New kill rate stats for killers (including top 5% MMR)
Comments
-
Why don't they show the kill rates vs. 3 and 4-man SWF groups?
2 -
Just shows how meaningless tier lists are overall.
2 -
idk if I would say they are meaningless.
There's still some truth in the tier lists people make, but people can easily underestimate or overestimate the power of things when making a tier list.
2 -
Remember, lesser-played characters are more likely to have kill rates that are less representative of how good/bad they may be. Pig, for example, has pretty consistently been among the least played Killers, if memory serves correctly.
0 -
There are so many elements missing that are important in this analysis.
For example, how many games played by each killer to get that kill rate ? Sadako for example, is not strong imo; However she is #1 spot.
Could that be because she is played much less given her weakness ? And those who play her, only use her strongest build with passive condemned. Not to mention that she is a basic killer to use, so she can stomp new players more easily which is still not an indicator of her power.
Same for nurse, Imagine 100 nurse veterans who would 3-4k on average. This will be negated by another 100 new nurse mains that will probably get 0-1k at most.
These stats don't tell much tbh.
2 -
This is why kill rates are a pointless stat. How many of those nurse games had someone who just DCd? of course the nurse gets 3 kills after that. They need to show hook counts and/or WINS. And actually define what a win is.
Additionally, "top 5% mmr" is a pointless stat. Because MMR is capped at 1700, which is ridiculously low. To put that into perspective, my chess rating is around 1600. Magnus Carlsen, the world champion has 2856 rating. So follow BHVRs logic. Magnus is only slightly better than me. When in reality, i could play him in chess for 1000 games and probably never win. I'd be lucky if i even got a draw.
0 -
Matches where someone DCs aren't included in the kill rate stats.
4 -
It's quite surprising the top 5% is higher than the global average for sure. But I can't see a figure for any other MMR brackets besides the top 5%.
We know that the lowest MMR bracket, maybe the lowest 5 or 10%, is heavily killer sided as well.
This would suggest the lowest kill rates occur somewhere in the middle.
0 -
The stats are misleading, the top players are not.
0 -
That's not the only reason the stats are misleading, that was just a single example, but do they factor out hook suicides? What about survivors that go afk? What about SWF?
0 -
Can they factor out hook suicides? Or any of the other things (SWF I get the impression they can)
0 -
May I ask where these stats come from?
Are they official BHVR stats pulled from the server logs of every single DBD game played during that month?
or is it 3rd party website, self submission reporting from the few players who even bother to go there and cherry pick what they upload?
0 -
Maybe the could with SWF, but hook suicides maybe not.
I posted this in another thread, but the biggest reason they stats are meaningless is:
Top 5% is a meaningless stat, because the MMR cap is 1700. To put that into perspective, i play lots of chess. My rating is around 1600. Magnus Carlsen, the world champion, has a rating of 2856. But BHVR would cap him at 1700. The reality is though, if i played him 1000 games, i'd never win and would maybe get a draw in a handful of games. Because of that statistics are skewed no matter what. Because a pig with 10k hours and a proper rating of 2800 will get matched against 1700 MMR survivors and will likely 4k without even a sweat. Just as a nurse with 200 hours and 1700 rating will get matched against a 2800 rated SWF kill squad like this:
Because of that, the stats are basically useless unless they specifically break down the top players vs the same top players. Like, let's say they actually track MMR about 1700 but it just caps for matchmaking purposes. What are the stats of the players with 2k rating playing against players with the same (or relatively the same) rating? That would be much more meaningful and useful.
0 -
More than half the time someone DC's in my games, is because killer downed them first.
This should count. Especially in cases like Nurse, whom players hate and get downed quickly against.
1 -
No, it should not count. Just because a survivor doesn't like a killer doesn't mean they are strong. If they looked at DCs and counted them then legion would probably be considered the "strongest killer in the game" even though they aren't.
2 -
They shared stats on their socials.
2 -
That wouldn’t explain her kill rate on the 5% MMR chart though, nobody being counted on that chart is a bad player.
1 -
Are hook suicides counted?
0 -
The game has literally thousands and thousands of matches a week even in the 5% bracket for every killer. None of these stats are from just a couple of people.
3 -
We have different goals.
You want to know which killer is "strongest"
I want to know which killers are a problem for the health of the game. If counting DC's means it catapults Nurse into 60%+, then this is a very important statistic that needs to be known and addressed for the health of the game. Why are people leaving her games, etc. The opinion of a large portion of the playerbase matters. Goal of game design is for everyone to have fun.
8 -
Yes she is. But she requires more skill than nurse. That's why more confident and more skillful killers pick sadako while the bad killers pick nurse.
2 -
Peanits confirmed above that matches with DCs are excluded from the data.
2 -
More like nurse numbers are "deflated" and actually much higher if it weren't for the bot nurses losing
2 -
It's hard to say for sure for really much of any of these numbers.
I've always been of the mind that these stats are fun to see but that's all they should ever be used for; fun.
Only the devs have the information and context to really determine what these numbers mean and what, if any, changes need to be made.
1 -
DCs do not show that though. There are much better ways to track that sort of thing. A DC is not the way to do it, because DCs could just increase due to a bug for example.
