Actually they have.
Links to both the dev notes about reducing SR range for mmr are even in Kaawumba’s thread. The counter argument on those statements is currently “they said SR but obviously meant mmr, while being able to distinguish perfectly between them every other time” or something like that.
Yeah, tending towards your MMR is done with differences in SR changes (like you see in decay) not with giving you bad/good teammates.
Not separating inferior players from superior ones!? Then why did the show matches between GMs and bronze players have 3v6? Even the 5v6 GMs vs Masters match was pretty close. Guess the rank videos indicated it is pretty easy to roughly guess SR based on even a SINGLE match. The professional players routinely dominate the highest parts of the ladder. Amateur leagues use SR as the skill cutoff to decent effect. Smurfs completely body players ranked below their true SR.
Your base premise is the ladder equivalent of saying the Earth is flat.
Your comment makes no sense.
You said SR isn’t part of mmr because no dev said so. I said there is such a post.
Your response feels like some strange parallel world.
Day 227; still no sign of balanced, working Matchmaking.
I’m too lazy to dig up “Kawumba” and his thread so paste the quote here. I guarantee it says nothing about creating teams based on factors other than MMR.
Matches are made by MMR, not SR.
“If you do decay, it only affects your current displayed skill rating. This decay does not affect the internal matchmaking rating we use, so we can still place you in fair matches.” – Scott Mercer (3)
“Skill Rating decays but your internal Matchmaking rating (the thing that determines who is matched against who – not SR) does not decay.” – Jeff Kaplan (21)
“We use MMR for matchmaking, not SR” – Overwatch official twitter (25)
“SR isn’t used to make a match.” – William Warnecke (40)
You may be confused about what MMR is. See How Competitive Skill Rating Works (Season 10) for a more complete explanation, especially the summary.
Unfortunately, there are also three posts that indicate or imply that matchmaking is either based on SR or matchmaking is limited to a given SR range (5, 7, 35). Scott Mercer is the author of two of the yes SR posts, as well as one of the no SR posts, so he contradicts himself. The contradiction must be resolved somehow, and I decide in favor of SR being used in no way during matchmaking for the following reasons:
Whenever MMR and SR are discussed in the same post (including by Scott), it is to say that SR is not used in matchmaking.
Master+ players regularly see decayed “diamond” players in their matches. And when someone like Seagull decays down to diamond on stream, he is still placed in grandmaster / top 500 matches. If SR was either used for matching, or the limits on matching were restricted by SR, this should not happen.
SR and MMR are closely linked (22), except for decayed players (who only exist in diamond+) so saying matching is done on one or the other is a distinction without much of a difference. This makes Scott being sloppy with the terms not as surprising and makes the answer to this question less important.
There hasn’t been any clear revoking of the old system on this point so the belief that we should only keep the newest non-contradictory statement (35) is implausible.
Generally the people who are insistent that SR and MMR are both used in matchmaking are using it to infer some broken or rigged system with horrible and implausible outcomes that are contradicted by careful analysis, common sense, and unrelated blue posts.
(3) Overwatch Forums
(5) Overwatch Forums
(7) Overwatch Forums
(14) Overwatch Forums
(21) Overwatch Forums
(22) Overwatch Forums
(25) https://twitter.com/playoverwatch/status/850435344457543680
(35) https://us.battle.net/forums/en/overwatch/topic/20759648155
(40) https://twitter.com/ww/status/867570441182826499
As stated many times before, it’s a semantics game.
SR is your visible rating.
Mmr is your rating used for match making.
Nothing stated in those quotes stops SR history (for example) from being incorporated into your mmr.
So every quote you have can then be correct without a contradiction.
You seem to think I mean there is a two step process, where SR is used to create a range then mmr is used within that range. This is incorrect, as SR is already inside mmr it is only one that is used in a single step.
No developer has ever said that this is the case. And it would be a silly thing to do, so I’d say no.
No. MMR is the “real” number. SR is a visible and friendlier approximation of MMR (Overwatch Forums) It has no more meaning with respect to matchmaking than tier icons.
Let’s assess the assumptions:
SR “Chases” your mmr
Quotation marks are used, so taking this word too literally is wrong. It is an indication of a pattern or trend, not similarity.
it’s possible to win a match and not gain any MMR. We make it so that if you win a match, you always gain SR
For this to be possible, the mmr can not correlate as closely to SR as you think. As soon as they are not close, then your model fails. They even admit earlier in that post the SR and mmr are calculated differently.
Also, Blizzard is famous for soon™ meaning minimum 6 months. They may also believe an effective variance of 1000SR equivalent mismatch between mmr and SR as close™.
Tl:dr - please stop pretending to know everyone’s mmr.
I generally assume that the developers words about how the system works are reasonably accurate, and that the plainest self-consistent interpretation is the correct one. If you are not willing to do that, there is not much to discuss, because you can believe pretty much whatever you want to believe.
