Supports and Skill Rating

Announcements
Prev 1 2 3 4 26 Next
04/18/2017 04:03 PMPosted by Scott Mercer
The determination of being “On Fire” examines not just your own performance, but your performance relative to your teammates. The calculation of your SR adjustment after a match doesn’t look at your teammates, but instead compares you to the performance of other similarly skilled players with that hero across an enormous pool of competitive matches. So, we compare your Genji play to the play of other Genjis, Ana vs. Anas, etc. Since we’re comparing “apples to apples”, we shouldn’t see any kind of support specific bias in SR adjustments due to player performance.


What if the player changes heroes? Why are you comparing hero specific performance when you specifically encourage hero switching as a way to balance your game? If I play Genji for 30 seconds, then Soldier for 30 seconds, then Winston for 4 minutes, what is SR really looking at and when?
Lucio is my most played hero this season & I avg over 11k heals per game. Most support players I see when I check their stats, barley avg 5k heals per game. I refuse to play support till this gets fixed. With a 55% win rate and the way the system suppose to be set up, I suppose to be climbing fast instead of tanking
Highly Rated
"You did very well this game.. but you did less well than someone else in another game playing that character where all the variables could very well have been different, so you get less SR"

Oh, okay.

How does that method of giving SR make any sense at all? What does someome else in another game have anything to do with the game I'm playing right now?
04/18/2017 04:03 PMPosted by Scott Mercer
Over the past few weeks we’ve been reviewing reports from players about how their SR gains after a win appear to be lower when they play as a support. I wanted to share some of the results of the investigations so far and clear up some community misconceptions along the way.

As part of the 1.9 Orisa patch, we made a change to how assist scoring was handled to address what we honestly considered to be a bug. Players were getting full assist credit even if the player being assisted did very little to the target. This change, along with other more significant balance changes in the patch, meant that we needed to recalibrate the tuning for the systems that calculate a player’s contribution to the match. This was performed for all heroes several weeks ago, and we’ve already recalibrated once more after the recent 1.10 patch.

The change to the handling of assists also affected how often a player might earn “On Fire” status during a match, but the amount a player is “On Fire” does not directly affect SR adjustments due to player performance. There is some correlation between the two because both systems are trying to measure “How well are you playing?”, but there are significant differences. The determination of being “On Fire” examines not just your own performance, but your performance relative to your teammates. The calculation of your SR adjustment after a match doesn’t look at your teammates, but instead compares you to the performance of other similarly skilled players with that hero across an enormous pool of competitive matches. So, we compare your Genji play to the play of other Genjis, Ana vs. Anas, etc. Since we’re comparing “apples to apples”, we shouldn’t see any kind of support specific bias in SR adjustments due to player performance.

We’re still seeing anecdotal reports of some players experiencing lower SR gains on wins, but we’ve also been seeing other reports from other support players that their SR gains look correct. Based upon our investigations so far doesn’t look like there’s a broad systemic issue affecting all supports across all competitive matches. There might be a more localized issue affecting a specific hero, or a certain type of play style or game situation. It also might be something completely unexpected, so we’re doing a thorough examination of all the code that affects SR adjustment.

In the meantime, we do appreciate everyone’s continued feedback and reports about this. They do help, we are looking at them, and if we find issues we’ll let you know. Thanks!


Great post and all but really.... What are your "reports" from other players? Your testers? Tons and tons of support mains have complained about the same thing so dont you think there is an issue here? It isnt just one or two people saying it, it is many across all ranks.

I appreciate the effort looking into this but on the other hand, it doesnt change the fact that playing a DPS hero regardless of my own individual performance im gaining 25+ per win and only 15-19 per loss even if i do incredibly bad. Now on the other hand when i play a tank or support, i lose at a minimum of 22+ (usually 25+) and only gain 19 - 22 per win. Not to add in the facts that the MM system itself is severely flawed in game matching.

Just remove the individual performance gain/loss in general and put in a flat gain or loss. Sure it would be a long grind, but at least you wouldnt feel completely robbed of everything you achieved in a day of comp by 3 losses. Then you have the question "well what about people who are doing extremely well and go on 5+ game winstreaks?". Simple award them more comp points the more they win.

The system then would be a simple solution to an incredibly frustrating problem for the whole community. Might not be the best solution there is but it is a solution.
Thanks for the post. But as a note, this 'solution' you have built to determine SR gains and losses completely overlooks a key issue with supports:

If your team does well, you don't need to heal/res/trans/etc. as much.

Therefore your team/friends all move up fast playing the other roles whilst, because you as support "did worse than other supports placed in more challenging, completely different and closer games did more than you to win", this means you get less SR?

What if the strategy in the game I'm in involves pocketing Pharah to flank, keeping up bastion or turtling with Rein? Does this all mean my contribution was less just because in another game with completely different people, the healer had to run around like mad as the team rushes the point repeatedly to die and need more healing?

Additionally, from your comments, playing as Mercy I get the feeling now that having my gold 45% damage healed card at the end of the game penalizes my SR as I didn't get to res people as often BECAUSE I DID MY JOB. But if I just let everyone die and res them, this is seen as better....
That was an incredibly long way to say "You're wrong". Just kidding though keep up the great work. Love you Jeff <3
The issue is, usually around gold to mid masters, either your team will stomp, or you will get stomped.

