That is the 3rd time you have completely ignored the majority of my post. If you aren’t going to actually attempt to debate then why are you on the forums?
I have already said everything that had to be said about it. no point of repeating myself.
1 Like
Currently the only security is the huge conflict of interest behind online game matchmaking.
Blizzard is the provider but at the same time the beneficiary and sole controller of the fairness of matchmaking (the regulations for online games are still to be developed in the various countries and these companies have total freedom to do what they want).
Speaking hypothetically, the matchmaking of the elo system (or its derivatives) serves to combine games with players within a certain number of standard deviations of each other (in common words, match players with similar skills).
Let’s say that a fictional company, blizzard, is in a moment of crisis with its game playerbase and can’t combine games in the correct way, then decides to program matchmaking not to match players with only principle of standard deviations, but create unbalanced matches and strategically insert players in one or the other team to give them the impression of having a progression over time of a balanced matchmaking; here at the moment it is possible without breaking any law.
Just some educated guesses: K/D ratio probably boost hidden MMR number a lot. Then your general stats will be compared to average players in that particular rank (iirc this was officially confirmed)
Once they built your profile, they score it and give you a MMR score. The MM then will try to even out MMR between teams. So theoritically you have to outperform expectation (your own stats) to tip the favor to your team to win.
Sometime you will have unwinnable matches because your teammates don’t bring their A games. Heroes choices and compatibilities matter a lot too, same rank means you can’t dominate any of them in a 1v1 set up, but if your team has better synergies you will win by constantly making the match 2v1, 3v1, 4v1 ect
Pretty much. People have theories but most have no evidence or draw flawed conclusions based on what we already know. They’re a few thread with just the compiled facts.
Ok click on that link…click on the subheading for matchmaking…and read everything on there
There’s plenty of info on matchmaking (not sure how up to date that is but I know Wyoming updates that thread from time to time)
But I do want to stress that MMR, the thing that basically decides what matches you get put in, is a number…
a number that tries to define everything about you as a player…ie, your aim, your game sense, how well you communicate and work with others, are you grouped…are you having a bad day…etc etc…
it is never going to be accurate for you or anyone else…so people need to understand that it is limited in what it can do (not just OW…all games)…and it’s doing this for 12 people…even if you had the same MMR it doesn’t mean your = players…it means you’re = to it based on its calculations (2 very different things)
Like people understandably get frustrated with how bad matches can seem to be sometimes…but I go the opposite route…I’m more amazed at how many good matches I have considering what it’s trying to do…especially in something like QP where it’s an absolute crapshoot with the grouping and such
Hmm, from what I can tell it works a bit like this:
Basically, whenever people are matches, the game tries to get the match to go 50/50 as close as possible with the available data.
What data is that? We don’t know. They’re not sharing it because it means people will focus more on improving the numbers and not their skills. (while I would argue that you need to improve your skills to improve those numbers, but ok, just a choice they made to prevent potential cheesing. League basically allows you to climb out of iron just by placing more vision wards)
Now, of course, the game can not do this to an exact science. So some team might have a 51.2% chance of winning, the other team might have a 48% chance of winning, and there’s a 0.8% chance of it being a draw.
If the team with the highest win chance actually wins, they will, on average, get less SR for the win, and the team with the lower win chance will on average lose less SR for the loss.
But if the lower win chance team actually wins, they will get more SR, and if the higher win chance team loses, they will lose more SR for the loss.
Makes sense, right?
There’s another thing that affects your personal SR gain or loss, and that is your “hidden MMR”. Basically SR is where the game says you are right now, and hidden MMR is where the game thinks you should be, based on your performance.
In other words your SR is lagging behind your MMR. MMR can be more variable, probably, and will wildly shift around its values, while SR is more consistent.
The paradoxical thing about this, which annoys players, is the illusion of ‘being stuck’:
If your MMR is lower than your SR, but you win a match, your MMR will be adjusted upwards, but if that new MMR is still lower than your current SR, you will still lose more SR for a loss than that you gain for a win, because the game wants you to arrive at your MMR.
Same in reverse, which is why those “unranked to GM” games don’t take all that long: if your MMR is higher than your SR but you lose a match, your MMR is adjusted downwards, but you will still gain more SR for a win than you lose for a loss if that MMR is still higher than your SR.
It’s the MMR that is a mystery. Again, it’s hidden, and blizzard doesn’t tell us what data they use to determine any of this.
I think MMR is entirely dependent on personal display of skill. Which means that, even if you lose but you played really well, your hidden MMR goes up, while your SR might go down.
If I am informed correctly, MMR is not a number, often players asked why this number representing MMR is not displayed and one of the developers or even the game director himself at the time said that it cannot be displayed because MMR is not a number but rather corresponds to a very large matrix.
I don’t think you can map SR and MMR on a graph, nor would that make any sense.
Number, calculation, algorithm…doesn’t matter…point was that it’s an attempt to define everything about you as a player…it’s never going to be entirely accurate for you or anyone else playing…
A matchmaker can only do so much…
And yet overall I think it does a decent job…even in QP (that’s the really shocking one to me to be honest)
1 Like
They just stick random people together like I know someone who fought a gm when they’re silver in qp so anything can happen
Ok, I see.
Do you play on US servers?
Yes I do and quite often
Ok, I play on EU servers and some days I win 20 games in a row and then there are days I play a whole match with 4 players against a sixstack and lose 10 games in a row.
good to hear that matchmaking works for you 
1 Like
Well it’s also worth mentioning that people tend to exaggerate the hell out of everything…
Yesterday my team got stomped in a koth round 1…like it was bad…immediately the matchmaking cries starting going up in chat…smurf this…crap mmr…blah blah…
We went on to win the next 2 in similar fashion …
I think it can be a bit of a scapegoat at times simply because people don’t like losing…we need to blame something right?
I don’t think it’s “rigged” but I do think it has problems. I see way, way more loss streaks than win streaks with it, and that is not statistically possible unless something somewhere is being fudged, either by the players or by the system. I did a 500 match test about three years ago now, and it was about 48.8 to 50.2 wins to losses. What was interesting that looking at all of the matches in order, it was obvious that loss streaks were 1) more common than win streaks, and 2) outnumbered them easily by about 1.5 times the rate. That is not statistically normal. I still see it when I bother to track matches, no question something is not right in the matchmaking.
I am not exaggerating!1
i won these matches + 7 more very easily on that day
(https://imgur.com/a/iFPPFDw)
The games that followed 2 days later were hell, it was impossible to win.
Everything was red, not a single win out of 12 games. 
Player slots were left empty on my side, players were AFK, players insulted each other from the first second.
You should have been there, it was a blast!
If you feel like it, you can always join my games.
1 Like
Well…toxicity, throwers, leavers, cheaters are not things you can blame a MM for
Have you ever heard of Valve’s Rainbow of Trust?
I believe that major game companies started using such concepts to control the gaming experience of players a long time ago.
The official answer is that it matches you with players if similar skill but there’s a loud minority of people that don’t believe them when they say it.
1 Like
System already abused with smurfs. Bliz just don’t care.
1 Like