Visit the Kill Switch Master List for more information on these and other current known issues: https://forums.bhvr.com/dead-by-daylight/kb/articles/299-kill-switch-master-list
We encourage you to be as honest as possible in letting us know how you feel about the game. The information and answers provided are anonymous, not shared with any third-party, and will not be used for purposes other than survey analysis.
Access the survey HERE!
Why MMR tho?
This game tracks your altruism from unhooking to healing,
boldness from chasing, escaping chases to cleansing totems,
survival from self healing to struggling, wiggling and surviving,
objective from gens and exit gate.
You get rewarded for that,
you also get rewarded for dying as Obsession +1000.
And the matchmaking only cares about surviving/dying, and killing/not killing,
eh....
Comments
-
I guess people wanna have unballanced matches..
3 -
Because the MMR is actually more accurate than the old rank system that you’re talking about.
10 -
The win is all that matters in a practical sense. And it is easier to design.
0 -
No it's not? Tell me why is that you think outcome>performance
2 -
Because those things are great at giving immediate feedback, but get unwieldy if they have to function as a matchmaker. This is provable, because the old matchmaking sucked.
The reason MMR systems only look at win/loss for the most part is because over a large group of games, winrate is the most universally applicable proxy. You aren't punished for playstyle, different characters with different abilities (killers in this case) aren't penalised based on what their abilities are designed to do, and it's hard to continually get false positives-- and if it isn't hard to do that, that's a balance problem, not a matchmaking problem.
TL;DR because it works better to do it this way in most cases.
5 -
It is more accurate if you are new to the game or not very skilled. The whole reason they changed to MMR was to keep new players from being swashed until they learned the mechanics of the game. Most players that get steamrolled early wont stick with the game and they prioritized those players over those that have been longtime users.
2 -
It's easier for BHVR to pretend that this system is working, as opposed to admitting that the matchmaking has become significantly worse than back with the emblem system. And with a slowly decreasing playerbase, they are going to keep on expanding the range of the matchmaking leading to even more and more unbalanced games. SWF it up or leave it at this point.
2 -
Because completing tasks out of context doesn't equate to skill. It shows that you *can* do those things, but you need to know when you *should* do things in the greater context of the trial. If player A and player B both score consistently the same on emblems, but player B escapes at a 30% higher rate, then player B is probably more skilled.
5 -
MMR sounds very cool in theory...
0 -
Because tracking all of that stuff is probably hard work, and these devs onky care about quick and easy solutions to problems.
Much easier to track kills/escapes than anything else. Just like hockey, they onky track the scores a player makes right?
1 -
For a few reasons, including:
- The sole purpose of MMR or rank is to make it so players of equal rating have reasonably close escape/kill results at the end. Trying to measure everything else under the sun just dilutes away from what you're actually trying to measure, namely the kills and escapes. Too many parameters does not necessarily make a rating system "more accurate", it just makes it more complicated.
- When matches involve players who are all in the same MMR bracket the results are normally close. This is evidenced by the experiments the devs did where, when the MMR brackets were severely tightened, the consensus was the players were all closely matched (or, as some people said, "I'm only facing the same sweaty survivors and having to try hard to win". Which implies it was a tough, close match for the killer.) When the brackets were loosened the resulting matchups were much more chaotic.
- Anecdotally my matches are noticeably better now than under the old Rank system, especially with the rank resets screwing up the rankings every month.
- Also note than every killer character in your roster has its own independent MMR, something that doesn't happen with the emblem system.
4 -
Absolutely not true, as shown by how a bunch of high skill players now complain that they have to "sweat to win". The obvious reason they have to try hard to win is because they're being matched against equally high skill players due to the MMR, something that didn't happen as often in the old system.
5 -
MMR is the worst matchmaking to give to any team based game,
What is this "more accurate" thing you're talking about?
SWF it up or leave it at this point.
Sounds like a very healthy game when your option is to either suffer or have friends that play at the same time you play.....
If player B escapes at higher rate than the player B will get more escape emblems, thus climbing ranked ladder faster, thus matchmaking is working..
Hard work, yet it used to be like that,
yet games like CS:GO have GLYCKO2 instead of ELO...
Emblem could have included ranks that didn't reset,
system could have been heavily rewarding wins, but also taking into consideration when 1 out of 4 players did basicaly all the work,
etc. etc.
I have to sweat to win so hard as survivor, because one DCs, the other gets downed 5 seconds into chase and the 3rd is discovering lockers for the first time.
Meanwhile, I switch to killer, play Michael Myers with 0 Perks,
boom, easy 4k, no pressure, wasn't even sweating or tryharding.
Yeah seems that the MMR is working guys...
