sometimes I lost one game because dc or I cant 1vs4, but still lost 100+mmr, how could that happened?
Because it doesn’t matter whether or not you disconnect. If you lost fewer points on a didisconnect, everyone would DC when they’re losing.
MMR gains and losses are based on the MMRs of the other players in the game. Exactly how it works isn’t really known but this is the only rational metric to use.
that’s 100% wrong, I know dc or first leaver lost more mmr. At least for my account it is. Yesterday I win one game (my team 6k,2900,4k+,4k+) against 4at (7W0L 5k+), I got 31 mmr. And I lost one games (my team 6k,4k+,4k+,3500) against 4x 4k+, i lost 70mmr. The only reason is I was the third to leave not the last. My feeling is that the first one to leave or DC will definitely drop more points (even if his team wins in the end), the second or third to leave may drop more points
it’s 100% correct. The order in which you leave is irrelevant. The match processing isn’t completed until the game is over.
You’re going to see different numbers every time because every game has a different combination of MMRs amongst the teams.
This isn’t something you can analyze by eyeballing. You need to record the MMRs of all players and the mmr change over many games, you can’t just look at it and go “oh it works this way.” Provide evidence.
It is possible that match length itself may be a factor, though I don’t have enough data to say if that’s true or not so I will refrain from doing so.
Obviously you need to play more games. As a 6k+ player of all races, I have encountered DC at least 10 times, and a lot of points will be deducted each time, even including me and 3 players with 4k points playing 3 at close to 6k. In the history -0, points are actually deducted close 20-30. The first player leave game and DC will definitely deduct more points.
OK, let’s break this down.
I’ve played hundreds of games in each ranked season so far, as well as probably at least 1000 before they restored the feature.
This is irrelevant information, Neither your MMR nor what race you play has anything to do with disconnects or how much MMR you gain or lose (Each race is tracked seperately anyway, with random counting as a race)
This isn’t even that many, but I have never had a disconnect that wasn’t the result of my own internet connection failing or battle.net getting DDoS’d. Which was like twice since the ladder reset, over like 100+ games.
The number of points you gain or lose depends on your MMR compared against the MMR of the opponents. You have no evidence that the first person to leave or disconnect has a different rate of point gain or loss. The game has no system to punish leavers or disconnects (and quite frankly, it should, but since the game doesn’t distinguish between the two, it can’t). The game doesn’t know when you disconnect. If you DC, it’s the same as hitting quit on the menu. Further, ATs and RTs match together, and so this simply can’t work, because an AT has one MMR for the team, so there’s no way to penalize one person for DC’ing or leaving before the others, and it wouldn’t be fair to treat the RT players differently with regards to leaving or DC’ing.
If you want to convince anyone of anything else, you need to collect data. Keep track of the MMRs of everyone you play with (screenshot the loading screens). Check everyone’s profiles after the game (assuming the server reports the match in a timely manner, which it doesn’t always…) present the results of several games and caclulate the changes in MMR for each player in those games.
Without doing this, you don’t have a case, because you’re just eyeballing it. “it seems like i lose more if i DC.”
Like i said before, It honestly should take more points away if you quit early or DC, but I assure you that it doesn’t.
lol forum cant put links or picture,
so i lost 157 mmr against teddstar at must because he get only 6900mmr:)
If you put the link in Preformatted text (The symbol that looks like </> above the area where you write) it will let you post it just fine. Doesn’t matter where you link it.
www.google.com Like so.
`www.google.com` <-- This is what it looks like when you type it in your message box.
https://tieba.baidu.com/p/8537080497?pid=148264595687&cid=0#148264595687
-157 against teddstar and his at:)
I told you, we should NOT put AT vs RT.
In any circumstances. If AT will not find games , than they must be patient.
If they are 3 in AT then they should go 3v3 not 4v4 … Or if they get a 4 th player than wait
to get a game.
Like 4x Beach volleyball, once a local champion woman team put a prize (eg 1000 euro for the other random team - if defeats them), so Random people tried to make a team, even athletic ones also mens, but every of them lost… This was going for 7days when I spent my holidays on Black sea, like 5 games / day. Is just not fair !
AT vs RT is ok, I beat mass AT, some 5k-5.5k AT is not so pro.
But teddster(6955), changer (6189) are high mmr player and all other player in my team are 4k noobs, should be a game -0, but -157 really?
on this rare occasion, I honestly agree. But it’s not even the worst matchmaking issue, as the game, especially in 2v2, is rigged to help keep winners winning and ensure that losers can never come back even if they improve.
At a bare minimum, ATs need to be matched against stronger opponents- i.e.e treat them as if they were like a couple hunded MMR points higher than they actually are.