crit accuracy.
These numbers are not bad. They actually aren’t far off from some top Hanzo players (which really just means that I play him well in my rank, not that I am as good as they are). But yet, I have this ABYSMAL WINRATE with him and now that you’ve said that, I realize it’s because Hanzo is my rage swap. He’s who I swap to when nothing else is working and I need something that can bust shields AND one shot. He’s the one I swap to when I have a thrower or a super smurf on the enemy team because I know that if I play Pharah I will die to said smurf when they swap to a hitscan, but I can probably go toe-to-toe with them on Hanzo. Doesn’t mean the rest of my team can, though, and it’s not like I’m GM so I’ll probably get pinged in the head no matter how much I strafe spam.
So I’m not really all THAT bad at Hanzo for my rank, I just pick him under duress (since I am a Pharah main) and so between being tilted into picking him and the fact that he’s my emergency hero, I’m probably already just playing him in a game where I was slated to lose by fate, anyway.
Well I do know this is true as well because my winrate with Zenyatta is quite high but it’s also high over a large amount of games. I have a 65 percent winrate with Zenyatta over all competitive seasons - that’s years of data. It was actually lower at 63 percent until a couple months ago when I started one-tricking Zen whenever I queue for support, because… well obviously I am a good Zen or just better than the rank I’m in and should, theoretically, climb if I just one-trick him because of that. (58 percent winrate over 30 games this season so far!)
But still, if you exclusively play one hero there won’t be any of those weird half-games where you swapped out of desperation skewing the average.
With the tradeoff of getting those weird half-games where you get countered and almost have to choose between defeat and switching for a victory. OW is not black N white for anyone with this 