1 -
Seems to work out alright for games like Overwatch, League of Legends, Smite, Rocket League, etc?
1 -
Seems
alright
0 -
Escape emblems, sure. But what if they only did gens that game because that's what the team needed? No altruism, no boldness, no chaser. Are they worse? What if they played around their emblems more but then the gens didn't get done? Is that somehow more skilled? The devs have run tests and looked at their data before. There's little to no correlation between high emblem scores and high escape/kill rates
3 -
It doesn't in Overwatch. Its matchmaking is absolute garbage.
I can't speak for the others you mentioned. I don't play them.
1 -
I would like to quickly cut in and say that those games are not a good comparison to DbD at all.
The big difference is, that all the mentioned games are symmetrical games, whereas DbD is an asymmetrical game.
This means, that in a game like Overwatch you get into a 6v6 matchup, where every player has about the same strength level. Therefore in any form of 1v1 engagement the player who outplayed their opponent, so the player with more skill, gets to live while the other player dies, meaning tracking for kills / deaths alone is an accurate indicator of that individuals skill level.
In DbD we do not have this kind of thing. What we have is a 1v4 matchup, meaning one player has to hold a rather significantly higher strength level than one individual from the other side. In this scenario the one player playing Killer will defeat any of the other 4 players in a 1v1 engagement not because they are more skilled than they are, but because they are heavily favored by the game and its mechanics. Just looking at who lives and who died in this case is therefore not a good measurment of skill, at least not for an individual Survivor. It could work if it was changed to judge the entire Survivor team based on the overall outcome like it does with the Killer (to allow a victory even for that one individual that got hardfocussed and stood no chance but therefore allowed everyone else to escape), but that is not how it works right now.
Now of course we could balance DbD around the 1v1 interactions, as the Devs have been trying to. In this case seeing who lives and who dies from the Survivor side would actually be a decent way to determine their skill level, however thisno longer works for Killers who are judged soley on kills. You see, if we balanced this game around the 1v1 interactions and made the determining factor for a hit be whoever outplays the opponent, then the Killer would struggle immensly against the Survivor team due to being outmanned as heavily as they are. Killers would end up getting obliterated by teams that are even just somewhat close to their skill level, meaning now this MMR system would put the Killer up against a bunch of much worse Survivors, because those are the only ones the Killer can achieve a victory / draw against in this system.
In order for DbDs MMR system to be comparable to games like Overwatch, LoL, etc. we would need to change the MMR system first to actually judge both sides on the same basis, meaning Survivors would need to be judged as an overall team rather than four individuals.
0 -
They're not directly comparable, no, I wasn't comparing them to DBD- but they are team games, which was what MMR is supposedly a terrible fit for.
Though to be fair, Overwatch and other team games don't measure kills/deaths, they measure wins/losses. DBD has its win condition directly tied to that, but other team games don't, it's all about tracking who wins.
As for the 1v4, I think that it's not actually as insurmountable as you're making it out to be. If every survivor goes in with an expected 50% chance of winning, that means that the killer's expected outcome is a draw, and the actual result is actually dependant on the specific actions taken within that trial instead of one side being massively favoured.
0 -
Just FYI one of the things the devs mentioned in that big MMR blog post was they actually are working on tweaking MMR so that survivors go up or down as a group versus as individuals. That way someone who sacrifices themselves to save everybody else still can have their MMR go up.
0 -
Yeah idk why they added mmr when teammates isnt the only RNG thing in the match.
1 -
The match making has always been a joke in this game. That's the real problem. Not gen speed, not camping and tunneling. It's the garbage match making.
2 -
Rank based matchmaking was way better, sbmm is just causing frustrating games, sweatfest, cheaters, sub cheaters, stale meta that consist in facing the same builds and killers over and over again.
I had 7000h back in September when sbmm came out, today at the time I'm writing this post i have 7100h, i think this says a lot.
1 -
You could just straight up ask, What about current Grades?
Some games, I repair basicaly all gens in the lobby, get some saves and heals,
barely see the killer, and dodge him stealthily, using stealth.
And get 0 pips, only safe pip, but no progression...
Nah, matchmaking based on outcome in team based games is just cringe,
sure it works, but does it work perfectly? no.
Yeah, exactly, and don't forget just because other games use ELO doesn't mean it's good..
And what actions are you talking about good sir?
The individual or team reliant?
I just played one game where I almost died as I bodyblocked to the point where they finaly wiggled out, I got 2nd stage hook,
and with great efforts, and more gens done, even pickups in the powered exit gates phase while injured,
I escaped through hatch, leading in draw,
I could have went through gate but you see the killer was kind of standing inside it and the other was closed....
