The confidence levels are included in my post, and they’re good when you take long term averages. Especially in masters and diamond.
Also, when Mercy is in a bad place win rate signals low when only one side has her and high when both sides have her. Dual Mercy situations are actually inflating her win rates.
Data selection is much less likely to be biased when you have a large data sample size. This happens with low pickrate heroes in GM (smallest population of players, and also the only place where you really see these aberrations in trends) a lot, and I noted that Bastion is a particularly nasty example. But Overbuff tracks a lot of games on a daily basis. A hero with 15% pickrate and 56% winrate, like Mercy had, is very obviously broken. If you work out the numbers, OB ends up being a reasonably accurate statistical model for the game state at large as long as the hero has ample representation.
Overbuff also gives you the tools to check out trends. You can tell the difference between a perennially bad hero and a hero like Bastion by checking out how they perform on a day-to-day basis. If a hero jumps wildly between extremes then you know you’re not going to get much out of the eye test, but if a hero is pretty consistently staying at a very low winrate you can probably work out that the hero isn’t very good.
To be fair Mercy never had the win rate Ana has right now (at least in GM). If I remember correctly it would usually hover between 49 - 52%, while Ana is consistently around the 55% mark, with a pretty close pick rate (13.8%)
In Peak Mercy Meta it was around 55-56% as well while getting very close to the 16.666…% (100%) magic mark for pickrates.
A GM winrate of 49% is very bad, as you need to win about 53% of your games up there to keep rank. It’s also what Mercy is currently at right now, and why people are alarmed about her.
The number you need to stay in GM is actually about 54.4. I got this by averaging the win rates of all characters weighted by their pick rates.
Might be true, I just remember people using her low - average-y win rate to say that she was balanced which was usually counteracted with people mentioning that if every team had a Mercy one would obviously win and the other would lose.
there are a couple of things that people seem to miss
now that ana is currently the strongest main healer we might as well just keep her since you know: skill should be rewarded in this competitive focused game that wants to be an esport so badly and what we should aim for is to diversify the meta with different off-healers combinations
It isnt about skill tho, it s about the hero appearing in “every game” as op puts it. If it s about skill we wouldnt have people complaining about widow being in every game
yes, thank god better ana wins. It used to be immensely skilled ana vs mediocre mercy, mercy wins. Thank god, let’s appreciate good anas and skilled heroes being better than LMB+w heroes.
Problem is that Ana specifically synergizes poorly with a lot of compositions, her being a must pick isn’t good for the overall health of the game, same applies to Moira, same applies to Mercy (although to a lesser extent).
This will drive the winrate towards the average for that tier (which is also how I should have defined high winrate as - above the % you need to keep rank), not necessarily towards 50%. A hero seen as must pick in Bronze would be far in the negatives because Bronze players lose a lot - it’s a necessity of the tier.
Agnostika noted that the average winrate for GM is 54.4%, so you would expect a hero with perfect 100% representation to sit at 54.4% winrate – if they are very close to 100% then the hero’s state becomes exponentially more alarming the further away they are from that average winrate because of the certainty of a loss it implies when you don’t have that hero.
Anyway I’ve derailed this thread enough with laymans stats lol, I’ll leave y’all to it.
I never said above plat. That’s their average winrates across all ranks, not that it makes a difference because winrates really do matter at all ranks in tandem with pickrates.
Symmerta 2.0 had a high winrate due to the circumstances she was picked for. People generally brought her out for timebank holds on 2cp, which was the one thing she was good for. This meant she got a good winrate for being picked appropriately, but her niche status was reflected in her pickrate as it was extremely low. Compare that to Brig’s current statistics where she has a high winrate AND a high pickrate. The high winrate indicates she’s being picked for circumstances where she is strong, and the high pickrate indicates she is being picked a lot. Individually either of those statements would be fine. The first indicates a great strength in a situation and the second indicates viability in many situations. Combine the two statistics, and you find you have a hero great strength in many situations. The problem with being strong in a lot of places is that it’s hard for the rest of the hero roster to compete.
Basically:
Low PR + Low WR = underpowered
Low PR + High WR = situational/niche
High PR + Low WR = Versatile
High PR + High WR = overpowered
Ive seen teams in world cup using ana in dive and ana hates dive
well, someone has to be the must pick
have you seen the world cup? mercy still gets played. it isn’t the strict support meta of zen mercy anymore. it’s better. just a shame blizz sucks at balancing and refusing to acknowledge rez on E what made her too valuable so they gutted her instead. welcome to ana’s world. they made her the best healer thast could deal damage and nerfed both those aspects
Ana isn’t a must pick tho. She’s just way more fun than Moira or Mercy and for the first time in years people don’t blow up at an Ana on site. It isn’t her fault she’s finally viable again and also has one of the most fun kits in the game. If you’re gonna reluctantly play a healer in solo queue might as well enjoy yourself right?
meh. Might as well just give Mercy the ability to shoot and heal just like zen. They could replace the rez with a healing orb and bam, no more “hold left click”.
Since support mains love to contribute to elim so much, we might as well have every supports attack and heal at the same time.
its just the meta, Lets deal with it for a while…
Some heroes go, some come, lets have ana shine for a bit as we think how to fix the support imbalance
That’s the bunny hop, not ga cancel
And every game was decided by who has to better mercy or zarya, rein, widow, hanzo, your point is?
That’s a fair point. This game was marketed as an FPS. It’s not outrageous to think the mojority if the playerbase wants to actually shoot things. And given how much everyone hates their teammates around here shooting your own team would seem even more appealing.