It’s not a thing, you’re just really bad at understanding statistics and how MMR/SR matchmaking works.
3 Likes
Ask yourself another question, why does bronze(8%), diamond(12%), masters(4%), GM(1-2%) aka 25% of the population take up 3500 out of a 5000 sr range?(outdated statistic, don’t even have a link anymore but nothing was fundamentally changed about competitive since).
Bronze players’ variation in skill needs 1500 SR to seperate but somehow 75% of the playerbase can be “placed correctly” within an equal 1500 sr range…?
does it matter?
it’s a normal distribution, and where they decide to delimit the ranks is completely arbitrary
1 Like
It’s not a normal distribution.
To start with show me a single game which starts you off in the middle of the pack, has placement matches that DO not matter after the initial ones, puts 75% of players in an equal range as 8% of the population(bronze)
None of it makes sense, it never has.
That’s why matches from silver to platinum are always a tossup, the skill variance is ridiculous.
Not just when you go higher but from match to match can be an astounding difference.
Hell, it’s been seven years and still leaving during a placement match will just make that match not appear in your 5 placement matches.
They didn’t care then, they might now, hence why I’m saying something.
I suggest you come up with a response that actually matters if you wish to stand in my way.
fresh accounts are easier to rank up than accounts with a lot of playtime, except if the play-style is so completely different that blizz can’t estimate it anymore. Lets say hardstuck mercy gives his/her acc to ML7.
I have multiple accounts in master and one account stuck in low diamond. I don’t play differently on that account. I would not mind it being in diamond. Fact is that the older your account gets and the more information it has about your play-style and therefore the harder it will be to climb.
Smurfs are also a factor that shake how good matchmaking can be. Lets say you always team up with 4 friends. You are not smurfing. You have the chance of having potentially 2 smurfs. The enemy could have an assembly of 6 smurfs or even worse a platin mercy getting boosted by the smurf teammates. Guess who will win.
Blizzard is very open about the fact that teaming up makes matches harder, but I think the higher the amount of smurfs is the less attractive it makes teaming up.
I honestly think that there are often matches in platin that could as well be in diamond just because each team has 3 smurfs which makes ranking up in that elo much more inconsistent.
One game you have to go completely aggressive and play more risky because its a platin game and otherwise nothing gets done. Another game you have to play more passive and strategic because it has a bunch of smurfs where its hard to estimate what actual rank they are and gameplay needs to be adjusted.
Therefore the higher your rank is the easier it gets often. I swear to god, Master games are easier to me than Diamond games, just because my team is pretty good as well and smurfs start to have less impact unless they are top 500.
It literally does? Its been tested and proved a loong time ago.
you could try hard for several matches and still lose, and then straight up int and still win.
Basically, you can consistently win and lose at a fairly same rate
of course it is, the central limit theorem applies
I would link it but this excellent forum prevents me
google “overwatch rank distribution” instead
No it doesnt, my win rate isnt 50%
What it actually does is ATTEMPT to make every match between a 40-60% chance for both sides. You could be anywhere in that range or even outside of it too
Its not hard forcing 50 on everyone, im not at 50 myself
Sr isnt a perfect measurement of skill. Its a point system tracking wins and losses in comp
The difference between 500 and 1500 is not as big as 3k and 4k
2 teams of equal size one loses one wins. No matter what 50% of the players in any given lobby will lose. Its not forced its just common sense.
1 Like
The system is a modified ELO system. As you get to the ends of the distribution scale, it becomes harder and harder to reach the end because normalization effects try to pull you to the center. Really, the only reason it looks unequal is because from 0-1500, the ranks aren’t separated except by the <500 band. Bronze at 8% vs. Master+ are both 1500 SR ranges, and should be roughly equal in a PERFECT distribution curve.
Where the extra 3% that are bronzes comes from is a combination of people deliberately keeping their ranks low and people leaving the system at a low point, but still being counted.
There’s no magical formula designed to keep some people low or preventing them from ranking up. The only formulas used are trying to balance team MMR to produce a roughly equal winning chance, and given how variable everyone’s performance is game to game, it’s never going to be 100% accurate. The closest analogue is team chess - they could form teams with roughly equal ratings across boards, but if Billy is playing well that day and Bobby isn’t on the other team, it feels like a slaughter even though the best numbers you have place them roughly equal.
1 Like