What else is RNG, because I don't think maps are RNG.
Don't forget camping and tunneling killers.
0 -
Both. Individual good play can lead to escaping, and good team play can lead to escaping.
In fact, I'd go so far as to say that at a certain level of skill, both are required for escaping.
That's just for survivor, of course, for killer it'll be both different and much simpler.
0 -
Oh another one of these threads.
A couple of months ago they turned off MMR for two days and used whatever grades people had acquired throughout the month. The matchmaking became so bad during those two days that Otz, a well known streamer who does kill streaks, immediately won 25 games in a row as Artist where he mostly didn't have to try. If you consider what this means, it means that he, not intentionally, bullied baby survivors for 25 games for a Youtube video.
Again, he didn't do this intentionally. But, that was the result.
Whenever MMR has been on, he has a very difficult time doing this. Based on just the experience of watching this, I think it shows that MMR does a much better job performing matchmaking than the rank-based system.
So you can argue about what "should" or "should not" count but the results tell the story by themselves.
If you believe that this is an OK state for a matchmaking system to produce, you and I have fundamentally different opinions about matchmaking. It seems the developers do not believe that the above is OK and have chosen to go with a system that, while not perfect, also does not produce the above.
And, I feel this is also important, a higher MMR has no reward. You don't get "rewarded" for surviving or getting a 4k beyond that those actions get you more pips and bloodpoints.
0 -
Grades don't work the same as the old emblem system. There's no deranking, and everyone resets back to the same starting point.
They also did this test right before a grade reset, meaning everyone that plays survivor somewhat regularly was stacked up at iri grades. This lead to a system more akin to random matchmaking than the old rank system.
A better test would have been to start few days to a week after a grade reset, where presumably, players would be the most spread out across the grades. Or just re-enable the emblem system to get a true test.
The way this test was performed almost seems like it was set up to fail, so they could tell us "see how much better our mmr system is?".
1 -
What do you mean both? I just had a game against Oni, where he injured survivor, than was too close for her,
so I sandbagged him in door, than I ran him for a hot minute, he gave up,
did half gen, unhooked like 5 poeple, healed like 3,
360d him more than once durning his Demon Dash, made him leave me from chase, and deplete his whole power after that,
than after he killed all 3 he just let me slug on ground instead of hooking me,
so after all that, I lost because I died, and I didn't get any pips in grades.
You clearly need good team..
Well you clearly see how wrong of a move that was, and how stupid it is,
rank people based on literally how often they play the game, because even a noob can have Iredescent grade...
Yes it was definitely set up to fail,
weather if it was intentional or not,
You see, @rebel581 should know this too, that since MMR was based on escapes,
people that died alot had easier games, people that didn't kill alot had easier games,
and what do you get when the games are easy?
you get pips and higher grades,
And what happens when a player that has easy games, is Iredescent 1 because his games are easy,
plays suddenly in Grade matchmaking? They have chance to meet beast players like Otzdarva...
So yep @Rogue11
0 -
Well... yeah. Sometimes you get bad teammates and you lose the game because of it, that's how every other multiplayer game works too?
That being said, the devs did say something about wanting to look into team-based adjustments to MMR so situations like that are less 'punishing', for lack of a better word.
0 -
Tile generation, pallet spawns, and gen spawns.
0 -
I played with Ranks too. Ranks didn't meaningfully go down with that system either. This idea that grades are substantially different because "with ranks you could derank". It was incredibly hard to derank. In order to derank, you basically had to not play the match. It was a constant source of complaints.
The pip requirements are the same so I can just link this. For rank 17-20, it was literally impossible to derank beyond rank reset. For ranks 9-17, you had to be mostly comatose in order to derank since getting 6+ emblem points could be done passed out drunk. Once you got to purple ranks, 5-8, it got a bit harder, but safety pipping was still easy at 8 emblem points. The only point where deranking even became a thing was at red ranks. Combined with how the system biases to +1/+2 pips for a lot of games and you ended up with a system where it was easier to rank up than it was to go back down. Adding to that, ranks only reset one color in the previous system so it was easier to go back to your previous rank on rank reset. This inevitably led to most players occupying red ranks which meant the entire matchmaking system had to operate within 4 ranks. In terms of matchmaking, there wasn't a meaningful difference between ranks and grades.
I mean this is such a common topic on the forums and it's strange because the rank system matchmaking was terrible. It's not like Otz got his 50 win streak only on the one day where they used grades and that wouldn't have ever happened with ranks if they had kept that system. He did that for about half of the killers in the rank-based system before they switched to MMR. When we talk about their experiment with grades I've heard people acting like they don't have years of data from the old system.
