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!
The kill rate argument
Kill rate is a flawed metric for determining balance for a few reasons.
1). It doesn't take into account the skill difference between killers. Huntress/Blight/Nurse are a lot harder to play than Legion/GF/Freddy. Despite having a low kill rate nurse is still considered as top two if not the best killer in the game by a majority of people. If kill rate is all that matters shouldn't nurse be receiving some major buffs?
2). Kill rate lacks context about how the game went. A very close, back and forth game that has 9 hooks in total can end with 3 escapes and 1 kill. A total stomp that leads to someone getting downed in the end game collapse with NOED and dying on first hook after being face camped also ends with 3 escapes and 1 kill. These two games despite having the same kill rate of 25% are not the same.
Kill rate should stop being used in arguments and as a balance philosophy as a whole for BHVR. Doesn't go deep enough.
Comments
-
The thing that always bothers me when they show their "kill stats" is the fact they don't bother checking if there was an afk survivor or if somebody suicided on the first hook or which perks were used durin the game.
19 -
Your reasoning is solid, but you need to suggest an alternative to kill rate then. Otherwise, the whole conversation is moot.
3 -
BHVR has the ability to dig in as deep as they want to. For example, for Nurse they have stats that Almo has shared, that says that when comparing players dividing them each by their strongest killers and then sorting by percentage of kills, and then looking at which killers are performing the best, then what you find out is that, players who main nurse and are the top 1 percent of nurses have kill rates comparable to that of the other 4 top performing killer.
What this means is what everyone already knows, in the right hands Nurse is a beast and is comparable to or better than other high tier killers.
This does not mean that the statistics are incorrect. Nurse's kill rates are low, and even when looking at red ranks her kill rates are low, but the developers have the measurements to see that she is not in the need of buffs to her kit.
Just because you don't understand how to use statistics to get reliable information about balance does not mean that the developers can't.
6 -
Also, there are more red rank survivors than ever before. Most of these survivors have gotten there purely by playtime, you can tell by the way they hold M1, and can’t compete in chases at all, and typically their load out (some still use Premonition).
Saying that the kill rate is higher than ever before is a flaw in itself. If more potatoes make it to the kitchen, more potatoes are gonna get cooked.
4 -
I mean they do for internal balancing reasons.
However the stats they post publically don't show any of that so everyone will go "Omg whiny killers look this killer has a 65% kill rate so that means they are OP" and it's like...no....there is more to it than that....
2 -
Can't speak for anyone else but I've been encountering similar 4m swf groups that troll the killer hard then with a full team escape ready , 1 runs at me and forces their death. It's stuff like this that makes those numbers inflatable or inaccurate.
2 -
That's assuming they only look at kill rate and ignore everything else, which it was indicated that they don't do that. They take everything they can track into account.
4 -
For balancing I agree that's not the only thing to look at. But ask yourself when was the last time they used kill rate as the sole means of enacting balance changes to a killer? Freddy currently has the highest kill rate that we know of and for the last 2 years and he hasn't been nerfed in more than a year. The last nerf I can recall for Freddy was the jump rope nerf, which was a pretty minor nerf.
BHVR uses a lot of things to determine what they buff and nerf. They take into consideration pick rate, kill rate, feedback from the community at the very least. Usage statistics is actually what they probably look at the most to see what needs adjusting. If something is being used by 90% of the players and it's causing problems they usually nerf it.
Anyway, in my opinion it's thinking like yours which has led to the current mess of ranking and emblem nonsense that we have. The result is perpetually bad matchmaking. Kill rate should definitely be used for matchmaking at least.
1 -
Sorry guys, but kill rate is accurate. It could be 10 min games , survs that kill themselves on first hook or people drunk, but it’s the number of times survs escape. And they say killers are overpowered.
1 -
The Data Almo, what does it all mean?
6 -
It can be used to work with a point in an argument. I wish BHVR showed more graphs or added more variables to existing graphs to give a general idea of Killer performance level.
