Imagine if MMR was visible, and you could compare your MMR to the rest of the players in your SR. Do you think there people would feel satisfied or cheated? Now Before you comment “People would just say the MMR has rated me unfairly…” That’s not my question. Do you think people who are of a high MMR for their SR would feel cheated that the system is constantly arranging teams with lower MMR for their team and a higher or an equal MMR based on their personal MMR for the opposite team. In order to create an “equal match” or a 50% win chance. This is admittedly how the system works. This not only prolongs wait times in order to rig matches against higher MMR players but it also creates bad players to win games they shouldn’t and good players lose some they shouldn’t.
I’d like to know how mmr is calculated info they do this, because I played game with 0 deaths as Brig had an amazing team save where I whipshot junks tire in the air, pretty much frontlined the whole match and held enemy at spawn and I gain a mere 23 SR, yet when I lose its 38 SR or something ridiculous. MMR is a bunch of broken bologna.
It takes a long time for the system to accurately adjust your MMR if you make improvements. If a new player performed like you described in that game it would probably reward them double. But because you have been at your SR assumptively for awhile you are being held at what SR it calculates you should be at by many seasons. So if you all the sudden make massive improvements to your game instead of rewarding you with more SR it will start arranging matchups against your favor first before it starts to drip a little extra SR your way.
But then you’d know that it’s rigged to make you lose with teammates 1k points below you
It depends.
Visible MMR could look like this: 1.37347284278456823468723468364832 and so on, which isn’t really useful to us.
It could also be that it’s -3 <> 3 and that is per ranking.
So you’re 2 in Platinum, means you’re high on the Platinum ladder. Though if you are -4 in Gold it would mean you’re actually 3 in Silver.
The whole reason why SR exists is because it can group you in certain rankings, and it pretty much shows you where you are instead of having to read a long number.
The MM also tries it’s best to match you according to your rank though sadly enough this also depends on player availability.
I think it’s also a good thing if it isn’t disclosed how the MMR system works because then most likely people would just abuse it so they play in a way MMR favors them.
It might be better to keep it as it is because maybe people will start making excuses more on the system whilst their actual performance is lacking. Maybe creating the issue that we have with Medals. Might happen, might not though i don’t think it’s the best thing to do.
I think they would be suprised that MMR is practicaly same as their SR 
I am confused by this. What do you mean? Matchmaker is creating games which have two balanced team having basicly same average MMR.
Good thread and good questions. Of course I agree with you.
Blizzard owes players transparency and fairness in competitive ranked matches. MMR is criminal manipulation.
I stopped believing in “the System is keeping me down” (more so than ever before) after seeing a streamer with no arms reaching Master’s while playing with his feet, literally.
I see some of the new replies don’t seem to understand the premise of the question. I getting the feeling it’s because they have a loose understanding of how the MMR system works in general, so I have linked this thread that explains it in further detail.
Yeah the medals system is completely useless, I really hope they do something about that for OW 2.
If they don’t want to give us an actual proper scoreboard, the least they can do is change the medal system so it shows you how well you are doing compared to the average performance of a player in your current SR, instead of comparing you to your current teammates.
I imagine I’d feel cheated. I allready feel cheated: I play in gold mostly, though some of my games are plat averages. I know my MMR - I play some games of QP and look at the profiles of my team and enemies. Heavily weighted platinum to diamond - on average. There are golds there too and occationally silver. Very rarely Master and GM.
So My MMR is typical of someone somewhere in mid - high Plat, and those are the enemies get in GOLD…yeah I’m angry already.
From what I read -3 is bronze and 3 is masters so a silver game everybody would have the same mmr.
Also mmr (forgot the name) only steps in .5 standard deviations. So between high silver and low silver every is around -2 to -1.5.
It’s not drastically different is my point. And I think that might be a mistake. It should be more granular- smaller deviations and larger range.
That’s a good observation but that’s not my point. My point is that if you are of a higher MMR you are placed with teammates of lower MMR so that the other team can have a fair chance.
That’s never been the case. The only time you or the other team will have players who aren’t of the same or extremely close rating as the rest of the lobby is when people in groups become involved. The game doesn’t see you as a 24 and need to find a 20 to put you both in a 22 average lobby unless you’re in a group. It just sees you as the equivalent of a 24 and puts you in a 24 lobby.
People have imbalanced games for three reasons, nothing more nothing less. Anyone who thinks it’s anything malicious or intentional that these games happen are really thinking too hard about it.
-
Humans are inconsistent. It doesn’t matter how accurately someone is rated, whether by their average, best or worst performance, everyone has the potential to over or under perform their rating. All it takes is one or two people on a team over or under performing what they usually do to make a game lopsided. Especially in ranks that aren’t the highest tiers because if a player was more consistent towards their peak they’d be ranked higher. There’s nothing that can prevent that unless you can perfectly predict the future beyond just guesses and averages.
-
Tilt/chemistry, it doesn’t matter how well someone plays or even how consistently they play as an individual if they are fighting their team in comms more than they actually fight the other team. Same reason why some “super teams” in sports fail to win a championship (looking at you Brooklyn Nets).
