when will mmr really be measured by skill?
its a joke! how that be "calculated", since bhvr made the soft cap so low, you either have to, play swf or killer whatever you do, never play solo q! that is the way to go
but I mean you can't really enjoy a swf either because you get a hundred hours killer in your lobby against you, for example a killer who had recently gotten myers, I was so sorry for him, he said "I got myers for the first time fetched me and will be matched here with people who have thousands of hours" (you could also see this from the perks tier 1 etc.), I apologized and wished him all the best, because your crap "mmr" (which isn't one) works so well ... he said "It's okay, it's not your fault, it's bhvr's fault" that day breaking my heart poor killer :((
finally make it really to skill! chase duration for survivor, how many hooks you make as a killer, and there are plenty of other things and options too (doesn't have to be perfect!) but it shouldn't be dull after bubba who facemcaped with dead lock, noed no way out and very often 2-4 kills makes = skill, or 500h urben calud hiding and escaped +mmr and in the same time the 5k h Nea looped the killer minutes then the killer got mad and camped away = -mmr bad player, and YES! that happen often!
Comments
-
All MMR is supposed to be is a relative indicator of someone's skill vs the average. The definition of skill isn't up to the matchmaking system to decide. You end up with a really janky system when you ignore win rates in favor of arbitrary measurements of "skill".
Players get way, way, way too caught up in single game MMR adjustments. Yes, sometimes that Urban Evading survivor gets out and the 5000 juker looper god dies. It's about as loud as an ant sneeze over of 100+ games. I bet that 5000 hour player escapes at a higher rate over the course of 500 games. I think I've escaped over 2000 games last I checked. Not even total games played. Just escapes. The outlier game where the player with 10k points escapes and a good player dies is not causing this global shift in MMR.
A single game MMR shift is hugely insignificant. And trying to legislate what skill is through MMR would just lead to more bizarre matches. It's only supposed to measure your win rate over time relative to your peers. There are 30k+ people on PC alone at any given time. The single game results aren't even statistical noise.
6 -
MMR was never designed to be an indicator of skill, and the correlation to skill is only because skilled play tends to lead to more kills or escapes. What the Kill Based Matchmaking system in use was designed to do is provide a function that can predict the chances of a survivor escaping/dying given everyones MMR and then attempt to match people in games where any given survivor has an equal chance of escaping through the gates as they do dying.
MMR in this system is not a measure of skill, it's a measure of your likelihood of escaping. Whether this happens because you're a god looper who can loop every killer for 15 minutes or because you Urban Evade around the map and then open the gate with Wake Up + Sole Survivor when hatch gets closed doesn't matter, only whether you escaped.
This does start to break down as you go over the softcap of 1600 MMR but you being able to get matched with a completely new Myers would indicate that at least one person in the SWF wasn't over it (Even in the worst case scenario of the Myers player already having another killer over the MMR softcap, it would still only bring their Myers MMR up to ~1200 MMR [assuming they haven't changed MMR in the past 3 months since how MMR works was revealed by dataminers]. We don't know exactly how big the expansion range for MMR is [How mismatched of an MMR lobby you can get thrown in] but before the March/April MMR test it was 500 MMR and was lowered by an unspecified amount after the test and during the test they tried lowering it to 250 MMR and then 350 MMR. Even in the 350 MMR expansion range case they'd only be able to match with up to 1550 MMR players, which is below the softcap)
As a final note, it's absolutely possible to play Solo Queue past the MMR softcap and still do fine. You need a lot more gamesense to simulate the effects of comms, but the good part of being past the softcap is that you can generally reasonably assume your teammates have some idea of what they need to do and so makes it easier to predict where they are and what they're doing.
1 -
I fired up my fresh EGS account for the first time a few days ago, and when I started, I could just 4K every match at 4 gens as Trapper with no perks or addons or even using any beartraps.
Switch back to Steam, and I can run the sweatiest builds on my best killers, and some games are just rough.
I'm not sure exactly how MMR works, but it does seem to be working. No complaints on my end.