0 -
this is a killers forum, but you can’t think 1/100 games against a 4swf marine squad is an argument to buff killers but average escapes provided by bhvr should not be taken in consideration
0 -
once they drop the MMR update it will be easier to use statistic for balance
0 -
I think the opposite, I think they know exactly what they are dealing with down to the last number, basically doing it with tracking programs like the automated system that issues a ban if you disconnect, you think they have that automated system without gathering the info? heck they probably have the real life info of anyone who makes a purchase using a credit card ( I work for a big company and we have all that info stored despite the fact that we tell customers we don't have it) they just don't want to reveal all the data.
They probably have exact data about every match and that's the reason why they are constantly buffing survivors and nerfing killers is because they know the exact numbers of optimal games played taking away disconnects, afk and stuff like that, If numbers were on the survivor's side I doubt they would nerf killers as hard the way they do it.
0 -
If the devs can get so much reliable info that we can't, why don't they make much better balance decisions?
1 -
Yes because CLEARLY coup de grâce and gearhead needed nerfs.
3 -
DBD is a complicated game, which requires complicated data analysis to properly understand what's going on.
Chase duration, kill rates, gens completed at certain breakpoints (like how many gens done by 6/12 hooks, any survivor that has reached second stage on hook counts as 2 hooks), killer skill level vs individual survivor skill level, killer skill level vs survivor team skill level (combined), addons, map, map configuration.
Although I'll say that maps and map configuration doesn't strictly need to be in for balance purposes, as those can remain the RNG factor of the game if all the rest of these variables are accounted for.
0 -
Also, the lack of proper rank/matchmaking makes this data completely unusable.
- Because there's no matchmaking, most games will have a random skill distribution, meaning that the odds of having high level matches without at least one 15 hour Dwight in it is impossible.
- Because there's no ranking, it isn't even feasibly possible to go in and purge data from low-skill matches after the fact, since there is no way of sorting for skill.
1 -
Not really, companies often do things that may look stupid to people (Had that happen to me at the company I work for, for example if you order something and want to cancel it before it even ships my company would rather stick to their policy and ship it first to you and then you ship it back , you would think it's a stupid idea because why lose money shipping the item and sending it back if it has yet to leave the warehouse if it's costing the company money?) but that's just how it goes for some big companies. pretty sure some dev had the ingenious idea of nerfing those perks for some stupid reason and the idea went trough somehow...
I think you are also underestimating how deep they can dig, with the company I work for I can tell by a simple look at my computer screen how often you go to one of our stores, if you went there alone or with friends, at what time you arrive, at what time you leave, what items you purchased, at what register you paid for the items, which employee attended you, if you use a credit card we know your address and stuff like that. it's not even expensive or complex, it's jutst part of the system most companies have... if they have the data for kill rates pretty sure they have everything else just that some times companies do stu*id shi*....
2 -
There is no need to complicate stats. A kill is survivor losing and killer winning, no matter how it is obtained.
According to the official stats released, killers are way overpowered because they win a lot more survivors.
Even the lowest killer, the nurse, wins almost 50% of the time. This tells me that ALL killers should be difficult to use and that killer by in large is entirely to easy.
For example, you shouldn't play freddy and within a few hours win a majority of the time. That is HORRIBLE balance.
3 -
The kill rate is flawed. Other metrics would be much better suited, even the total number of hooks in a match.
1 -
They share stats with us and tells us not to draw any conclusions from them.
what do we do???
0 -
The Devs themselves told us not to draw any conclusion from the stats. These arguments are silly. I'm 100% sure the Freddy nerf is for multiple reasons outside of the killrates.
6 -
A killer that averages 1 hook 1 kill a game and one that averages 9 hooks 1 kill a game isn't the same and if you think they are you should never work for the balance team on any game
3 -
I don't think taking kill rates into consideration is a bad thing. The problem is when it's actually the only thing you're taking into consideration. Say, for example, the argument for keeping Object of Obsession as it is. The devs consider that OoO is fine because survivors who run it don't tend to escape trials significantly more frequently than survivors who don't. However true this may be, it does not change the fact that, as a killer, it's an absolutely overwhelming perk to play against, that some killers suffer from it more than others (mostly Trapper and Hag) because their main power is easily countered by it, that SWF can abuse this perk, etc. Saying "nay this perk is fine because these stats look normal" is completely missing the point.
