SBMM DOESN'T WORK. KILLER STATS FROM 100 MATCHES TO PROVE.

The reason his data is different than the OP's -- and both are different than the results I'm seeing from my match results -- is again, sample size, which is ridiculously small considering the number of matches played of DBD daily.

The steam charts show that, on PC alone, over 45,000 were playing DBD 5 minutes ago. If we were to assume that there were NO console players playing the game at that time (which would obviously, not be the case), with 5 people per match, the rough estimate would be that thousands of separate matches were going on at that time, even if some or most of those players were simply sitting in lobbies at any given point in time in that short period. When you start adding in the console players, we're easily talking tens of thousands of matches a day. Statistically, using 100 random matches over any period is going to be such a small sample size -- we're talking literally fractions of fractions of a single percent (assuming all of those 100 were in a single day, and potentially matches overall numbering in the six-figures in that same period of time), that you simply cannot use it to "prove" any point. Whatever stats the devs throw out -- assuming they're using real data and not simply making things up on the fly -- will be far, far more accurate than anything ANY of us can come up with, even if we pooled all of our numbers together.

The OP's experiment is simply one person's brief snapshot of a much larger event that he (or Pulsar, or I, or you, or anyone here) simply do not have enough data to draw actual meaningful conclusions with. They're interesting to read what ONE person might be experiencing over what amounts to a brief snapshot in time, but statistically, they're not going to be necessarily representative of the complete, larger experience.

ETA: And all of the above ignores all of the other variables that could come into play to skew numbers from any of us. SBMM, any of our individual skills, all of the RNG involved in the game, whether people are playing to meme or playing serious, what perks are being run on either side, etc., etc.

Yes this is right. It doesn't even have to be as many as 100 but if you play a few matches of Trapper, only his MMR changes and only by a relatively small amount. As soon as you switch killers, you switch to that killer's MMR.

These aren't stats aren't for 100 killer games these are 26 seperate stats of 4 killer games.

It all depends on the data that was recorded and how relevant that data is to the subject. For example if the OP believes hes in high MMR games then it makes perfect sense -- he recorded 100 high MMR games and these are the stats we see.

If hes not in high MMR games, which he almost exclusively confirms (by saying hes an average killer at best) then he just recorded 100 games of stats that are irrelevant to the subject

So, I have 200 matches recorded on my own as survivor which show escape rates -- both for myself and for people on my "team" -- that are wildly different than either the OP's, or Pulsar's (mine are in the middle of the two of their results).

Care to explain -- without using a biased take -- which one of the three of ours are relevant and whose aren't then? How are the OP's numbers statistically significant, yet mine or Pulsar's aren't, and why?

I worked in a casino for a number of years. This debate reminds me of the people who would come to a roulette wheel, look at the board showing what numbers had hit already, and then betting certain numbers because they were able to "identify a trend" based on a handful of spins on record. That would pretty much work out for them the way rational people would assume it would (and is why casinos stay in business -- they're built on people who see numerical trends that simply aren't there or accurate and pay for it -- literally).

The more I see, DBD's issues are less about Killers not being able to get kills when appropriately matched... it's more how it feels to play Killer and big disconnects between what BHVR thinks is a "good showing" for a killer, what the community thinks a good showing is and how BHVR rewards less than stellar killer games.

As a personal ancedote, I remember 2 kill games where I've depipped, as an example. 3 kill "draws" where I gain no pips. 4 kill games that award only 1 pip. For a team that says "2 kills is a draw game, 3 is a winning game for a killer"... their scoring system doesn't really align with that philosophy.

That and this game does have one of the most toxic communities in general - killer gets exposed to 4x more of it than survivor does. Math basically says Killer has to deal with much more toxicity, which can easily lead to decent/ average matches leaving a bad taste in their mouth.

I'd argue it's rarely happening for anyone at this point. As I stated earlier in this thread, I'm of the belief (without actual proof, obviously, just anecdotal evidence) that queue times are the primary factor in matchmaking. During one of the previous tests -- when someone with a presumed high SBMM rating like Dowsey -- would literally sit in a lobby waiting an hour to find a match, I think the devs adjusted things to keep that from happening, and queue times are prioritized over everything else as a result.

Again, that's an opinion. My "data" is that, during peak survivor times at night, when I play killer, I can play someone I'm presumably very good with (like Pig) and my lobby time is instantaneous. When I play a killer I never play and I get clobbered with regularly, I'm still getting instantaneous matches, and survivors whose skill level are pretty much no different than what I get as my "main". (Not EVERY match goes that way, but far more than not, leading me to believe that the rarer lobbies of lesser-skilled survivors I go against are simply happening randomly).

Same as survivor -- when playing for 4-5 hours during the peak periods (either as part of a SWF or a solo), the quality of the killer ranges from God-tier to complete potatoes to everything in between. The ONLY thing that stays consistent (since I've been actively keeping track of it as well) has been lobby wait time to find a match, which tends to differ by 15-30 seconds top from the longest to shortest times. suggesting that those times are determining who I/we are matched with, and NOT any SBMM ranking.

200 survivor games is a much larger sample size than 4 games as Trapper. So yours is the most relevant.

You said in your OP that you play an average killer at best -- why would these stats show you play an average killer...? Something doesnt add up to the stats you provided. Either it's a brand new account or you dont play your best every game

We don't even need to argue that SBMM is not working, but these stats are not meaningful at all. Even just playing at different times at the day can change the results massively.

Well i don't need statisticians to tell me whats the reality when i see the reality every day. 🤡

Which honestly is the issue with all of these "experiments" being done by people hoping to prove all sorts of things on both sides. There is pretty much no variable control, so you are left with something that isn't at all repeatable or falsifiable (the hallmarks of an actual good experiment) and instead have done nothing but record anecdotal data to present as evidence.

That's a fair enough answer.

I will still state that it's simply small sample sizes by all of us that are causing deviations in the results, as well as an incomplete look (again, by everyone who's chimed in so far, myself included) at all of the variables that can go into determining what results any given match will have.

A perfect example in my own data is killers versed against. The most common killer in my matches overall has been The Artist. Could I use that "data" to claim that The Artist is now the most popular killer in DBD, or that she's clearly OP and therefore being picked in numbers I would have previously seen Blight or Huntress or Spirit picked in? I could make those claims -- but considering that I started compiling my data right at the time The Artist was released, that single fact alone would go a large way to explaining WHY that killer was the most picked by my opponents, and it's a perfect illustration how statistics can be completely misinterpreted if there's not enough of a sample size to determine accuracy of those stats and the relevance of missing context for said statistics.

All of that said, if the OP wants to continue to compile stats and update, I'm a stat-head (I've played fantasy baseball very successfully for close to 30 years now), so I'm always interested in analyzing numbers. I plan to keep up with my stat keeping as well, not to "prove" anything, but if for no other reason than it's interesting for me to do. I'd hope Pulsar does the same, and yourself if you want to put the time in.