"Forced 50% Winrate" Is Real - And It's YOUR Fault

I remember vehemently disagreeing with that the first time I heard it and it sure hasn’t aged well.

Purposely rigging matches to get you to play longer is pretty damned corrupt in a game with this kind of monetization model, especially so when there is so much variance in the gameplay already that lends itself to the peaks and valleys they are artifically creating.

…is or was using some weird match making algorithm that matches teams with similar team mmr against each other.

at least it used to be this way. but the way team mmr is calculated can be pretty messed up.

  • first of all, it tried to even out the mmr of your team to your current rank. (lets say plat 1 mmr is 2200. so it tried to create teams with an average of 2200 mmr from various players with higher and lower mmr)
  • your mmr can rise or fall faster than your current rank. you can be plat 1 with diamond 1 mmr. this mostly happens when you go on win streaks
  • the matchmaking algorithm compensated your high mmr for your rank by putting lower and lower mmr players on your team as you kept winning, to still average a 2200 team mmr to create a “plat 1” game

this created games where the more you went on win streaks, worse and worse players would be put on your team against a usually evened out enemy team. this put basically all the pressure to carry games on your shoulders, and created situations where your team got regularly destroyed in lane and you were the one supposed to win the game despite the odds.

there were limits to how far people could be apart in mmr to be matched in the same team, and this was pretty much the breaking point you had to reach to break out of this matchmaking and get equally skilled teammates and finally climb.

Of course it’s forced.
Whenever i am at lower LP (from 0 to 80) i get teams of all types, good, sometimes bad, most of the time average. And it depends a lot on how i play it.
BUT whenever i come to one game from promos or in the promos itself. I must by default get literally inting team.
Dont let them fool you. Everything is tracked. If they can show you at the end of the season how many times you died to Yasuo and how many times you killed a Yasuo. Trust me they know everything.
They know which accounts have tendency to disconnect and go afk.
They know lose streaks and win streaks and everything that matters.
And when you are near to promote to next division you will get these people from elo hell in your team if possible.
The system is not designed to help you climb. They dont give a damn.
The system is designed to annoy you and to stop you from climbing so you get tilted and play even more.
Thats all thats important to them. That we play as much as possible.
Not our climb, not our mental health. Nothing. Only play time.

I see a lot “Oh, match making wants me to lose this game because I’ve won some”.

No. Your opponents get better, your teammates get better. At the same rate.

Why do you have to play against plats/dia when you’re gold? Because you have high plat/dia MMR and so those are the opponents the system thinks you can beat. You’ll climb in that situation with a 50% winrate because that’s how LP/gains losses work. Or are at least supposed to work, bugs with positional matchmaking LP gains notwithstanding.

Match-making is trying to put you in games at your skill level. You’ll probably find that a lot of you have games just like the ones you complain about your team having, or are making other critical errors in your gameplay and just don’t want to admit that the idiot that is 0/10 at 10 minutes into Nasus is the same skill level as you.

MMR isn’t perfect and could definitely do with being visible . Invisible match-making systems make people suspicious - especially in team-based games when their teammates don’t do well every game.

… themselves said that they will make it so that both teams will have 50% win chance. Which means if someone has more than 50% win rate will be matched with someone who has less than 50%.

This is the reason why sometimes people have the most ridiculous win and lose streaks one after another. This doesn’t happened to me in any other moba or multiplayer game.

When I play good. And manage to get about 10-15 wins in a row. the mechnism comes into play. and since the past 3 days. All I got were sad team mates. 2 wins and 14 losses.
just so that I get the 50 percent win rate.
The matchmaking should be as per to the experince but within that, random.
but it seems more like moment u cross 50-55 percent win rate, suddenly, all you get are terrible team mates. gg valve.

Good that you understand that and protest against this corrupt matchmaking system. It is just a lottery. You will have more luck or chance at the casino than at [this game]. Oh, and those who [spent money] I bet they get a better % so they will feel like they are actually improving, but its just how they are matchmaked.

People who don’t believe it’s rigged - why don’t [company] make their MM system completley transparent and open? Then there would be 0 discussion. There is nothing secret about a MM system.

For those that can’t tell, these are comments from dota 2 and LoL regarding the perception of “Forced 50%” matching alongside the lack of visible data and company transparency. I edited a few choice words to remove game names or companies mentioned, but the point here is pointing out all the magical repetition that comes up from similar games for similar complaints, faults, and claims that such things don’t happen in other games cuz the system is rigged.

For reference: my choice of quotes are fairly recent, so I’m not plugging quotes from 2012 or whatnot even though similar complaints have plagued these games for years.

If you’re one of the peeps that likes to associate MMR with “skill” or be under the impression that ‘match makers’ should match people of similar “skill” then you’re probably misunderstanding what “MMR” actually means, and how it is applied to these match makers.

MMR, like the ELO system, is an appraisal of probability, not skill. Having a higher mmr against a lower one would shift the probability of victory for one over the other (generally phrased in the context of one winning out of [y] games) but doesn’t actually convey any sort of relation to “skill”. The ‘goal’ of a system is to match players so the ‘probability’ of outcomes between two sets would have either side capable of winning or losing similar sets of games against each other.

ELO-based metrics have three key flaws:

  1. these are multi-player games, not 1v1
  2. match games aren’t played out in a series of multiple games to see the effect of the probability realized beyond just one instance.
  3. a pre-determined starting point despite players having unknown ‘skill’ levels for initial seeding.

win/lose metrics tend to not weigh who managed to provide worthwhile contributions over others, or discern from an otherwise dominate side/player suddenly being upset at just the right moment to shift victory the other way.

These systems just compare effective probabilities so the result can end up being “chance” both on what ally/enemy competencies end up grouped together, and what sort of value distribution is giving at key times to establish the initial rating: getting ‘lucky’ wins for favorable mmr spreads will be far better than loses at unfavorable spreads being detrimental. Or similarly, bad matches are initial seeding will make a climb significantly harder than bolstered estimations.

4 Likes

Had a game today. I looked at all my team. We were all on win streaks.

My team:

Plat 1
Gold 3
Silver 1
Silver 3
Silver 4

Them:

Plat 1
Plat 4
Plat 5
Bronze 1
Unplaced

Then we lost