The developers have already read this exact thread a thousand times. Another thread saying the same thing is not going to change their minds.
0 -
Well, all your points are true.
Skill Based Matchmaking? More like Escape Based Matchmaking.
Devs announced they wanted "fair" and "fast" matchmaking.
But queues are still slow at certain time.
After MMR, this game also became more "stale" and "stressful".
There's no "fair" thing in asymmetrical games, which devs refuse to accept.
1 -
That's an excuse, most maps have fairly similar spawns of everything even if it's RNG,
you run for 10 seconds, pallet, you run for another 10 seconds, pallet, you run for another 10 seconds, 2 windows...
Well yeah... that's like saying,
"well it's just how it is you go to work every day, but sometimes the day you worked 8 hours in isn't gonna add to your paycheck"
0 -
Not even slightly, man. There could perhaps be an argument for losing in that circumstance not decreasing your MMR as much as losing under other circumstances - an argument I'd agree with! - but you still lost.
That's how it works in every game, you don't get handed a victory screen just because you got some sick headshots or pulled of a great flank, you and your team still have to win the game. Same for DBD, if you lose you lose, and bad teammates can lose games.
0 -
Yes thats exactly why mmr is not good in this game. Its less of skill and more of "who can get to the tile faster"
1 -
Yeah but that's exactly the only real and biggest issue team based games have.
If the survivors though, don't understand that they have to run towards pallets and windows, and rather choose to run to an open area, I am the one being punished with them MMR wise..
0 -
It sounds like your big issue with team based games is that they're... team based?
0 -
mmr system is not working. why then does the survivor, who always loses, not meet the killer from whom he can escape?
0 -
I mean it should be common knowledge not to run into the open unless you know the killer goes for some crazy mind game. Like survivor isn't hard to play as but people still have no idea what they're doing even 200 hours later which baffles me.
0 -
I wouldn't be mad if the team was a good team, but bad teams exist and frustrates the heck out of me,
because they run to open area instead of loop, because they get stuck on tree and get revved by a chainsaw, because they get easily mindgamed, because they get camped/tunneled, because they play against killers that don't care about kills, etc.
Yes, it is almost you see 100 different players do the same mistake over and over again,
or as killer, go left, nope go right, oh hi survivor that is in my breathing distance
0 -
Because when collecting a lobby, the killer has priority. They give him the survivors, not the survivor of the killer. If the killer loses, he is given survivor potatoes. If the survivor has always lost, he will never be given a killer potato. Unfair.
0 -
Old matchmaking sucked because the devs made it suck.
The safety pip was a bad idea.
They increased the range you can be matched with to alleviate wait times.
They made getting a pip easier, and it got easier the lower your rank was.
They practically took away rank reset meaning everyone eventually got bunched up at red ranks.
1 -
If you're going to make a ton of changes to the emblem system so that it functions purely as a matchmaker and not as a progression system as well, isn't that basically the same amount of work as implementing a basic, standard MMR system?
At least this way the emblem system gets to exist as a progression system too, without that part being sacrificed to make it actually work as a matchmaker.
0 -
It would reflect an individuals skill better than escape or die 🤪
You have no influence over the other 3 survivors and what they do, but you are being judged as if you do. If the old system was refined then it would better reflect the effort you put into the match.
1 -
Would it?
Would it actually reflect your skill at consistently winning the game better than just measuring how often you win the game?
0 -
Yes.
How often you win isn't up to you as an individual. You can be the best survivor in the world, but you won't win if the killer ignores you and just kills the other survivors, leaving you for last.
On the flip side, you can run a killer for 5 minutes and it won't matter if the other 3 refuse to do a generator. if the old system was refined, it wouldn't matter if the other 3 survivors were awful, because you would be able to show your ability to contribute towards the win and hopefully be matched with like minded survivors, leading to easier victories.
1 -
Those things don't happen every match. They're outlier scenarios, and any given time that they happen isn't going to swing your MMR all that much.
It has to happen to you almost every match for it to actually be a flaw with this matchmaking system. Otherwise, it's just bad luck. Bad teammates lose you games in any game where you have teammates, it doesn't mean that you didn't lose or that the game itself needs to be changed to accommodate.
That being said, I believe the devs have said they're looking into adjusting the system so losses like that don't lower your MMR by as much, which I think is a good solution to that 'problem'.
0 -
Yes it absolutely would?
0 -
How so? Bearing in mind that we're looking at consistency, not how someone played in one specific match.
0 -
Is it though... Not from what I've seen, honestly. I'm really glad I started playing this game under the old rank system. I remember actually going with/against fellow newbies when I still was one for about 2 weeks until I was starting to see meta perks every game. That just doesn't happen anymore.
1
