MMR is accurate and its affecting your sr so it can be closest to your MMR, mmr will reward you with more sr for wins if you perform good. Its like this below diamond. MMR is your friend and will help you with climbing.
But SR isnât accurate⊠Thatâs kind of the entire point, itâs an estimation and a rather dumb feature since people focus so much on a number rather then the actual game. Losing or winning a few games wont really change your level of skill and many people would probably find it annoying to have a number that would change the tiniest amount or in other cases by a landslide.
MMR is the accurate value and itâs not a way to rig the system but to offer fair games. Youâll have an extremely high win rate if playing below where you should be and by no way being forced to stick around for too long in letâs say gold if you are plat.
Could you elaborate on how removing the accurate system and replace it with an inaccurate would offer more fair games and be âless riggedâ? I mean surely if you think MMR is some sort of matchfixing to force you to 50% youâd think the same if SR did the exact same thing.
MMR is there to try and offer fun and fair games between equally skilled players. No system will get it right every time, thereâs too many variables to take into consideration but thatâs at least the concept. It wont hold you back until you reach your level.
Or you want some wild wild west in which you put 12 people in the game with completly different skills in which the team who get the 1-2 diamond or above player wins? Pretty sure people would find some master genji obliterating their 4 bronze-gold and 2 plat skilled team quite infuriating.
Ok, now I have a little time to answer you.
To build a theory you need a lot of empirical data, which you get through observation.
With the data we have, it is not possible for us to make any certain statement about the matchmaking system.
Iâm also not interested in putting a theory together.
I have claimed something that only Blizzard could disprove with statistics. Thatâs my point. The following questions and official statements lead to my conclusion:
If a small development team is able to extract statistics that are necessary for all players and prepare them on one platform, why does Blizzard prohibit this project? (pursuit.gg)
And why doesnât Blizzard implement these statistics into the game itself?
Jeff Kaplan says that winning is more fun than losing, and a very long, close game can be more frustrating than a game where you get stomped. What does that mean for matchmaking?
Scott Mercer said that from now on the games are more balanced and the maximum chance of winning is 60%. (Based on SR)
What does a 20% difference mean? What are the actual chances of winning if a team consists of only DPS or support players?
Adaptive SR calculation for players higher than Platinum Rating has been removed. The official reason for this was that the system had already been exploited. (e.g. Onetrick Bastions)
The determination for MMR, however, still takes place to a large extent in this way.
I donât mean to say that a skill evaluation of players is generally not possible. In my opinion, a system which is secret and can freely decide with what probability a player may or may not reach a rank, clearly contradicts the idea of competition.
We donât know with certainty whether the matches are actively manipulated to give as many players as possible a âgoodâ experience.
This is an open question, many players ask themselves this question, and Blizzard could answer this question by providing detailed match statistics for previous matches.
Just the fact that we canât check whether MMR is calculated correctly contradicts the rules of any competitive game. Imagine that such a mechanic could be found in a âREALâ sport. Rumours of manipulation would immediately spread. In this way, e-sports will certainly never be taken seriously.
And yet somehow it does happens.
Because regardless of what you think enough people upvote and agree with it to stay relevant but hey if you can create a thread of similar popularity explaining how the mmr system is indeed fair you are free to do it but i doubt you can do that.
And yet i not just seen but actually played in a team like that last year. The top playet and the lowest player had 1300 sr difference between and while the system indeed tells you that such match is ptohibited all we needed to do is to regroup with the same team with thr weakest as the group leader.
Its simple really the matchmaker, the sr and the mmr system are all flawed messes.
Those people didnt just decay. They deranked the normal way.
Something being popular doesnât automatically make it true.
Blizz has changed the grouping system several times through last year to further close the gap between group members, especially at high rank. And 1300 gap is completely different than your 1000 Sr grouping with 3500 player. Thatâs beyond exaggeration.
And my rebuttal would be that you canât reasonably make claims without evidence. To quote Hitchens, âWhat can be asserted without evidence can be dismissed without evidence.â
I think pursuit likely went to the wayside due to the way it functioned in practice and how allowing that would make detecting cheats/hacks more cumbersome than it has to be, rather than Blizz attempting to hide the stats on it. Stat tracking is still a thing that can be done to identify strengths and weaknesses in someoneâs game.
It doesnât inherently mean anything at all. It could be an admission of what could be called âcommon knowledgeâ when it comes to all competitive games/sports.
This is a known issue theyâre working on, as evidenced by past comments about giving each account separate SR/MMR values for different roles.
This is conjecture. We donât know how they calculate MMR or if one-tricks are unfairly rewarded for their playstyle preference.
It not being open publicly truly is a shame, but that doesnât mean it automatically contradicts the idea of competition, nor does it mean anything nefarious is going on.
And on a final note - âREALâ sports donât consist of matches being played by a bunch of random players. Thatâs why there is a need for an MMR system. Itâs not perfect, which is why all the major e-sports use set teams as opposed to groupings of MMR-based matches, but itâs a better option than having no matchmaking at all, imo.
lol.
