Exactly! Unfortunately, you view it wrongly, but regardless, what you just said is FORCED 50.
Darak’s View - You have a 50% chance to win any game
Real Logical View - You lost too many games (MMR dropped), we will place you on a team with higher MMR teammates (know the game, can help carry any mistakes), therefore you have a higher chance of winning the game - vice versa - You won too many games (MMR raised), we will place you on a team with a bunch of really low mmr teammates (don’t know the basics, only play quick match) have fun trying to carry this garbage.
Again, I will say, this is not about climbing, this is about just playing the game with people who understand it so that it is enjoyable and can be played correctly.
I played from Gold to Master and saw Bronzies in action as well (thanks, QM…). The sad truth is, that no matter how high you are on the ladder, those mistakes will happen. Just not as often. And it’s not because those ppl are not where they belong. I saw my Master friends and myself (also Master) to do those stupid things. We just make them less often.
Because humans can be tired, deconcentrated/defocused. They can have bad days/moments. So you having teammates who are “bad” might be due to ppl having their bad moments.
(And that’s why it’s impossible to make a system which can force any kind of winrate with 10 ppl.)
that would be the ideal goal of the matchmaker from my post earlier. its not the job of the matchmaker to force you to be at any particular win rate. however, the matchmaker will attempt to find players of equal skill and eventually this leads to, or near, 50% winrate since you are at a relatively balanced place.
If your mmr dropped it won’t place you with higher mmr teammates. It will place you with and against the same mmr players. If you are better than that mmr you should win that match.
If your mmr raised, it won’t put you with bad teammates to force you to carry or lose. It will put you with and against same mmr teammates. If you don’t belong there, you will probably be the one to lead the game to a loss.
Forced 50% doesn’t exist. It’s just an excuse that people use to justify their mistakes/losses.
“I lost so it’s my teammates fault/matchmaking”
“I won, I’m so good at this game”
Noooo, I have already heard that from Vision. That is not what I’m talking about. I am talking about, let me see if I can give some better examples…
So a game with Braxxis holdout starts, we have zag, diablo, bw, zuljin, jaina, the zag goes top and sure enough the diablo is behind him, the diablo stays top while multiple people are pinging him to join the bottom 4 man, he isn’t trolling, he just has no idea that zag should solo and him as the tank should be with the 4 man. This is not a “bad” day, this is a player that just doesn’t understand Braxis holdout and our team comp.
The current system imo mess up the functioning system of the MM because you can literally que a master player into plat matches with silver allies and utterly destroy the match.
It doesn’t prove Forced 50, but this isn’t a courthouse, and nobody has to prove beyond a reasonable doubt, all you have to do is use common sense. Yes people are in the wrong ranks and blizzard know who they are due to their poor play and they put the people that are doing well on their teams to force a loss. This isn’t on purpose, but it happens.
Blizzard knows nothing and don’t do that lol. Excuses excuses again to justify a loss.
What Blizzard (by blizzard = the matchmaking) knows is this: it has to create the most balanced match possible with 50% chances of winning for both sides. How does he do that? By matching players with the closest mmr possible together.
Now, if a player is placed wrong, the system can’t tell that.
No it does not match players with the closest mmr possible together. I think it takes an average, which is precisely the problem I am trying to get you to see. I wish it placed players with the closest mmr possible together. But, it takes the average MMR of the team (I think I read this).
I had a match where my team were all platinum 3 with everyone solo que’ing facing opponents where 2 opponents were diamond 2 gold and one plat 1
4 of the opponents were in a party.
So we faced a team that had massive cordination and experience and we lost that match so hard i forgot what game i was playing.
Lots of strawmen in here. OP never mentions “forced 50/50” win rate, yet most of the comments are mocking this argument. OP is saying the game is very streaky, so let’s examine that claim instead.
One point people are missing is that, while it’s difficult for the matchmaker to make even games, it is almost trivially easy to make lopsided matches. This means making streaks is easy for the matchmaker.
I see a lot of people asking “what does blizzard stand to gain from this?” Its simple: streaks make people more likely to play more games in one sitting. Win streaks make people want to keep playing and getting wins; losses make people keep playing because they are hungry for the next win. Less volatile matchups have less of a psychological effect on players.