0 -
okok i got it now, MMR or skill based matchmaking (sbmm) called, its not a skill based matchmaking understand it! btw football has nothing to do with a ball you kick, you just call it that and when i walk into a candy store I hope I only find fruit
no matter how bhvr wants to call the alleged MMR, it is nothing more than a hidden embem system in the guise of the mmr
Scott Jund do a well video for that, its dosent exsist that is what i say too (logical mind says it when you deal with it)
if you would take a quick look at what mmr stands for then I wouldn't have to tell you that but I mean the word itself so (sbmm) it becomes obvious what this is about
here for you again what "MMR" stands for: - Matchmaking Rating, or MMR is a value that determines the skill level of each player. This value is used in matchmaking. Winning increases a player's MMR, while losing decreases it. Ranked Matchmaking allows players to earn seasonal ranking medals based on their MMR.
you are welcome to say "in all games where there is an mmr it will be measured by skill but not in dbd" i would accept that and it is byfar an insult to any mmr what we have in dbd is a simple matchmaking from 100-200h will you are matched with everyone if you have less you will also be matched with newbis
Post edited by DudelPuma on0 -
Cause "skill" is way too vague of a thing to track. Not to mention not a good one for matchmaking. Which is the point of mmr. Matchmaking, not some badge of honor
In the end of the day if you have skill you will win the mayority of your matches against those lesser skilled.
If you aren't then you aren't as skilled as you think you are.
2 -
If 'skill', however you choose to define it, doesn't correlate with win rate, then 'skill' isn't a good metric for matchmaking.
2 -
Actually, MMR not directly measuring skill is a property of every MMR system on the planet. What any functioning MMR attempts to do is to be able to predict the result of a match, of which skill is only part of what influences that.
As an example, if I start all of my chess matches with the Bongcloud Attack as white (1. e4 e5 2. Ke2, widely considered to be one of the worst possible openings for white) then I will have more skill than other players at the same ELO because my opening is bad and that causes me to lose more games compared to a person with similar skill and a more sensible opening. Similarly if you play CS:GO and only ever use the Desert Eagle then you're going to be miles more skilled than your opponents who are at the same MMR as you because your opponents are using actually good guns.
The reason why MMR directly measuring skill would be unusable for a matchmaking system is pretty obvious when you think about it; people don't want matches of similar skill, they want fair matches. If I have the same amount of skill on Trapper as I do on Nurse, I'd prefer getting the seal team SWF games on Nurse instead of Trapper because in those games I can actually do something to win compared to Trapper where I'd want less skilled opponents because I have less tools to deal with the other side.
1 -
As another poster said, it's a marketing issue on BHVR's end. Calling it SBMM was a bit of a mistake, mostly due to the community's bizarrely narrow definition of skill and complete lack of introspection in the face of conflicting results.
For example, rather than accepting that game sense and positioning are more valuable survivor skills at the top 5% than looping, the community simply decided those players were somehow bad despite winning a majority of their games. And they decided MMR doesn't gauge skill at all.
1 -
You will never manage to convince me a killer that let 4 survivors escape is any good, they got destroyed by those survivors, they absolutelly wrecked him since he was so mediocre he couldnt even get a single kill, bot trapper and they can get the same results.
0 -
I know you’re a biased killer main or whatever but this isn’t correct. Survivors don’t fair well on killer-sides maps, and you can’t say “as long as you run the right build, you’ll do well” while also, in that same sentence, saying “survivors for the most part on any map with any build do well.” Those are oppositional statements. Additionally, it’s much harder for survivors to gain MMR than killers because survivors rely on other survivors to escape in a way that increases MMR (through gates). A single survivor cannot carry an entire game to victory under standard conditions.
1 -
The RNG doesn't matter at all over the course of hundreds of games though. You'll experience it both ways hundreds of times. In a single game, sure. But single games don't affect MMR to a significant degree. MMR isn't meant to judge someone's skill in a single game. A better player is generally going to win more than a worse player over the course of hundreds of games.
1
