Sounds good to me. We are constantly discussing this issue in our group because it is really frustrating to lose SR and it not being your fault. An other issue is the getting a disconnect within the first minute of a match start. You cannot get back into the game within 10s and get penalised with SR reduction and a 10 minute ban from competitive. It makes no sense to have the match cancelled when it hasn’t even started.
True, if a person DCs in the first minute of the match, he is 90 percent of the time not a thrower or anything. So when no one is getting any negative impact from that person genuinely getting DCd in the first minute as the game is cancelled it is only unfair that the person is being punished for something which is outside his control.
The thing you didn’t address is how SR works when you in a group of varying SR, you said everything is even when your all the same. I play exclusively with my wife and or friends normally a group of 2-4, I’m 2700ish and my friends vary from 1700-2500, in games I reguarly gain less and lose more than everyone else in my team, and I normally end the season lower than my placements. This should say that I’m too high and I’m not playing well, but when I check oversumo and other similar sites my performance is reguarly in the 70-99% of my rank. And then when I place it puts me back up in mid plat every season. There is definitely SR adjustments going on here and the fact they haven’t addressed this in the post gets me suspicious because it will stop people from wanting to be the highest SR in a group.
TL;DR you get SR punishment for being the biggest SR in your group and they haven’t mentioned this in the post
1 Like
Since we are on the subject of SR and win rates, two questions i have had for really long are:
-
does the total amount of SR between players on both sides remain the same after a match? so if on one team all player gain on average 25 SR, do all players on the losing team lose on average 25SR, or does that not apply?
-
how does the system handle leavers? say team 1 has a 40% expected win rate, and a player on that team leaves, does that mean the remaining 5 players on team 1 lose less SR than they would lose if the leaver hadn’t left? or is the amount of SR lost for losing dictated at the start of the match?
i love the moment of whiners consternation when Blizz publish data that totally obliterate their ‘argument’
they should do it more often… this way we will know real state of the game and forum whining and toxicity would decrease
id allso suggest that blizz should do polls on forums, etc.
this way we would know that they care, because there are allso people who throw left and right accusation that ‘blizz dont careeee’
1 Like
I actually came here to talk about this, and found this great post. I suscribe all this, but just doing a plain +25 SR on win and -25 SR on loss it’s ok too, maybe better.
If there are no performance based statistics then everyone in the groups always get same amount of SR, even if they have varying skill differences,
which is bad. And if there are performance based statistics they are going to be flawed in many ways, which is also bad.
This is sole reason why there should be pure solo queue. You cant “cheat” solo queue because it is purely based on your own performance alone. It makes the matches more balanced because there is no people who were boosted playing within groups.
No performance based stats are good cause people don’t have to try bad plays to get better stats and ultimately risk not winning. Winning should be the only goal in everyone’s mind when playing a game. You can have gold medals in everything and still be a bad teammate and be a huge reason on why your team lost the game.
Only solo makes no sense, because you have players that main heroes like any tank or support, who require the team to have an interest to work together. The current SR system below Diamond has no incentives of doing so. If you can deliever stats, which might not be relevant for the outcome of the match, you will get more out of a win, and decrease less on a loss. Which is why some people stick to one-tricking niche heroes to play the system and get higher ranks despite win rates below 50%.
I can see where you are coming from, tho. I would like solo queue to only put you with solos. In fact i would love the system to always try and mirror the group constellation such as 3-2-1 or so. Moreover, ideally the matchmaker wouldnt put you against groups bigger than your own. No worst experience than duoing and then having a 2-2-2 with no voice coms against a stack of six. On paper that might be a 49 to 51 chance of winning between these teams, but if the three duos come with a support and DPS main each, you are already at a disadvantage.
Naturally i know that system would mean longer wait times when queuing and so that will not happen.
so when will this hit live?
1 Like
Just lost 50 SR and was benched for 10 minutes because failed to connect to server… great…
I hate this SR system so much. Please change it!!!
I appreciate the post and new LFG feature. That said I want to clean up a huge misconception too. As our respected developer, Scott Mercer said.
There are almost no flaws in the sr system. Let’s remember what SR means “skill rating”.
The biggest issue that everyone was and still having is “the synergy” problem. If our developers are telling the truth. There is no other thing we suffered for this long. We had good teammates as individuals, but as we all know every people has their own opinions and even we all have to work together to win the match most of the time our “individual” “similar skilled” “teammates” cannot find the synergy in the team they’re having on that moment.
So, in other words, they (developers) already know what the problem is but wanted us to fix it. The expectation they got was high but they expected us to do it. As we all know and LFG’s existence is the reason for it. Even we all are individuals and we need to work together to win. We just couldn’t do it. Because of many many reasons.