-
A rating being inaccurate, I think is both more and less common than a lot of people think because of everyone’s inherent inconsistency. In the same way that it takes just one or two people performing better or worse than their accurate rating to make a game uneven, it also only takes one or two people to be over or underestimated by their rating to make it uneven.
I think you missed the point - HeyZeus probably said it better.
In that game your playing, YOUR ALL the same MMR. with the standard deviation being “huge” compared to the range there is not enough difference in MMR. For your original point. Just cause I have -1.5 and I see someone with -1, mentally its too close for me to “Feel” anything.
I mapped the scale out.
| Bronze | Silver | Gold | Plat | Diamond | Master | Grand Master | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| -3 | -2.5 | -2 | -1.5 | -1 | -0.5 | 0 | 0.5 | 1 | 1.5 | 2 | 2.5 | 3 |
Silver is starts at -2, Gold starts at -1… So with Std Dev being .5 if your playing Silver in that lobby and its a “high” game (meaning its silver to gold - a “low” game you will only see silver, or silver and bronze) you will see MMRs of -1.5 to -1.
My whole thing is, the range is too small, and std dev is too big. But what if everyone in the lobby all have -1s, is that one of those “epic” games? or is that still a steam roll? If its still a steam roll - then they need to pivot to a new system lol.
It would be interesting what the queue times are really waiting for, if we’re waiting for other people to finish their games, or are we waiting for the matchmaker to actually make the matches. - If its people waiting to finish thier games, I would rather wait 2% more in the matchmaker system so that its an epic match.
I don’t think that really even plays a role in it.
Why?
Because there’s no reason for it to just do the standard deviations in whole numbers or even just single decimal points. They can go to as many decimal places as they want and by doing so create as specific a rating as they want without creating a larger range of deviations in the system.
Simply put, as much as a 1.5 in a system that went to 3 would be like a 3 in a system that went to 6. The rating relative to everyone else doesn’t change, just the specific number assigned to it.
I don’t even think there’s a limitations on how far the deviations can go, it’s just statistically speaking, very few data points ever go past 3 standard deviations in a bell curve. Changing the size of the deviations would make them no longer standard deviations and wouldn’t really change how people are relatively to each other any more than just adding extra decimal places in where they land on the plot.
Could there be a reason behind putting a stricter lenience on what it considers a match worth starting? Maybe. But it won’t suddenly make someone’s rating more or less accurate than it already is.
At this point its all spec… No dev is going to come on here and say exactly how it is.
My greater push is to at least try something different. But its been 6 years… I’m mostly fighting for OW2 to be better.
For this post unfortunately the thought experiment doesnt hold for me. In any lobby there are going to be good players and bad players because of the “time” they waited, and the ping they have. its the “classification” of good and bad I think they need to rethink.
My main thing is no rating system or matchmaking system will ever produce consistent games, even consistently close games for the simple reason that, even if it manages to rate everyone accurately based on either their peak, base, or average performance, people are inherently inconsistent. All it takes is one person on one of the teams, over or underperforming their expected output to make a game that was otherwise close to a lopsided one.
Even within determining if a player is inherently good or bad for their rank is tricky. One player may have the mechanics to maintain at a rank no matter what mistakes they make or how uncoordinated their team is but they themselves can be very poor when it comes to their own decision making in game, not just in terms of hero choice but also cooldown use, positioning, ultimate use/tracking, and target priority. While another player can have rather weak mechanics for their rank but be great at tracking their opponent’s cooldowns, finding opportunities to punish their opponents mistakes, and positioning themselves and using their abilities so they minimize their own mistakes and give their opponent as few opportunities to punish them as possible. Both can be at the same rank but function better or worse on different teams. The player that is strong mechanically but struggles mentally needs the team to play around them more because they will inevitably make mistakes that their team can compensate for by creating synergy with their picks. While the player that struggles mechanically in their rank will need a teammate or two with strong mechanics to play around and to help pull off any plays they see the opportunity for, because no matter how smart they are in the game there is a good chance they see an opportunity to take someone out but its not something they themselves are capable of.
A lot of players are specialists but even simple things like the map they are on or the specific matchups they are in/with can drastically change their performance. For example, I consider Junkrat to be my “comfort pick” when playing damage, but know there are matchups and maps where I’m even better on Reaper or on Sym or Mei, while there are also matchups where I know that I can do more to disrupt my opponent by using Torb or Sombra. My mechanics aren’t spectacular so I have to leverage any advantage I can find that I have the ability to pull off.
Source? Also there’s groups in every game… Plats queueing with low silvers in my games… Except when i que with some friends on their alt stuck in gold/low plat - suddenly it’s sweat heaven and intense games…
It’s blatantly obvious to me that the players on my team and enemy team have wildly different skill at the same SR
In fact I know it to be soo, just finished a winstreak and one loss with some people:
1 a widow player with a main in DIA 2 a support main playing tank with a T500 account on console.
Both “stuck” in Gold/low plat and matched with me in high gold games. There is no way on earth or hell that the imbecile feeders I get in 50% of my matches are the same skill as these two guys in either tanking or DPS.