Why Bug? Its normal loss why it should be Bug?
Did you read the bugs section of the original post?
I dont think so, you mean in this thread?
Yes.
20 characters…
But its something different, you are talking there about next match after game with leaver.
What happened to me wasnt bug, its how the game works. People just dont know that you can lose less by staying and not leaving like others. Thats why i said it here so you can maybe implement this in your post so more people know it. Its another reason not to leave games. Well, if you performed well and your mmr is not lower than SR 
. I didnt find this in your post, but its possible i missed that. I dont i ever experienced the Bug you are talking about btw.
It was for the diamond placed acc. I didn’t tryhard for the last 5 games since 5th game had 2 masters during placements. If went full try hard mod, I’m sure it’d place me at 35xx-37xx
Unfortunately, I need much clearer data (or a developer statement) to put a theory like this in the main post. There are too many random things that people say, and too many random things that can affect SR gain, for me to go on anecdote alone.
IF this is true it’s explained quite simply by Kaawumba’s guide. After six to eight games the MMR has risen substantially but the SR hasn’t risen at the same rate - it’s still chasing the MMR. Given that matchmaking uses MMR rather than SR the matches get harder faster than the SR rises. Your SR is still gold but you’re playing with and against much harder opponents.
Could you please link that. I can’t find it.
EDIT: If you’re referring to the post by Scott Mercer, then I agree with Kaawumba that it is simply sloppy language. There is so much evidence for MMR still existing. The fact that this one post doesn’t even mention MMR suggests he used “SR” interchangeably with / instead of “MMR” - possibly as a simplification for the layman or possibly just being lazy.
If Scott was referring to a change to the system then he most likely would have talked about both MMR and SR.
I have no way of knowing how fast MMR moves. I know that when MMR is more than 50 away from SR, SR gains on a victory will be significantly different than SR losses on a defeat. As this usually doesn’t happen, we do have verification that MMR and SR are close for most people.
But yes, one possible exception is bronze to GM type series, so you may be correct that MMR moves faster in this case. The long term effect will be zero though, because SR gets dragged behind it.
Yeah, the stated scenario was diamond/master player playing on a silver account. I consider that to be the same as bronze to GM…except it starts at silver.
Also note that it was a big “if” I began with. I’m not convinced the story is realistic in the first place.
What do you mean by “theory”. Both things i presented can be tested. That thing with leaver on enemy side can be tested anytime when enemy team has leaver, just try it. I wish i recorded other thing because i really didnt expect you would not know it, i just thought you forget to implement it.
Btw you had no problem to write there theory about how are people matched in games by mmr, while blizzard said people are matched by SR.
I will record both these things if you want next time i will have them. And i already have like 500 recorded games on my Channel, you can see pretty much in every game that SR of both teams is practice ly same so i dont know from where is your theory about matchmaking by mmr coming from. With exception of placement for sure, placement. Placement matchmaking is working with mmr.
For the first one, I’d have to see enough data for it to be significantly significant. It could take 20+ games for this to be the case. And all the data has to be written down. If I believed every time someone said they saw something, my post would be a bunch of random contradictory statements about how things worked.
For the second one, I guess I could test it. But why would anyone want to leave a game when they are are up a player? Even if true, it’s pretty irrelevant to how anybody plays the game.
MMR is usually close to SR, so if matching is done by MMR, then SR will be close as well.
The one time that the developers have confirmed that MMR and SR are not close is decay. In this case, the SR of the two teams are not close. See Streamer Data - Google Sheets → Seagull 10, Team SR compared to Enemy SR.
Blizzard has also said that matching by ability is by MMR only. My statement on the topic is in the original post → “But I just read this post from a developer, and it said matchmaking is based on SR, not MMR”.
The system blizzard has bellow Diamond is there to help people climb and get rid of smurfs as fast as possible. So placing people with similar mmr in game doesnt make any sense. I have done some runs in past on accounts placed lower. I pretty much spot another smurf/alt instantly in lower ranks. And in majority of games there was nobody on enemy team to balance me while my mmr was probably 1000 + higher than SR. How would alt accounts climb so fast if there would be someone same as you in term of mmr on enemy team every time? I have similar experience in low plat now. Sometimes i have games when we are getting steamrolled because enemy team has much better players, basicly all 6 of them. Its just happen because its luck based or random you can say. Or another example. I had many games where we had new accounts already placed in same rank. Once enemy team had 3 smurfs from high elo with level 27+ while we had same amount of people who were new to the game. There was insane mmr gap in teams. So there is no way system is working with mmr instead of actualy SR outside of placement.
System just find another 11 players and place you in game in a way where teams would have same SR, mmr is Irelevant for matchmaker if he has SR. Mmr is used there for SR gain or loss, . Just like blizzard said in past. And its my experience too. Lvl 2500+ in here
Below diamond, it is impossible to have MMR 1000 above SR. The only way this can happen is severe decay, such as happened with Seagull in season 10.
You seem to be under the expectation that MMR is a very accurate representation of your skill, even on a new, recently reactivated, or shared account. This is not the case.
If you have a diamond player, who dominates a few bronze games, the matchmaker does not know that he is a diamond player. It only knows that he is better than the people he is playing against. So the game raises his MMR a significant (but not giant) amount and he plays another match against and with players of his new MMR. If he again dominates, the game continues to push up his MMR and he continues to play higher opponents, until he stops dominating.
This happened here: Streamer Data - Google Sheets → The500IQ 14. SR is going up 50-100 per game, and MMR is moving up a similar amount. If MMR was moving up much faster than SR, then team SR would be a lot lower than enemy SR (like happened during Seagull’s recovery from decay).
Once The 500IQ started having trouble in plat, his MMR and SR stopped shooting up (even though a GM is playing the account).
You seem to be accusing me of believing Cuthbert’s handicapping theory. I do not. Read the second post in his thread (by me). Again, once you acknowledge that MMR is not accurate on new, recently reactivated, or shared accounts, things will get much clearer.
1 Like
I was wondering if Kaawumba or anyone else knows if PBSR is used when you are still ranked diamond but your SR is below 3000. You have to be below 3000 for five matches before you’re put in plat, so during those matches is PBSR used?
It doesn’t really affect much, I’m just curious.
My humble contribution to the topic focused on Placement Matches:
Thanks again Kaawumba for your help!
1 Like
My data doesn’t have a clear answer.
In Overwatch Forums and Groups and Matchmaking in Overwatch Scott describes it as being tier (plat/diamond) not SR (3000+) that is the relevant quantity. But Scott isn’t always super precise in the way that he says things, so take this with a grain of salt.
As promised, 100 games worth data.
https://docs.google.com/spreadsheets/d/1IDNzfQU5XOgwAfC7YF_uHh5KKVpjRW3rAW6YRZteMwM/edit?usp=sharing
To add to that, even though it’s not super reliable data, for my Pharah and my Orisa, my individual performance has always been way over the average player of my rank according to, Oversumo, Omnicmeta and Overbuff…
Thanks. Your average gain on win is 23.10 +/- 0.49 and loss on defeat is -21.78 +/- 0.45, so this is a bit below the expected value of +/- 24. Usually when this happens I blame performance metrics, because that is the most opaque part of the system, but it could be something else.
Okay, but we don’t actually know how performance metrics are calculated to influence SR, so it’s really hard to make statements about what sort of changes we should expect based on a given performance.