This isn’t what I said. It takes the sum of all heroes (Pick rate*winrate)
It’s not some it’s all heroes and looking at each rank to see a trend.
This is straw-manning.
I have an alt which I do “experiments” with. Try and play with different styles etc.
I pushed it WAY WAY higher than I have ever gotten on my main by abusing the matchmaker and Mercy.
I was trying to come up with a method to measure impact, which Blizzard could apply, and once I came on an idea, I thought… this is actually testable.
The test was, I would in each match try to hook up with the highest ranked DPS player in the match, totally regardless of how well / badly he played, and hard Mercy pocket him next match.
The theory was, I would be QUITE a lot lower in SR than he was, and would drag the match SR down, and because I would be “low impact” and he would be “high impact” then the team would be playing closer to his SR than mine.
Rinse / repeat. No real brain being applied, just boost the highest ranked DPS player, and try to move to the higher ranked DPS player as a duo each match.
I got 2 ENTIRE ranks above my peek SR doing this. The value I got by being lower SR and giving them easier matches was more than Mercy was going to ever give any other way.
It was pretty horrible. But it also gives a way to measure impact 
If a hero’s win rate is boosted if they are the lowest ranked member of the team compared to the boost from normal if they were the highest ranked member of the team, then they had less impact.
And the reverse was true.
Like I think the BASE argument once you get outside of trying to talk SR numbers and the like that Zephrin is making is a pretty good one.
You can argue how SR works, and matchmakers do things, and various calcs, but you know that Zephrin is right that supports tend to be low impact.
Now… is that a problem outside of Mercy? I’m not sure.
But like, I think the problem is, can we get them to queue? If they FEEL like they are useless, then they won’t, and them not queuing is a problem.
But it looks like you’all have moved on to an argument about the mechanics of the SR calc system, so I’ll make my exit, because it is impenetrable. Especially after Sr resets.
I calculated 3 averages from Streamer Data - Google Sheets (using the non decay summary sheet)
all average calcs done for the rows whereby SR column was in the interval (3600SR,4200SR) because that was what your previous point really came down to as I’ve already addressed the >4200SR case separately before as I argued a different way there.
so from those filtered rows:
average(|SR change|) = 24.50607287 (the first value I reported)
- average SR gain in (3600SR,4200SR) i.e. average of all the only +ve SR changes in that SR range = 24.53642384SR
- average SR loss in (3600SR,4200SR) i.e. average of all the only -ve SR changes in that SR range = -24.45833333SR
I didn’t say some heroes… I said some HIGHER RANKS.
I didn’t mistype. you misread.
For average on win
=SUMIF(E2:E710,">=0")/COUNTIF(E2:E710,">=0") =21
For average of loss
=SUMIF(E2:E710,"<=0")/COUNTIF(E2:E710,"<=0") =-24
You did it wrong.
Nevermind I missread it. Again my bad…
You cherrypicked a range
again 
I think there is a way to measure impact, but I don’t think that is it.
And the reason is this…
Say you had a hero which was strong but had 0 impact. like NONE what so ever.
It sat in the spawn room and made everyone else on your team stronger.
Any buttons you pressed did NOTHING.
It was strong, but your skill wasn’t involved in the game what so ever. The hero would have a high win rate, but no impact at all.
No amount of looking at winrates in isolation is going to detect that.
So winrates is disconnected from impact.
Now… if you could know which position in the team a hero was being played. (the lowest SR player, or the highest SR player etc) - then you could measure impact. because…
The 0 impact hero, would be just as good in the hands of the worst player as the best player.
So, if hero had a high win rate, when played by the lowest sr person in the team, and a low win rate played by the highest sr person in the team. Then, you would know they were a low impact hero.
Because the skill of the person isn’t making a difference.
The issue is, we don’t get to see that data.
You can’t even say “well the hero would be super strong in lower ranks” because you would trivially make a hero which the abilities they boosted on the team would be the ones which are more useful to people playing in higher ranks.
So, because a low impact hero can so effectively hide in the stats. I don’t think you can use the stats that we get from overbuff to detect them.
It doesn’t mean we can’t detect them, it just means we don’t have the stats to do so.
2 Likes
Yep misread it. My bad.
It’s not enough data though to prove your point and cherrypicked it.
That said even my over buff data showed Masters winrate dipped.
I would guess this is were the SR that is lost in +4700 comes from…
This would mean yes my model falls apart at 3k+.
Still doesn’t prove that you can have 50% winrate at all ranks.
Ignoring hero switching…
My point was over an infinite number of games a player WR would reflect there skill at the hero and their impact.
If you look at winrate in-game and use that as a gauge for their impact. I’d agree it doesn’t relate.
That said I was just using WR as a measure for skill it doesn’t really matter what it is… SR ect…
My point was simply…
Over infinite games, the skill of the people you play with won’t matter as they will average out and the only thing that will matter is your impact on the game.
I just watched a YO video where he basically says near the end that “Moira got a lot of flak over the years and it’s something we now have to get used to”
He advocated with other CC against Moira back when that certain youtuber went on about her damage orb lol, but now she’s ok because it’s OW2, funny how our playstyle is now “good and the proper way to play” but in OW1 we’re the “biggest throwers” ever
Also he said that the supports just sat in the car being carried, but that isn’t the case for top tank Moira players. So what he saying is Moira players are the only ones who didn’t travel in the car? LOL
1 Like
My point is, I can construct a hero with 0 impact which looks the game as far as the stats you get show.
