Devs: Your Update Indicates MMR kill-rates are changing
The average kill rates (how many Survivors are killed each trial, on average) are a little lower than we’d like – a sentiment echoed by many community members.
Since Killer kill-rates are hovering around a ~55% (depending on the killer), according to BHVRs data. This statement indicates that Killer kill-rates are going to be increased, so it's less likely that survivors will escape.
Survivors currently lose ~55% of the time, what will be the increased survivor loss rate percentage after the update?
Comments
-
It would be interesting to see the data of swf vs solo q escape rates because im assuming a lot of the lower killer rates are because of swf
2 -
To be fair, the survivors are the real killers of DBD. You lose more due to what your teammate did/did not do rather than the actions of the killers themselves.
Not to mention the trolling, dc'ing, and general BM that conflates the kill rates immensely. It only takes one to kill everyone.
The amount of games that are ruined by poor teamplay/toxic behavior is what im going to assume is larger than 5%, so they are increasing potential kill rates to counteract the overinflated stats.
5 -
I want them to post what the escape rates are at high MMR where it's sweat squad central. I wouldn't be surprised if it's 70%+
4 -
As far as I'm aware that's actually old data so we don't actually know what the current rates are with mmr.
8 -
They have released global stats, with all MMR level included. Be sure at high MMR the kill rate is ridiculous.
0 -
This. Survivors ######### around, screw around, and self-sabotage so much and kill rates STILL aren't where BHVR wants them to be. That is ridiculously telling.
Also this. Did we ever get updated #'s after COH?
4 -
can someone post these stats? I see them referenced in threads but wanna look myself
0 -
That data is from January 2022. MMR is unable to change the stats in a radical way, since game balance stayed the same. If you match players in a slightly different way, it won't change the global picture. It can only be changed with much longer que times, for example if stronger killers stay the the que for longer, waiting for a match, while at the same time weaker survivors are matched to weaker killers. But it's not the case, so that data is still valid, maybe +/- 2-3%.
2 -
Can you actually back that up with something, or are you just guessing?
1 -
Back up what exactly? Stats/my reasoning?
1 -
I don't believe we do.
To be honest I'd much rather them show us hook rates(not hook stages) rather than kill rates. Kill rates really are a flawed metric, especially for how much they're used in terms of balance changes.
Split out the data from different mmr brackets.
Show us the hook rates when the 5th gen is popped and compare that data to hook rates at end of match. I'd be willing to bet you'd see literally a doubling of hooks and even kills for that matter from 5th gen to end of game. IE end game hook bombing and NOED account for literally half the games hooks/kills and are skewing the data tremendously. I consider this a skew and not good data because at that point it isn't really representing balance as much as it is just face camping and not wanting to let a teammate die and ultimately the entire point of this data is for determining balance.
On other words the data as it's generally been presented in the past is very misleading to most people who take it at face value.
1 -
Your reasoning.
0 -
They will never split out the data from different MMR brackets, LMAO. But we can hope.
1 -
And we all know why..lol
4 -
I actually think this is very inaccurate.
"MMR is unable to change the stats in a radical way, since game balance stayed the same. If you match players in a slightly different way, it won't change the global picture."
Before we were matched much more casually which covered up the games balance problems as you were much more often facing lower skilled players. With the introduction of MMR you are now facing much higher skilled players and much more SWF groups on average at the top. Matchmaking can have a very large impact on kill rates.
1 -
I have a feeling there is a chunk of the roster more or less missing from higher brackets. If I play with my 4 man (combined 20k hrs), we pretty much only see Oni, Nurse, Blight, Twins, Hag, Billy, Spirit.
0 -
I think this is the latest stats, but I think there was actually a newer stat release where kill rates were broken down by MMR. I thought it had released sometime close to when the meta perk usage stats were released. Maybe someone has that.
0 -
Yeah pretty much. Except Spirit, we haven't seen a Spirit in weeks it feels like.
I don't want to be mr tinfoil hat over here but I've always felt like the stats were intentionally presented in a way to give a specific view that caters to the balance changes they want to make anyway for validation. The issues with how the data is presented just seem too obvious for them not to realize it which is why I can't see any other rational explanation for it.
0 -
I can show you a generalized example.
Take two decks of cards with 100 cards each. Yours is numbered 1 through 100, mine is numbered 5 through 105. Now we both draw one card from our decks. Whoever has a higher number - wins. Repeat 5 times. Now we have the average. I have a higher chance to win on average. Now shuffle the cards. Repeat the same experiment. Nothing has changed, I will still have the same chance to win (in reality when "law of large numbers" kicks in).
Shuffling (changing the MM) can have some local influence, but the global picture will stay, my cards still have higher numbers on average than yours. Same applies to DBD, as long as average que times PER PLAYER haven't changed that much. Changing que times would mean that we can take some card (like "105" from my deck) and remove it from the deck for some matches (for example strong killers or survivors will stay in que for longer and will be skipped during matchmaking).
All we can achieve with a different MM is a more fair distribution of "cards". 1 will be matched to 5 and 100 to 105, instead of 1 to 100 and 2 to 105, but it doesn't change our OVERAL chances to win after x repetitions.
0 -
I don't want a generalized example. I want something to back up the position you've taken.
0 -
I just did. Use your brain.
0 -
You really didn't. You basically just strawmanned it.
0 -
Ehm, I just explained you the principle, that's how it works. If you are unable to understand it, I can't help it.
0 -
You haven't actually demonstrated the relationship between your example and the position you've taken. You basically just created an unrelated scenario and then said that it ties into your original position because the scenario you created holds up.
1 -
Tag: show me that the sum of 1 and 587 apples is bigger than 587
Math teacher: proves that the sum of x and y is bigger than 587.
Tag: you strawmanned it
Math teacher: I showed you the principle
Tag: you haven't showed the relation between x and apples
That magnitude of illiteracy and inability to think keeps amazing me. You have to think too if you want to understand something, don't expect someone to spoon-feed you, unless YOU want to strawman it, which you probably do.
Anyway, I proved you my point, whether you are able to grasp it or not only depends on your mental abilities. Try harder.
0 -
I only ever see Spirit with MDR/Cherry Blossom. Or the AI bot, but that's another story.
I get why they present data the way they do. They could make it more granular, but it would probably show problems they're aware of and not trying to fix. I think it would be interesting more than anything just to see a census more or less of killer population and pick rates per bracket. Even if they wanted to hide win rate.
0 -
i'd say it's more like you tried to prove that 1+ 587 > 587 by showing that 5 + 5 =/= 9.
1 -
"I only ever see Spirit with MDR/Cherry Blossom."
Pretty much the only way I see her as high mmr viable.
"Or the AI bot, but that's another story."
Yeah we've ran into that bot a few times too.
"but it would probably show problems they're aware of and not trying to fix."
Pretty much my thoughts.
"I think it would be interesting more than anything just to see a census more or less of killer population and pick rates per bracket. Even if they wanted to hide win rate."
I'd be interested in seeing that as well.
0 -
The 55% kill rate was from the old figures. They did not share what the actual average kill rate is, as of June 2022, but apparently it's low enough that they're actually willing to tweak some core numbers.
It's entirely possible that average kill rates are below 50% currently. The fact that they DIDN'T share the number sure makes it seem like it might even be lower than that.
0 -
Yep. I'm convinced the high MMR escape rates are easily 75%+ and that's with teams sometimes giving an end game kill because they refuse to let their teammate die. I think 3 escape average is highly likely
1