Talking about when features will be complete is a very different matter than talking about how the system already works, and any developer (myself included) knows way too well.
P.S. You clipped that quote to suit your dubious purposes. It is: “So SR ‘chases’ your MMR very closely, except in a rare case of severe SR decay at GM/Masters/Diamond level of play.”
“Similarity” can be a vague word too. A pickup truck is similar to a sports car in that they are both motor vehicles. I’m willing to believe Bliz is using “similar” for things that are actually closely related.
How much separation do you consider “not close”? One match with a change in SR, but not MMR, would probably still keep both values close. SR decay would be the only time when SR and MMR would be drastically different. I think it’s reasonable to conclude the following:
I think number 2 is true because we know MMR is used to form teams, SR is not. The average SRs of the two teams are usually very close. Therefore, SR and MMR must be close together.
The devs might be purposely vague about MMR calculation, but I doubt they’ve outright lied to us. It’s not their fault if people are foolish enough to jump to wild conclusions on very little evidence.
Whoa this thread still alive?
The SR/MMR close relationship is a myth busted by reality:
People can go up or down huge amounts in placements (I once was put in diamond 400SR higher than my previous season). Clearly my MMR was WAY wrong for the previous season, for all my between seasons matches were low diamond rated, while my in season ones were low plat.
This happens to alot of people, not just me, just search for placement results and you’ll find a bunch of outliers like this. It mostly works, but not really.
The system sometimes amplifies losing streaks by giving you rectal secreationer and rectal secreationer teams wich throws you off causing more losses untill you hit a hard skill bottom where you get to bash as many seal pups as you want.
A neutral system would just match you on SR not giving 2 fracks about your skill, and thus not amplifying or nerfing anything. Games would become more consistent over time.
We can’t know how the MMR system works exactly. We can only see its effects, and personal reports can be unreliable. Here’s what I’ve gathered from this epic-length topic.
Cuthbert’s theory: If you’re playing really well, then the matchmaker looks for people playing poorly to place on your team. In this way the matches are “balanced”. I am in agreement with Cuthbert that the matchmaker shouldn’t do this. The difference is, I don’t think the matchmaker works like this in the first place.
My theory: The matchmaker tries to find 12 people with close to the same MMRs taking into account latency, avoid lists and other factors. It takes the two highest MMR players and places them on different teams. Then it takes the two lowest MMR players. It places the lowest MMR on the same team as the highest MMR (2nd lowest is teamed with 2nd highest). It proceeds like this until all slots are filled. In this way, the teams should be balanced.
I’m not convinced the matchmaker is completely responsible for streaks. I think that psychology plays a big part, especially in losing streaks. Losing causes you to tilt. Then, playing while tilted will make you more likely to lose, which makes you more tilted.
You’re right not to invest too much in anecdotal evidence from individual posters. But I respect that people are trying to relate my theories to their personal experience with the game. Pattern recognition often produces false positives. But it’s all we have, as human beings, and it can be true.
Thank you for giving this so much thought.
I wouldn’t call this first part a theory, but rather a logical conclusion based on what Principal Overwatch Designer Scott Mercer has told us. There is no other way for such a handicapping system to function. The game does not make matches with different numbers of players, for example, or adjust player damage output, health, etc., to create the handicap. It does this by segregating the best players from each other (and the worst players respectively).
We are on the same page. You’re right that the matchmaker typically places the best 2 players on opposite teams. But it’s even less fair than you image, because MMR is free to place the next-best 6 players in the match to oppose the best individual player.
I like your idea of placing teams in order of best to worst, alternating between teams. That would be slightly more fair than the current system. I still think that all forms of handicapping are wrong for competitive play.
You shouldn’t be convinced at all. How could MMR possibly engineer streaks for individual player accounts? Every game takes place between 12 players, and anything can happen. MMR seeks to ensure a 50% chance for players in every game to win/lose. If anything, MMR keeps streaks from signifying any kind of true pattern in players’ careers.
Out of curiosity, how often do you think this is the case? Do you believe it’s somewhat normal, or just a rare occurance at the extreme ends of the ladder?
Bump. This thread shouldn’t die because it’s 100% true. It’s infuriating.
the devs will never remove “fair matchmaking”, the only way they will change is if people start leaving the game. And the mass majority are ignorant to the backroom dealings that blizzard is playing.
Just today I’ve finally found evidence against MMR.
https://us.forums.blizzard.com/en/overwatch/t/i-have-3-accounts-gold-plat-and-diamond/135435/39
MMR literally allows a player reside in 3 ranks with the same stats. 3 ranks, same stats, Blizzard, cmon…
So you want the game to stop placing 3 strong players and 3 weaker players on each team and instead place 6 strong players on 1 team and 6 weak players on another team?
Summarized, you want the game to intentionally stack all the better players on a single team?
Re-worded, you want games to be as uneven as possible?
Re-re-worded, you wants games to be unfun face rolls all the time?