So incase of a team stomping, they will be performing "average" in terms of other players at their elo, because stomping or getting stomped is the norm, so they will get a low SR gain of 20-ish.

But when you get stomped, you're performing way under average, so you lose 30-ish SR. So that might be the reason why wins give low SR and losses make you lose MUCH higher SR...
Honestly, since this individual reward system is just too RNG at the moment, I think they should just remove it as a whole and start testing it in PTR to make sure it works properly before releasing it again.

If you can't do something right, don't do it at all.

Just make it so that everyone gains a flat say 25 SR for a win, and then an additional 1-5 SR depending on their individual performance...

Instead of the current gain a flat 10 SR and then an extra 0-25 SR depending on their individual performance...

I have no problems with the way the current gain is, but if you're going to have something so significant in a game, at least MAKE SURE IT'S WORKING RIGHT.
Next may I kindly ask the developers to actually get in game, and play roadhog for a couple games. Watch the hook get broken most of the time if there's even a wire mid air inbetween the target (Yes, this actually happened when I hooked pharah out of the sky). Tell me if that's how it's going to be, landing a long range skill shot and basically getting spat in your face.
I know that this has been a hot topic for a while now and even the Overwatch team responded to it but I am still confused on what they want from us.

Here is the thing, yeah they can match certain players against certain players who play the same characters but you aren't staying the same character the entire match, some switching to react to something or fix something wrong with how you are pushing or holding Meaning some of your stats will change mid match from hero to hero.

Even if you stay one hero then the comparing you to other heroes can work.
Taking into account the following:
Kills
Deaths
Accuracy
Damage Dealt
Use Use Of Ult
And that can work for DPS, tank, and even defense heroes, yet supports can't really do that unless you are planning to front line support.

And not only that as a healer if my team pushes in hard and fast kills fast and takes no damage what the hell am I suppose to do? They took little damage so healing them I won't really be doing much and yet I can't go aggressive support cause they aren't made for that.

And then if they don't and just can't base it off the amount of healing I do, what do they expect us to do as supports? Unless my team is taking a lot of damage or dying a lot I can't support them like I should be able to meaning I will go up in the ranks slower still.

If you try to focus on damaging the team to up your stats your team might loose heath of that important buff they need to keep fighting. And you don't have enough survivability on most supports to be attacking people or getting in their face unless you are dspstanky.

What do you want us to do?
This is purely anecdotal but I rarely get more than 17-19 SR(rarely play supports) per win but consistently lose over 24 pts on a loss. regardless of avg SR of teams. Regardless if I played well despite losing. I am pretty sure you guys can investigate individual match history and see that trend.

It also strikes me as odd that you would simply compare my solider in a game vs the avg of all soldiers. You said that calculation doesn't take into account your teammates, but this is a team game. Team comp matters. If you have a leaver on your team, that matters. There are factors beyond personal performance that dictate how well you do in a match.

If you're telling us that it doesn't matter that the other dps on my team is playing junkrat into zaraya/pharah, then I would tell you that your system is broken. This system punishes you for losing, indiscriminately, but rewards you by merit? Doesn't make much sense to me
But your statistics on Mercy doesn't depends from you. It depends only from team.
Is it fair?
Hi Scott,

Whether working as intended or not there is a Psychology to this on a player.

As MERCY main at rank 1707 I got 2 SR for a win to go to 1709 then next game lost 30 SR to drop to 1679.

The psychology of that makes me not want to play the next game out of fear I may lose another 30 or even if I win there is a chance I may only get another 2 SR.

Also reports from other Mercy players getting 2,4,5,7,10 SR for a win and losing 30+ for a loss adds to that Psychology fear.

I know it's a stupid fear (it's just a game after all) yet it still exists and has an affect on the support player base.

I think no matter what system you use you are eventually going to have to add a MINIMUM GURANTEED SR GAIN for a WIN.
I think perhaps it may be better if you compare a player's stats to the specific game stats. The reason I say this, is because as people have mentioned, when you play support for a team that dominated and you hold first point, your stats will not reflect good play despite a win and perhaps game winning actions made by the support. I can understand why you compare it to the community stats because it finds the outliers quickly like people on new accounts with high skill. However, wouldn't it be better that you also compare it to the specific game so these kinds of issues are avoided for supports.

I also am concerned what happens when you swap. I am a support main but there have been games where I have played 4 different healers because my team wanted to switch things up and we were trying different strategies to win. It is impossible for me to perform as well as another masters player as Ana for example who played the whole game as Ana, versus me swapping a bunch to help my team win. I'm going to assume (hopefully) that you have thought of this and have something like stats per minute or per life. It is hard to say though. If I die 10 times in a game it applies 10 deaths to every hero I played. Even though I clearly did not die 10 times on Lucio when I played him only one round of KotH. I also just find I die a lot more in masters as a support because dive comp is really strong and I'm a high priority target. Does not necessarily mean I was ineffective.

