Does percent win rate tell me anything?

General Discussion
If I have a 45% win rate with hero X, and a 60% win rate with hero Y, does that mean I'm better with hero Y? Does it tell me anything?
Well yes and no

yes because if you are better with a hero you will win more often

and no because you can just get unlucky with bad teams and lose a lot
Doesn't really mean anything. You can be the best (what ever hero) and still lose. Unfortunately you are still depended on 5 other people to do their jobs.
win rate does nothing
It doesn't count ones own performance just how many matches ended in victory verses defeat while playing as character X and Y not your personal impact.
If means you win more often when you pick Y instead of X.

You can use a site like masteroverwatch or overbuff to check your stats to see if you actually play X or Y well, relative to the rest of the community.
Yes if the number of matches is high that means it is not luck or randomness.
09/19/2016 07:39 AMPosted by Bullgrit
If I have a 45% win rate with hero X, and a 60% win rate with hero Y, does that mean I'm better with hero Y? Does it tell me anything?


If both heroes have the same or similar number of games played, then yes.

If your 45% hero is across 200 games and your 60% win rate hero is across 5 games, then no, there's still too much margin of error.
It's supposed to determine your "skill", but it's merely a way of telling which character you had the most luck with regarding teammates.
Stats are a useful tool but you have to be careful not to read too much into them. I have a 70%+ winrate on one hero that I know that I am not particularly good at playing. But it's the hero I tend to play when my team is already doing really well or I am on a very good team. One of my best characters has a very low winrate because it is my "tilt" character -- the hero I switch to when the match is probably already lost and my team is not coordinating. The end result of both of these behaviors is some pretty skewed stats.
I find eliminations per life to be a pretty good indicator for dps. I wish we had heals per minute, damage per minute, and damage blocked per minute stats.
Whats important to differ in Stats are to take in account from what skill level they are taken - or even what mode.

Game has lots of Characters which will have a high Winrate at a rather low skill level, while them being rather bad at a high one.

Then again - I am sure Bastion, Torby or Symmetra have a 50% winrate in +3300SR comp - since they were only picked twice.
09/19/2016 08:23 AMPosted by Jeff Kaplan
Stats are a useful tool but you have to be careful not to read too much into them. I have a 70%+ winrate on one hero that I know that I am not particularly good at playing. But it's the hero I tend to play when my team is already doing really well or I am on a very good team. One of my best characters has a very low winrate because it is my "tilt" character -- the hero I switch to when the match is probably already lost and my team is not coordinating. The end result of both of these behaviors is some pretty skewed stats.


Two very good points to consider as well.
But competitive rank is based on games won. If one is great at hero X, but lose, one's rank goes down. If one is so-so at hero Y, but win, one's rank goes up. Or is it possible to lose a game and go up in rank because one did well personally?
09/19/2016 08:23 AMPosted by Jeff Kaplan
Stats are a useful tool but you have to be careful not to read too much into them. I have a 70%+ winrate on one hero that I know that I am not particularly good at playing. But it's the hero I tend to play when my team is already doing really well or I am on a very good team. One of my best characters has a very low winrate because it is my "tilt" character -- the hero I switch to when the match is probably already lost and my team is not coordinating. The end result of both of these behaviors is some pretty skewed stats.


Holly crap jeff, nice analysis.
Actually a perfect example of what I do myself to keep my precious stats lol!
09/19/2016 08:23 AMPosted by Jeff Kaplan
Stats are a useful tool but you have to be careful not to read too much into them. I have a 70%+ winrate on one hero that I know that I am not particularly good at playing. But it's the hero I tend to play when my team is already doing really well or I am on a very good team. One of my best characters has a very low winrate because it is my "tilt" character -- the hero I switch to when the match is probably already lost and my team is not coordinating. The end result of both of these behaviors is some pretty skewed stats.

Well put.

Everything has to be analyzed in context.

Was the loss because of you? Could you have prevented it? Answers to questions like these are always "it depends" and make numerical analysis of player skill pretty difficult.

With the intent of a balanced match being a 50/50 outcome, there's no telling what impact someone playing a different character that they don't know will have on the match, or the impact of two people teaming up and working together.

I don't know how much Blizzard factors in but there's a gazillion variables: who you've teamed with, or your opponents or teammates have teamed with (positive or negative group synergy could have a bigger impact than individual skill), class selection and their relative performance with that class, the impact a given map and attack and defense has on those, etc.

If you're swapping classes to try to fill in gaps in the team, more often than not (at least in my experience) the rest of your team is doing something pretty dumb and rolling a bad comp. Your chances of winning still probably aren't that good.

That competitive seems so heavily based on win or loss and expectations therein is one of my complaints. I don't think it adequately accounts for all these variables, just because your team lost, should you really lose rank? Unless it's changed in S2, you'll still lose rank, just not as much. I don't think that's right. I also don't think a gain in rank should be guaranteed for a win, but either way, that's how it's been.
09/19/2016 08:23 AMPosted by Jeff Kaplan
Stats are a useful tool but you have to be careful not to read too much into them. I have a 70%+ winrate on one hero that I know that I am not particularly good at playing. But it's the hero I tend to play when my team is already doing really well or I am on a very good team. One of my best characters has a very low winrate because it is my "tilt" character -- the hero I switch to when the match is probably already lost and my team is not coordinating. The end result of both of these behaviors is some pretty skewed stats.


Any word on WM?
09/19/2016 08:23 AMPosted by Jeff Kaplan
Stats are a useful tool but you have to be careful not to read too much into them. I have a 70%+ winrate on one hero that I know that I am not particularly good at playing. But it's the hero I tend to play when my team is already doing really well or I am on a very good team. One of my best characters has a very low winrate because it is my "tilt" character -- the hero I switch to when the match is probably already lost and my team is not coordinating. The end result of both of these behaviors is some pretty skewed stats.


Why is no one asking which heroes???
09/19/2016 08:23 AMPosted by Jeff Kaplan
Stats are a useful tool but you have to be careful not to read too much into them. I have a 70%+ winrate on one hero that I know that I am not particularly good at playing. But it's the hero I tend to play when my team is already doing really well or I am on a very good team. One of my best characters has a very low winrate because it is my "tilt" character -- the hero I switch to when the match is probably already lost and my team is not coordinating. The end result of both of these behaviors is some pretty skewed stats.

Tilting? u mean u troll on purpose. Reported.
09/19/2016 08:23 AMPosted by Jeff Kaplan
Stats are a useful tool but you have to be careful not to read too much into them. I have a 70%+ winrate on one hero that I know that I am not particularly good at playing. But it's the hero I tend to play when my team is already doing really well or I am on a very good team. One of my best characters has a very low winrate because it is my "tilt" character -- the hero I switch to when the match is probably already lost and my team is not coordinating. The end result of both of these behaviors is some pretty skewed stats.


Can I get your word that this applies to competitive rank as well?

Join the Conversation

Return to Forum