0 -
Sure, these charts aren’t the whole story, and the devs even say that in the article. They’re not totally meaningless though, it’s harder to argue for instance that “Sadako is the worst killer in the game” when she has an average to above average kill rate in the top 5% bracket.
3 -
Is it possible that the killers with the top 5% mmr didnt actually face survivors in the top 5% in order to make matches?
I'd love to see killer and survivor performance by mmr. Show me completed gens, hook states, and kills.
0 -
They're releasing stats that reflect the reality of matchmaking. How is it somehow more meaningful to sort for a percentage of MMR that isn't actively used in matchmaking? The top 5% falls into the same bracket for matchmaking purposes, hence the chart. The chart isn't meant to account for skill variance among that. It's to show kill rates over thousands matches sorted for the soft cap. Just because it isn't displaying data on a granular level for X factor doesn't render it meaningless. It doesn't take into account the difference between a 2000 MMR Nurse vs a 1700 MMR Nurse because matchmaking doesn't.
If someone wanted to get a better idea of the top 1%, they could watch some DbD League or something, but that's not a public match environment.
4 -
Just sliding in here to let you know that we received many requests from players to include Pick rates in our post, so we've gone ahead and added those in.
22 -
Thank you for that.
Nice to see my boy Mikey have such a healthy pick rate in both MMR brackets.
1 -
DCs should not be used for balance purposes, because that would encourage survivors to DC even more often, in attempts to force the devs into nerfing a killer.
0 -
The Twins don't seem like high MMR monsters to me!
1 -
I agree with all of this
0 -
@Peanits is there a way to show escape rate for each killer? Which killer has the most escapes when played against.
0 -
Well these stats already show that. If a killer has a 60% kill rate then that means 40% of the survivors escape.
5 -
Nice! thank you!
0 -
Thank you, now I definitely understand why my games have so much Weskin' time.
0 -
I guess that sheds light on why the 100% bonus is almost always perma survivors.
Proud of trapper though.
0 -
Well, at least we can put to rest that GF is a weaker Myers and that Myers is nowhere near the worst killer in the game.
3 -
It would be interesting to see stats including DC’s as well. At least in my games the DC’s happen mostly when the game is literally lost while laying slugged etc. Not at the start of the game, that problem is way gone for me. The fact that those games don’t count could mean a little in the bigger picture - especially with Nurse who I get stomped by way too often.
0 -
Personally, I always felt like people really overestimated Pyramid Head. I didn't think he would be that low though.
3 -
i think it’s worth considering that there may be significant variability amongst that top 5% MMR. Like if you look at the top 5% income in the United States vs top 0.1% there is a tremendous difference. So top 5% doesn’t exclusively mean everyone is a 4 man Seal Team SWFs. And against really good killers it becomes more about teamwork and less you can do as an individual survivor, and killers like Sadako, Dredge, even Freddy can really breakdown the teamwork or macro strategy of the game.
1 -
This is such a toxic mentality. Everyone loses.
5 -
I want to appologise to community, it's my bad that Sadako is so high and Blight is so low. I dumped Blight really heavily, because I still dont understand how to control him. And Sadako.... I just play her a lot, so.... forgive me.
3 -
I REALLY wish we could see kill rates by killer against SWF
0 -
Compared to other killers? No. In a vacuum ? Yes.
0 -
Dc matches don't count in those stats I quess kill rate would be closer to 70% if they did. That also lowers nurse kill rates so many dc from her.
0 -
In high mmr there is lot less dc:s that is my quess why kill rates are higher there.
1 -
The pick rates pretty much confirm the reason for the discrepancy between perceived power and kill rates and also likely shows a general problem with using kill rates at all.
A good killer is going to be good regardless of the killer. They'll have an easier time on Nurse/Blight/Spirit than another killer, but they'll still be good regardless.
At the same time, a medium-good killer will be boosted to good by playing a better killer. A lower tier killer with a medium-good killer will have a harder time making it into the top MMR bracket because the killer themselves is worse.
This chart and the pick rate kind of prove that MMR is working. MMR will self-balance so kill rates are, mostly, meaningless if you sort them by MMR because your MMR is dependent on how many kills you get each match! By definition, you'll gravitate towards the median kill rate. It's a circular definition. What does change is your MMR rating and the pick rate demonstrates that more accurately than the kill rates. If there's more Nurse's at MMR it's because Nurse killed more survivors to get there.
The other component is skill floors and skill ceilings. Sadako has probably one of the lowest skill floors in the game, but I feel like her skill ceiling is probably also pretty low. Yea, there's maybe some strategies like passing condemn to different survivors and keeping track of where people are on the map, but it's mostly demanifest, teleport/move to area, manifest, attack. There's not really a whole lot of technical skill beyond standard killer skill but you still get the effect of your power (teleport/undetectable) after maybe playing 2 matches so she's pretty much at almost full power after a small amount of time.
Nurse isn't exactly hard, but she's not as easy to play as Sadako. And most of the top tier killers involve more skill and are more interesting to play. Which means there's more opportunities for messing up and losing a game because you played badly.
0 -
So following bhvr logic that means any killers that share the same tier as twins in high mmr or on higher tiers are now part of the absolute monsters in high mmr club
Now all the pig nerfs make sense lmao
1