Regardless, I am a much better support than I am at DPS but I gain more rating when I play these roles. Just think there needs to be more consistency between roles. I don't mind gaining a bit less than a DPS who performed really well. I do however have a problem gaining half the rating they do for the same game. I duo with a high masters/tapped GM player for most of my games and I am mid masters. I highly doubt he is doing double what I am doing in the same game. Especially since he wouldn't be able to do those things if I wasn't keeping him alive all game.

I'm sure this issue is a lot harder to fix with supports also because stats can be very similar between a silver level mercy and a diamond level mercy. What I mean is, they both could have done 15k healing but you can't see the details in the game that distinguish an amazing mercy from an average one. With DPS and tanks it can be a lot easier because you have accuracy stats and damage stats which for most heroes are affected by your accuracy. Things are a lot more easily measured.

I'm also just going to assume that you have rolling stats because the meta is always evolving. A support playing in the tank meta probably died a lot less than a support dies in the current dive comp favored meta.
The issue mostly relates to Rein and supports since their stats are largely dictated by teammates and the overall match. If a game is high tempo with constant fighting, dps oriented heroes like soldier can 'inflate' their stats from having more encounters. This hurts supports relative to others outside of Zenyatta since they can't keep up a comparable amount of healing. Then there's playstyles such as Ana using grenade to anti-heal instead of being a heal bot leading to less stats overall since you're guaranteed to get the healing stat up whereas anti-healing is only rewarded in the stats if a teammate follows up. Then there's Mercy, if you play safe with rez and hide you miss out on healing in favor of a well timed res as opposed to a mercy that just sits out there healing and revives no matter what the circumstance is that can be more detrimental. Then you have a Lucio who can sit on heals all game that can pump his stats up as opposed to the Lucio giving well timed speed boosts. Supports and rein just have a lot more complex production in a typical game that stats can't contextualize well enough to say 'ok this person did well'. You can look at a soldier going 40-5 with 30% of team's dmg and 50% kp and it's safe to say 'ok, this guy did well' regardless of context most of the time. Then heroes like Hanzo where the one who has no medals but got 2 picks every fight has worse stats than the one that just spammed arrows at a shield for gold damage and killed tanks for silver elims where he could've been more productive as soldier with that playstyle.

Overall I think transparency in how performance relates to ranking is the only way to say for sure since it definitely happens. I'll play Ana and do well, but only gain 20 sr for a win then lose 30 sr for net losses in sr going even then I'll get tilted and start playing dps and heroes like Zarya and get respectable gains and losses.
04/18/2017 04:25 PMPosted by KingSlayer
I appreciate this. I can personally tell you that I main Zen. I was always averaging 18 - 20 SR per win regardless of being on fire or performing well. Same was true if I performed very badly. Still got 18 - 20 SR regardless of skill. I have video proof if you want to take a look at my YouTube videos. Some of the matches I did poor, average and one I got four Gold at the end of the match we won but still only got 20 - 25 SR for the win.

I stopped playing Zen for a while and started playing DVA. Suddenly I am getting at least 30 SR for wins even if I did average. If I did really good then I would get up to 40 SR for a win and not from streaks.

Please investigate Zen.

I would have 1300+ Transcendence healing, 15,000 healing, 50 Defensive Assists and 40 Offensive Assists as well as 15,000+ damage and still only get 20 SR for the win. Then another game I did very little of anything with Zen and got 18 SR. Only a 2 - 5 SR difference from performing really well to poorly.

YouTube Video Proof

https://www.youtube.com/channel/UCuHFc4ET1ZpAQKxETupx7MQ


Same goes for Lucio. I have been a healer (mostly Lucio) main for the past two and half seasons. I would get ~20 SR on wins and thought this was the average, and what you are supposed to gain after a win. Until recently I stopped playing support because of these issues, and I am gaining 30-40 SR on Junkrat, Soldier, Dva, Orisa, etc. What's more frustrating is the losses. I would lose >25 SR when I used to play support, versus ~20SR when playing dps. I appreciate the devs investigation and response to this, but I still truly believe there is an issue.
Weird. The least SR I gained from zen is 27, in a control map and we went 3-0. I performed pretty poor at that match(14 elims only, which is bad for a semi-dps), but still got 27 SR. The other matches where I won as zen I basically gained 30+ SR without streaks or medals other than healing (sometimes a bronze in elims/dmg). The most SR gain as zen was on my smurf, 55 SR, from 2910 to 2965, 2 golds and 2 silver(on a 5-win streak but I don't think the streak bonus is activated).
I've seen a couple of posts complaining the SR gain of zen is broken. I am just confused. Even in my worst performed game I never get SR less than 20 (For all Heroes, not just zen).
PS. The games I am talking about here are all after hundreds of matches in that season, I.e. No bonus SR.
PSS. I only solo Q in competitive. What really surprises me is that, while I never gain less than 20 SR even if I don't contribute much and carried by teammates, I often lose less than 20 or sometimes even less than 10, in games I lost but get high k/d ratio, 3-4 golds, and on fire most of the time.

Join the Conversation

Return to Forum