âthe proof is that I donât have proof.â Uhh, âŠokayyy. Can I have some of what youâre smoking?
This has nothing to do with matchmaking. This just has to do with the nature of the game. Sometimes you may get someone playing a character they donât know very well. Or maybe the team comp doesnât work for them. Or itâs not a great map for that character choice. Or maybe the mercy is just pocketing them and they look amazing. Or about 100 other reasons.
Again, there are not wild skill differences. With the exception maybe of smurfs.
?
Again, not true. I guess you could call the devs liars? But theyâve said SR chases MMR. Meaning if your MMR is higher than your SR, you gain more SR for a win and lose less for a loss. Thatâs how they keep these numbers close.
If MMR was shown, people would simply spend the whole game trying to increase their MMR rather than winning. The system is actually designed really well.
uhh MMR doesnât decay. If you mean they are purposefully performing badly so their MMR drops, then yeah, the system will think they are worse. But uhh, I mean, it can only do so much.
Iâm pretty sure the system doesnât allow for that much variance in rank. So no, you canât.
Most games I play, every player is pretty close in skill. There is the occasional smurf, but thatâs rare, and it usually ends quickly.
If you have 12 players of very close skill, yes, that game is fair. lol. How would you make âfairâ matches? Just randomly throw people on teams? That sounds like almost every game would be unbalanced.
Even elementary school kids playing dodgeball have figured out that you should make teams even by âaveraging up player scores.â Iâm not sure why so many kids on these forums think otherwise.
-
Cuthbertâs idea about handicapping is wrong. Iâve shown why itâs wrong in that very topic.
-
What youâre presenting is not even what heâs talking about. What Cuthbertâs point in that topic was that you can have a 12 players, all of them SR 2500, but they all have different MMRs in order to differentiate how good each player actually is, so an example would be youâd have 2 high-MMR players at 2500, 8 medium-MMR players at 2500, and 2 low-MMR players at 2500, and the matchmaker would have to spread that out evenly (so each team would have 1 high-MMR and 1 low-MMR player, OR the 2 high-MMR players would have the 2 low-MMR players as teammates with 2 more med-MMR players vs 6 med-MMR players). He came up with that idea by twisting a Scott Mercer statement about 50/50 matchmaking WAY out of context AND he ignored a statement how MMR works like SR and that itâs closely linked to each other except when inactivity decay happens.
However, itâs also important to remember MMR works the other way around.
It very aggressively tries to push and keep smurfs up at their belonged rating. A âgoldâ stomping on plats will earn huge gains to boost them up to where they wonât stomp anymore. Likewise, if the same âgoldâ stomps on silvers but still loses (because not all games are carryable) they will lose next to no SR.
Also, letâs not try to over exaggerate here. Games donât have anywhere close to as extreme in skill difference like you say. Again, the matchmaker does its best to get a group of 12 players with nearly identical SR. The only time any notable difference is ever made is when the queue times become very excessive and the game starts to loosen up its restrictions just for the sake of the player being able to play.
Even with smurfs, if their SR says 2548, the matchmaker will try to find them other 2548 players to play against.
Not to mention the only time youâll get a situation close to what you describe is because all those players are grouped together and the system is literally forced to make do. Though, also be aware that thatâs also impossible, since you canât group and queue for comp further than 1000 SR in the worst case scenario.
You seem to want to look at the rare, if not impossible, exceptions and complain about them instead of acknowledging the actual rule, which in reality is not that bad at all.
Thatâs the whole idea. The integrity of a competition game can only be verified when all doubts about systematic manipulation can be dispelled.
I can agree with most of your answers. But please donât confuse MMR with a realistic rating or the division of players into leagues.
I dont know whenever i read throught the various threads about surfing, sudden streaks of impossible to win/lose matches, seen my own 1000 sr fluctuate from bronze to low gold back to bronze makes me think that there are some horrible stacked flaws in the system.
We know that mmr has somekind of decay system but how much exactly is a mistery.
We have no clear idea on what is our mmr we only know that its close to our sr.
We have no idea exactly how does the system calculate our sr/mmr gain/lose.
It would be greatly helpful if blizz would release all crucial information about these but they dont want to what also manages to further fuel conspiracy theories.
By basic logic even with long quene time a player should never be matched with lower ranking players because when you lose it still gets calculated for you. Theres no such thing as losing nothing you will always go down.
Youâre free to basically call the Blizzard dev team liars, but the playerbase itself can also be extremely problematic.
Here are examples of players who believe that the matchmaker intentionally gives them bad teammates:
All these players managed to do was prove how bad they are at evaluating themselves and their teammates. Especially the first topic. That player clearly had or still has an ego problem, because each video he posted initially was titled stuff like âPERFECT Mercy gameplayâ or âPERFECT Moira gameplayâ and then he ended up changing the titles on each of his videos after getting called out all topic long.
And if you paid attention to these forums, a lot of people on here think like these two players. A lot of posts here still about how theyâre so much more skilled than what their SR is yet Blizzard is handicapping them so that they donât rank up. Thatâs why I recommend that people go elsewhere like the OverwatchUniversity subreddit or coaching Discord channels if they are legit looking for help or actual discussion. This forum is just too poisoned and add to the stereotype that gamers are immature children.
