DBD's MMR isn't really MMR
I don't know if anyone else has noticed but in solo queues they match 2 good players with 2 bad ones (4 man SWFs aren't affected)
Comments
-
In just the last day I played with someone who had over 4k hours and someone with like 360. People play sweatier because of mmr and it's for zero reason honestly.
3 -
I think the really high amount of survivors that suicided/disconnected since 6.1.0 and the resulting stomp match just made every survivor lose mmr points. Like a lot of mmr points.
7 -
But like, it is
9/10 games have a ######### MMR system
2 -
I mean... the soft cap for mmr is so easily reached that basically everyone ist "high mmr" and so of course mmr does not work xD Every point above the soft cap does not matter for the most part... So what did you expect to happen?
4 -
How do you know? Where do you get this info?
2 -
Dataminers figured that out months ago, I'm not 100% sure on the exact numbers, but as far as I remember the soft cap for mmr is 1600, even cheater bots that win every game don't get past 3000 and many streamers that are actually good at the game are around 2200/2000. You do not start at 0 mmr but 650. And there is a little bit more to that, if you want the full information Scott Jund did a video on it called "Top MMR doesn't really exist according to New Data" , in the video description is the link to a guy who refers directly to what the data miners found out. I think Scott only refers to the video of that guy.
3 -
-body is one character too short
2 -
What does that mean?
0 -
Of course. People shouldn't be surprised it's not true MMR. You realistically don't have to do ANYTHING in a match, as long as you walk through the exit at the end! Lol
3 -
Gotcha thanks man
0 -
While those numbers are the latest available from dataminers, it's also been confirmed that the number of people over the softcap is way less than even 25% of all players. Keep in mind that the 2000-2200 MMR range of streamers are for people that literally play the game for a living, with comp swf's only hitting 2400-2500 MMR
https://forum.deadbydaylight.com/en/discussion/comment/3189257#Comment_3189257
2 -
I think its because the less Survivor playing made the MMR bracket so wide it doesnt matter anymore, dataminers already found the softcap is very low and when there are queues it broadens the pool a lot to shorten them so now it just picks wathever is there.
This was happening already before 6.1 but now is even more noticeable because queues increased.
1 -
that's why people with 200 hours and 6000 hours can be in one team and the 6000 hours guy goes down in 15 seconds
0 -
MMR now basically works the same as the old rank system. I don't think that's a bad thing necessarily. Some games are super sweaty while others are more casual. It's definitely less predictable than before and my games have overall been more enjoyable.
1 -
Can I refer everyone back to the subject title that DBD's MMR isn't really MMR as most people hear seem to be agreeing with this
0 -
Do you have any number of "less" survivors playing? Like real numbers? Just because there is a 100% bonus there, doesn´t mean that everyone suddenly stopped playing survivor.
0 -
I don’t think it’s realistically possible to have a good MMR system in this game. How can that function when both sides can technically win at the same time?
0 -
No, thats why I put "I think" instead of "its because" but a lot of people on the forums, reddit and generals stated they stopped playing Survivor after 6.1 and the queues did increase in lenght after 6.1 (before they were around 3 minutes, after they were around 5) and it is known (as BHVR confirmed this) the longer the queue the broader the pool becomes to find a match until it reaches a point where it just throws in the same lobby whoever is available disregarding MMR entirely.
I could ask you if "do you have any numbers on the "high" amount of DCs and suicides? like real numbers? Just because people complaining in the forum they saw an increase doesnt mean that everyone suddenly started doing it" but you also put "I think" so I know its just an educated guess like mine instead of someone stating a fact.
Now, do you want to quote me so we can start an absurd debate where nobody can prove if they are right or wrong or shall we leave it here and assume each of us has a different and impossible to prove theory?
0 -
How do people know how many hours someone has played? Is this something pc exclusive where you can get this data on someone?
0 -
In Steam you can check gameplay time if they allow their profile to show it.
0 -
Today I played with one Prestige 99 and two prestige 100 players (each one in a seperate game). I know that prestige doesn't mean anything skill wise, but someone who brought their favorite character above 50 shows a level of dedication that should equal at least a modicum of skill.
Well, all three high prestige survivors went down as easy or hard as the rest of their team, I felt no difference at all. I know that prestige ranks can be cheated, but nothing at this players made me doubt that they were legit. At the end of the day it all depends on your skills vs theirs.
0 -
That information from that video is off. For one thing they said MMR goes up or down a flat amount per game which was wrong, the devs already confirmed that (like most MMR systems) how much the MMR goes up or down depends in part on the relative rating of the individual players. If you kill a lower rated opponent your MMR goes up less than if you kill a higher rated opponent and vice versa.
Also the devs confirmed a while ago there was maybe 1% of people over the old soft cap and the distribution is a bell curve. From that we roughly estimated that the curve is a mean of 1000 and SD 400, and from that you can guess that the number of people over the 1600 soft cap is about 6%.
0 -
There was quite a bit of misinformation in that video. MMR gains and losses aren't static. They adjust relative to the opponent. https://en.m.wikipedia.org/wiki/Glicko_rating_system
They thought it was static because they used dummy accounts with the same rating. Someone handy enough could check their MMR right now and see a prime number, which renders half of Scott's video false.
1 -
Yeah on pc you can check someones hours through steam, granted their profiles arent private. So if someone was exceptionally good or bad I check profiles for hours if they were on steam
0 -
People have datamined how the system works since like week 2 of its implementation. Plus Bhvr has talked about it on their own dev streams.
0 -
People dodging lobbies also creates odd situations. I particularly like when I queue up and am immediately inserted into a lobby where everyone is ready to go. It's like "I'm likely to be either the best player or worst player in this match, I guess". And the matches seem to bear that out.
Not that it's necessarily indicative of skill, but another example is when I was in a lobby the other night will 3 other survs who were all P40+, and they all bailed at once and were replaced with P0s and P1s. The match was horrible.
But yeah, matchmaking in general is kind of a crapshoot, and I think it's gotten a lot worse since the 6.1.0 patch dropped. Trading queue time for accuracy, I guess.
1 -
High MMR duos/trios getting lower MMR teammates is something I've been noticing for quite a while myself, I do believe there has to be something in place where the matchmaker is trying to "even out the odds". Obviously I don't know for sure, but it's baffling how consistently it happens.
But in general, the MMR cap is really the culprit ruining matchmaking at the more extreme ends. It's like old red ranks, there's a pooling of players at the cap with significant disparities in skill and experience and performances. I don't believe that with how low the cap is (1600MMR, when the high MMR bracket spans 2000-3000MMR) that really only ~5% of the playerbase are at that cap. But even if that is the case, it only goes to show that even among those 5% the differences are still huge and the cap needs to be changed. I think experienced and skilled players wouldn't mind waiting a minute or two longer for a chance to get more competitive matches. I mean, maybe I am wrong and even of the advanced players most just want to play and win all the time, but there's obviously also players on the receiving end of that, so even in terms of player retention I'm not sure it's the right route even if it were the case that players rather want to win against worse opponents than compete against equals. Either way, I don't get why they don't make it so that the matchmaker looks for matches as close in rating as possible and gradually expands that range, rather than have a cap. If there really are so few players at the extreme MMR ends, just make the range expand faster the longer you wait.
Backfill does seem to also still be an issue, although I don't think the matchmaker entirely ignores MMR on backfills anymore.
0 -
Ok i think killers stopped playing.
0