Blizzard was one of the pioneers with WoW in studying how to psychologically manipulate players using addiction and gambling. I don’t think it’s too much of a stretch to at least suggest and discuss the possibility of win streaks used to incentivize long play sessions. The people blowing off the arguments come off as fanboys; the abundance of personal anecdotes and data gathering at least warrants a discussion.
It does feel like a forced 50/50 system. Im kind of tired of it to be honest. I play pretty much the same in every game i play, i make a few mistakes and also make a number of plays. I didnt go on a win spree per se, but i did go on a 25 game jaunt where i won 20 of them.
Now, im 5 for 5 in losses after 30 games and am hesitant to play again because i havent lost 5 in a row in a long time. The reason i lost most fequently was a bad draft. Too many squishy ranged dps like some combo of hanzo junkrat kt and liming on the same team. I also had a guy who randomly selected in the last pick slot and had no idea what he was doing with the character of choice.
People can say that ill probably win again, but im pretty sure algorithms are strong enough in todays age to calculate which member of the team will choose a wild pick. Put enough wild pickers on a team and you can obtain a loss because these people play whoever they want whenever they want. I normally oscillate between a few heroes so the match maker can algorithmically figure out what i am most likely to pick. I feel like im being purposefully placed with people who choose wild heroes sometimes. One game i was in, this guy chose Imperius even tho he was only level 3 (played 2 games) with the hero. These off-the-cuff players are incredibly toxic to the fabric of the game. Wins feel very smooth and the early incurred losses also felt smooth, but that the other team was just slightly better. Now, its blatantly clear im being sent back down the ladder for a little bit and my teammates are substandard of where i was before.
But that’s exactly what a perfect matchmaker would give you. 50% wins and 50% losses.
This is what I understand least about these people complaining about matchmaker while being up the “forced 50% winrate tree”.
And, scroll up to where I talked about Trueskill and Trueskill2.
Or I can link to the post once I get home and on a computer.
Judging a players “true skill” per Microsoft’s system in a shooter game like Halo, in a 4v4 takes 46 games normally, and up to 138 games in some cases.
Making it 5 players would increase that number… and that is before you look at the fact that Halo/multiplayer shooters generally don’t have vastly different characters.
So you might be looking at 50-150 games to judge you “true skill” at one character.
To use made up numbers, say that the MMR average of all players is 1500, with a standard deviation of 400.
If you’re queueing at say, 1900 MMR, it will likely at first:
-look for 9 other players within 1800-2000 MMR.
As the queue gets longer, that may expand, to 1700-2100 MMR. It is likely instead expanded once minor bumps.
But stating it just averages is a gross oversimplification and misrepresents how the system works.
There are a number of other factors.
What I have from a source is part of the reason they don’t reveal how the matchmaker rules work is because that causes people to alter their behavior, which causes the matchmaker even more trouble.
Like this:
This is all wrong. Because the level of skill difference is much lower. You’re more likely to get something like this:
1700 1800 1800 1750 1750 [1760]
Against
1900 1800 1700 1700 1700 [1760].
Or in an extreme case:
2000 1600 1600 1900 1600 [1740]
V.
1740 1740 1740 1740 1740 [1740]
Note how how the numbers as percentages are much closer than you example, this is because MMR/ranking/etc is far more nuanced.
Now, it might be 1.740 or 17.4, etc. but your example is hideous in how is misrepresents things.
No. Your example grossly misrepresents the skill differential, or rather the MMR of players.
Using unrealistic numbers to make it look far worse than it is.
My example is probably a bit too close in MMR, but, well, i typed it on my phone. Same with all my responeses in this thread except the original one about trueskill, and let me link it now:
And to get a more realistic wide range average here we go:
team 1 MMR: 1711, 1801, 1757, 1695, 1736 = 1740
versus
Team 2 MMR: 2015, 1500, 1596, 2005, 1584= 1740
Here we have an inbalance like you talked about earlier, although slightly differnet.
However, notice that the best player and the worst player are within ~500 MMR, or about 75% of the top players MMR. This is a far more fair way to describe an example than using a 1, 5 and 10.
This also puts the difference from the best player on team 1 to team 2 at about 200 MMR, or about 10-12%.
And the difference between each teams worst players at about 200 MMR, or about 11-13%.