Anyways, back to your post:
A)Get wildly different skill levels for the same SR
Thereâs going to be variances between âperceived skill levelsâ in the same SR, but the matchmaker doesnât know or care about that, because as Jeff Kaplan stated, MMR and SR are âclosely linkedâ with the exception of decay.
B)How they balance for a 50% winrate on such wild skill differences
And because MMR and SR are linked, it canât balance the way you think it does. Itâs very possible to randomly place all the âgoodâ players on one team and all the âbadâ players on the other team, because as far as the matchmaker is concerned, all 12 of them have the same MMR/SR and thatâs all that matters.
C)Keeps players wanting to climb because bases on SR
You canât reasonably expect this because for you to climb, you have to kick someone down the ladder. The entire point of this mode is to approximate how good each player is in relation to the rest of the playerbase. Itâs going to be a rough approximation because this game is balanced for 6-v-6 play between organized teams, but people wanted solo/dynamic queue matchmaking.
D)If MMR was showed and instead of SR, people would have a cow at how wildly different the skill levels are being matched onto a team so this gets masked under SR and noone is the wiser.
Dont believe everything the Devs tell you especially when they dont provide proof and take there word as gospel.
If I had to guess I would bet MMR veries by at least 500SR ± over your real SR.
As I said earlier, youâre free to call the devs liars, but it has been stated that MMR works like SR, and that theyâre closely linked.
If you really think that MMR is something that can wildly vary from SR, youâre going to need to explain why. But I bet that your explanation will probably go into the same kind of territory as the two examples I gave earlier.
Letâs talk constructively about this subject and not speculate about personal motives, it doesnât make sense.
The Devs have to prove itâs not rigged. Otherwise, this game can hardly be considered e-sport. The Devs are also the only ones who can prove anything in this matter. So far they have only proven that they are not interested in making matchmaking more transparent.
Every single Blizzard statement on competitive mode and its matchmaking has been compiled into the following topic (check the references section):
Blizzard devs used to post on the Competitive forums, but when they get ignored in favor of topics complaining about rigged matchmaking or being called liars and other immature behavior being committed by players that post on here, I could understand why thereâs no point trying to reply to everything or even coming on here to read.
We know that mmr has somekind of decay system but how much exactly is a mistery.
MMR is stated not to decay, only your SR can decay, and itâs for inactivity if your accountâs SR is over 3000.
https://twitter.com/ww/status/867570441182826499
We have no idea exactly how does the system calculate our sr/mmr gain/lose.
Way too much to go through, but itâs covered here:
The onus is on you to prove that itâs rigged.
Also, e-sports does NOT have randomly generated teams at all. Teams scout and sign players, and if the players canât perform, theyâre out of a job.
Everything Blizzard ever said about Competitive and its matchmaking is compiled here:
People ignore it though, or flat-out say that the dev team is lying, because they want to believe that matchmaking is rigged.
Looks like weâre talking about this right now. And it looks like a similar thread is being created every day and there are thousands of other players talking about it. Personally, I donât care what anyone else thinks about Overwatch. I didnât develop the game and itâs not my reputation.
As Iâve already said, Blizzard must make the algorithms transparent or include traceable, detailed statistics in the game. Otherwise there will always be these accusations.
Even if it may sound wrong and unfair at first, it is important what the players subjectively think of the matchmaker.
And Iâm convinced that with so many posts thereâs always a bit of truth.
In all current e-sports games, Blizzard is the only development studio where secret MMR and secret matchmaking algorithms are applied.
1 Like
I think i have already read that some months ago, the problem is that we cannot really see whenever a statement is true or not.
When someone wants proofs that the system is not rigged they in most time want to see something they can clearly observe and not just statements, like for example an application what shows exactly what is your mmr, how much you gain and lose pet match and exactly why.
The problem is kinda the same as what warframe has with the riven mod system. Its an equipment what is randomly generated for random stat bonuses and weapons. The devs claimed that the reason why a player can only owe 90 of them is because the data takes up too much place but they have an uprise of tech people who can only shake their heads on this because even the worst system they can come up with would still allow lot more rivens to have.
Because the devs there are unwilling to share their exact storage methods the conspiracies loom, same as here because we have no way to get exact info on how ot match gone expect the endgame sr± the conspiracies thankd to the different various statements loom.
1 Like
How can the MATCH MAKING RATING not be fair? All it does is place you in a game with 11 other people with the same or very similar MATCH MAKING RATING!!!
You think your MMR is too low - just go out and WIN some games and itâll go up and you get to play with and against better people.
You think itâs too low - just go and lose a bunch of games and you get worse opponents and worse team mates.
Why is this a discussion again? I think we all agree that a diamond player is better than a gold player and that a GM player is better than a diamond player?
Not in the rank you belong? Go and win some god damn games then and youâll climb. Simple as thatâŠ
1 Like
It makes sure the mecahnicallly skilled stay high rank
No mmr put silver with top500.