Winrate as a stat is kind of irrelevant, as highly picked heroes will always have a winrate close to 50% due to mirror matches. Also, low pickrate heroes usually do not have a large enough data pool of games to gauge an accurate power level from them.
Now, with regards to pickrate, I think pickrate is somewhat irrelevant as well, except at the highest level of play. Say for instance, there are 10,000 extremely dedicated Hanzo one-tricks in Diamond. Then, Hanzo gets a nerf that makes the hero substantially worse, and all of the one-tricks fall to plat. This would cause Hanzo’s pickrate to rise in Plat, but it would not be because Plat players decided to start picking an easy hero in order to win. That was somewhat of a drastic example, but hopefully you get my point.
tldr: Bad heroes may have higher pickrates in lower ranks, since the players of those heroes may have lost SR by picking them. This can’t apply to GM, since there is no rank to fall out of into it.
Nah people will still say that heroes are bad and bring up GM pickrates because apparently if gm players say a hero is bad, then it is bad in gold as well.
Orisa, Sym, and Moira mains need to look at this post. I’ve dominated so many games with the supposed weak Orisa and Sym. People also need to realize you can’t fit Moira and Sym into every comp or map, learn a different hero lmao (unless youre casual).
Those hero rankings are based on playtime, stats, and win rate… you can earn them at any rank if you are playing a lot and outperforming for your rank.
Hero Score/Hero Rank/Quick Score/Quick Rank are pretty useless stats by any measure I can think of. They are also useless as a gauge of overbuff participation numbers.
I wouldn’t go that far. I think the numbers are reasonably accurate, but the raw numbers can only tell you so much.
Clearly certain metas favor certain heroes, so a hero’s pick rate might rise because the hero themselves is strong, sure, but it also might rise because someone they synergize with on their team, or someone they counter on the enemy team, is particularly strong.
Nothing wrong with the numbers themselves, just with the way people sometimes apply them.
I guess the dev that said that doesn’t work on the profiles and doesn’t know Overwatch has an API for that…
They’re the same for the items that are the entirely the same.
They are of course different for items where your profile displays data in per-season total or per-ten-minutes averages where overbuff is displaying it in per-match averages.
Not everyone is a one trick, in fact the majority of players aren’t one tricks, so in the end it balances out. While the sample size is relatively small it can yield workable conclusions by extrapolation.
The pick and win rates rising with meta is 100% correct. I wish more people realized that. Reason why a hero like Zarya gets constantly nerfed when Rush meta heavily favours her so she’s played more. (Also why plays think Zarya is broken god hero bc all they play in that elo is rush- even when it’s bad)
You realise there are fewer GM players in the game right now than ever before, and less than 1% of the OW population is in this rank?
You assume that they instantly fall from diamond to plat.
Imo, winrates are completely irrelevant however pickrates can and should be used to support an argument. For example, the pickrates of McCree led to him getting nerfed by showing his insanely high pickrate over other DPS. Pickrates of Sombra, Bastion, Orisa all remain low because they are extremely niche or just useless in the current meta.
I don’t think win rates are completely irrelevant. E.g., a hero picked because it is “popular and fun” rather than being picked for being strong or meta would have a lower win rate.
They can be really hard to interpret though, such as, a very low pick rate hero with a very high win rate might be picked only occasionally and very successfully as a counter, or they might be played mostly by very talented one-tricks giving them an outsized win rate.
Pick rates are of much more significant though, and win rates need to be understood in the context of a pick rate… it’s very easy for a low pick rate hero to have a high win rate, but as heroes approach 16.67% pick rate (i.e. 1 on each team every match) they will naturally be forced to a 50% win rate. Usually long before that, since in a strong single-comp meta people will feel forced to mirror, driving win rates towards 50% long before getting near max pick rate.
Except McCree has a lower winrate than heroes like Torb, Symmetra and Bastion because he has a much higher pickrate. In terms of DPS heroes, its usually the higher the pickrate, the lower the winrate.
Yes. As I said, I do agree that the pick rate is much more significant, and you do have to take into account pick rate first to understand the significance of win rate since win rate tends towards 50% as pick rate rises.
I’m just saying, win rate can add a little more detail, if you are careful when interpreting it.