Can you answer questions on placements then? Because most of the time it doesnt really matter given the whole MMR situation. It just feels like it places you 50-100 sr lower than last season whether you win all or lose all matches or a mix.
Thank you for the interesting and informative post! This directly clarified some of the misconceptions and hesitations I had regarding queueing as a group. I’m looking forward to the LFG feature, and being able to focus on playing the roles I enjoy most!
Lots of good information here, none of it particularly surprising to me, and I appreciate you sharing it. It’s impossible for me to not make a comment though surrounding:
The greatest benefit of grouping is the simplest: You get to play together with your friends! There’s really no better way to play Overwatch.
I’ve got an SR that puts me in the GM level, and I have a bunch of friends who have essentially quit because the matchmaker doesn’t let us play competitive together. A bunch of the friends I game with most of the time are gold-diamond, and probably could get diamond/master if they played more, but they aren’t interested in playing quick play because it’s a dice roll what you’ll get in terms of team and players, and most people aren’t trying very hard. They’d play more if they could group with me, because playing with friends is more fun than queueing into gold level games with random people shouting at you for your pick no matter what it is.
It would be really nice if there were a way to play ranked-style (even if you don’t get a rank, or it has a separate rank to “true” ranked) with teammates who were outside of your available SR range. I and my friends play a lot less OW than I would otherwise just because my friends can’t play competitive with me. If this could be resolved somehow I bet you’d fix a lot of boosting/smurfing problems, too; I wouldn’t be surprised if some of the people who are intentionally throwing games are doing it to get into an SR range so that they can play with other people they know.
There are no flaws with Queueing at high ranks OMEGALUL
The devs do not know everything, just because they made the game doesn’t mean they understand how it works.
I like that you tell us the truth, but there are some questions that go unanswered.
This is actually very interesting because it contains data that you’re blind to atm. IF you get 10 games in a row with a 40% winrate, then it’s possible to lose all 10. EG, your SR will go down to a lower tier. This also means that even if you play really good and it won’t matter. Cause the opposite team has a 60% winrate over you. I feel like there’s data that should support this, but you just have not collected. Yesterday I played 10 QP games in a row and lost them all, BUT I was ALWAYS gold. Is it possible that you queue me up with players at lower skill rating to try and match the teams skillrating vs the opposite team?
EG:
Team 1:
Player 1: 40 skillrating
Player 2: 45 skillrating
Player 3: 40 skillrating
Player 4: 40 skillrating
Player 5: 55 skillrating
Player 6: 60 skillrating
Team total: 280 / 6 = 46.6
Team 2:
Player 1: 50 skillrating
Player 2: 50 skillrating
Player 3: 50 skillrating
Player 4: 50 skillrating
Player 5: 45 skillrating
Player 6: 40 skillrating
Team total: 285 / 6 = 47.5
So based on your system, this is a valuable matchup. Team 2 is on an advantage, but if you are the 60 SR person on team 1, you’re up for a really bad game. The reason being, you’re meant to carry the team in this case.
I can’t say that this is the case, but yesterday and many other times when I go on a loser streak with ALL GOLD MEDALS EVERY GAME, this is the way it feels. And if it is, it’s not a valid system.
Also, I’ve been stuck at Diamond since Season 2 and I think I’ve improved A LOT, but the problem is as soon as I reach 3.4 it feels like the team SR decay, it feels like the other team is simply better. Not necessary their counterpart to my role, just the whole team and therefor I think you use the example I took above. You place one person who are actually way higher than others players on that team or vice versa. Trust me, I’ve been into games where I feel totally carried as well, like the other players or at least 1-2 players on our team are just rocking the match, me just standing on a payload for 10 minutes and hear the voice “Victory” at the end of the match. Like “what was that!”
I also think there are games where everyone are at 55 (hidden) SR and these are the best games, where it feels REALLY balanced.
But as I said I can’t prove anything, I just think this is the way it works, cause that’s what it feels like.
1 Like
Why don’t they show us what the duo queue win rate and solo queue win rate is at gm/top 500?
Or 6 man at gm/top500?
They won’t, because it completely obliterates 90% of this post. Queueing is the most imbalanced thing about overwatch, and always has been.
2 Likes
top 500 players group up and cooperate, because they are good in team baised games as TEAM… and they are topp50 because of that
there are 40.000.000 accounts (not all active, some alts, etc,) and people were flaming about ‘matchmaker’ … so now they got statistics from matchmaker that blew up in their face
matchmaking works differently in top500 because there are 500 of players in top500, not 39.999.500
if they wont group up, they are out of 500… simple as that
Looking at this data it seems close to 50/50 if you’re in a group or solo. However, it also looks like Solo players do lose most of their matches, even if it is close to 50% and groups do win most of their matches, even though its close to 50%.
I wish the solo play experience could somehow be further improved.