Now, that does not mean that you can completely dismiss stats. They need to be taken into consideration with their context and all, but numbers do say some things sometimes. The best example of this is the Freddy kill rate, imo. The fact that he has a much higher kill rate than other killers (not just 1 or 2%, it goes all the way up to 75%) does, in fact, mean something. Denying it is as pointless as basing everything around stats and ignoring other factors that impact the game balance.
1 -
This is a terrible way to measure balance as it would show all snowball killers as being underpowered. It would show nurse as being incredibly underpowered! (heavy sarcasm)
According to stats of hooks per match the following killers will be getting major buffs - Nurse, Oni, Hag, and the Twins are all getting major buffs to compensate for their low number of hooks per match!
Killers play in different ways and use different strategies , the end goal for killers is to kill and the end goal for survivors is to survive. That is the correct metric to determine balance.
See I feel it's silly people like you that have made their way into the design team that hamper actual balance.
1 -
I'm interested in seeing how a working mmr system affects the kill rate. One of the reasons I believe the kill rate is generally high is because the current rank system is not doing a good job at separating skill and too many red rank survivors don't play that well.
Honestly between the really easy games as killer and the frustrating games as survivor, because of the huge skill discrepancy between red rank survivors, I've taken a break from the game because I'm just not having that much fun overall. I've had some really good games on both sides, but those are few and far between.
One of the biggest moments I remember where I was like "how are you rank 1?" was this bill that went down to the deep wound timer.
1 -
This content has been removed.
-
They don't balance around kill rates, those are just some stats they shared with us. Doesn't mean they don't have access to much more data than what they show.
2 -
It gives them a false impression
a dev commented saying they don't think legion is weak because "He is in the middle of the kill rates" or something similar but that completely disregarded why some killers like huntress were below him. If you took out the skillful killers like the ones who obviously don't belong under legion he has i think the 6th or 5th worst kill rate in ranking order again if you ignore the killers who are obviously better then the legion and are being held back by bad players.
0 -
They gave us those stats because it's too general to actually get anything useful from
Now if they could give us more stats we can (as a community) come up with our own theories
Stats like and can be applied to all skill levels:
Average number of hooks per killer (might tell us how the killer is being played)
Average chase times per killer (Might tell us how players use their power and mindgames with a side of survivor looping skill)
Average Gen completion per killer (Might tell us Map/Gen pressure of the killer and how the survivor play the objectives)
Perk usage (Might tell us the overall strategy for killers and survivors) -and how killers and survivors adapt to the counterparts strategy)
0 -
That's not how they work. Let's take for example the top and the bottom killer according to their killrate, Freddy and nurse respectively. They look at Freddy and ask: why is he so strong? They realise: oh! He does too many things too easily, got it.
Then they look at nurse and ask: why does she have such a low killrate? They realize oh! It's because you have to grind to get good at her, got it.
At least, that's my theory
0 -
- Nurse kill rate is like 50% and is the lowest kill rate of all killers. Devs said during that stream that hers was the lowest due to her higher skill ceiling but were happy with where she is. They are aiming for a kill rate as close to 50% as they can manage so how is her being the closest to 50% out of all killers means she needs a buff? Considering the devs acknowledged her high skill ceiling, how she's considered one of the best killers in the game and that her kill rate data is skewed because of the high skill ceiling so people either 4k or 0k with her shows they do take other data into consideration.
Balance isn't just one data set but when devs talk about balance of any game they generally keep things more simple and one or two data sets so people can see where it is and understand. Devs have given out data for the percentage of how often perks are used and several other things. They don't count games with dc's and I'm certain they're aren't using only one data set to decide on overall game balance. Just because that's the data they share in a one hour stream doesn't mean that's the only data they have.
0 -
I mean that's like saying old moris on old nurse is bad so we need an alternative, you don't need to suggest something an alternative to point out somethings bad. Quite frankly, I don't think we even need an alternative.
0
![[Deleted User]](https://forums.bhvr.com/applications/dashboard/design/images/defaulticon.png)