Well yes, but, that is different from “hero impact” which is still very much a thing the game has.
Yes, you can really only change your skill, and you should work on that, but… hero impact is still important.
If a hero doesn’t feel like they “get stuff done” then people will naturally be upset at that.
Or if a hero is super high impact, then it will feel like in games the person playing that hero basically gets to decide the outcome of the match.
It is still important from a game design point of view.
Yes though, as a PLAYER, your skill is what is going to determine where you land in SR eventually. Some heroes will take longer than others to get you to that place though, and some people may give up before they get there.
Or feel that the game is too “random” because they don’t have the same level of influence.
I just treat people like that with disdain ;).
The game is hard in high ranks, it is hard for everyone. No one gets a free ride. The more people say a hero is “no skill” the funnier when I get them to play as them, and they hit the wall, and can’t advance.
1 Like
Yep, I pointed out that mercy does have an issue with how many games it may take but it will happen.
The point was indivual impact on mercy and the claim I disagressed was…
Where the implication mercy has no impact on a fight. My point is over many games this won’t matter and the only factor that will matter is the skill of the mercy, not the people you’re pocketing.
The basic point is when you choose to pocket someone that decision has a chance of getting a result. Mercy’s individual impact is measured by that chance not by the result after it happens.
Skill can increase that chance by staying alive, reacting faster, choosing the dmg/heal at the right time, and who your pocket and therefore there isn’t a disconnect in a skill when pocketing someone when you consider it over more than one game.
In one game there 100% could be but this is true to any hero as if your team underperforms if your indival impact is high it can totally negate your skill and create a disconnect in that game only.
Yes, Mercy has impact, no it isn’t anything like enough.
I was playing Mercy without a problem in Diamond. I had an alt, where I couldn’t get it out of silver playing Mercy.
There is an obvious problem there. I eventually had to Moira the account to get it to the rank where Mercy started having impact.
I swear to god Mercy is the only hero which gets easier as you go up ranks. There is something wrong there.
I mean you can say “over infinite games” but that doesn’t fix the actual problem, and make no mistake there IS a problem there.
There is a very real on the ground practical problem with the hero.
it’s like 340 points there and not even really skewed in which SR those points are for.
and again, I’ve proving for various SR ranges separately as mentioned before, that’s just the bit that addressed what you claimed before about me “eyeballing” the +/-24 number as the graphs supplied before wer for 3k-3600SR. So the averages calculated before for 3600-4200 was to show that no, I wasn’t just “eyeballing”, it’s still largely +/-24SR in 3600-4200 part of the 3000-4200 range.
like it’s not me cherrypicking when you questioned that specific part 1st and I was addressing it.
your model falls apart for <3K SR case in the almost the same way it does in the 3k-4200 case.
If you’re playing a high individual impact hero, then “an infinite number of games later” you would climb to your “true” SR/MMR eventually (as will others) and when you reach there, you’re unlikely to win against opponents who’re actually better, and vice versa if they’re actually worse than you.
Performance based SR doesn’t really change that because it only alters your base SR change more if you play significantly different in skill compared to where you’re currently placed. i.e. if you reached your “true” SR/MMR, then performance based factor isn’t altering much at all.
so yes, even with a high individual hero, the individual will have a 50% WR “after” an infinite number games.
similar a low impact hero would also get the same WR “after” an infinite number of games as explained before, requoted below
i.e. no correlation between “WR after infinite number of games” vs level of hero impact.
I’m not claiming there isn’t an issue with mercy. I even said the number of games it takes to reach your rank is an issue as you’ll be playing outside your rank for longer ruining more games and having a bad experience of feeling powerless.
In comparison to other supports, mercy has problems.
see I tried explaining to them how the kit deterministically doesn’t take much of the user’s quality of inputs as much of a factor in how much value they get, but they hard insisted on going “but muh model of WR over infinite games correlates with level of individual impact disproves that” hence the all this talk of “whether one results in 50% WR or so after infinite number of games regardless of high or low individual impact hero used” etc.
…
But Mercy has a reasonable win rate at all ranks, and an awful level of impact. I mean, any measure which actually worked, would have Mercy stick out like a sore thumb right?
Totally, and if you DID have a way of seeing how much impact a hero had, then Mercy would be your perfect test case.
If it didn’t work to show her as pretty different from most heroes, then it wasn’t working as a measurement.
Like, you have a test case right?
1 Like
I’d disagree.
look man. don’t look at me.
I’m trying to disprove that there’s actually correlation between WR “after infinite games” vs level of individual impact for them. I know it ain’t right.
that’s a question for them
I know, I was just agreeing with you. It isn’t going to show it up.
1 Like
If you wanted to measure issues with mercy’s impact.
If you could play constantly at 3k SR skill level
Make 100 accounts and measure how long it takes to get to 3k.
Then look at the spread (not the average) compared to other supports.
It will be much wider.
Edit: Well actually be better if you were a 4.5k player
True, but that isn’t data you can get from Overbuff. Like, again, it is something Blizzard could measure, looking at smurf accounts and how quickly they get to their stable SR, but…
We can’t, not from Overbuff.