DBD is slightly killer sided (and that's great!)

Oh, I totally agree with you that the game absolutely doesnt aknowledge any teams existing outside the swf lobby. Maybe shouldve said it out loud ;P

No mention to any changes on matchmaking, or MMR, or anything team related. And the only mention on the dev logs are in this one before the patch was launched and that article you screenshot was written:

If you follow the link from this twit, that is supposed to be the same one that you found in Steam with an included explanation, you would see the link is broken as the article was deleted:

There is also no dev log for November in the list, jumping from October 2022 to January 2023. So, if you have a post of that exact same article / dev log here, post it, as I wasn't able to find the original source (I didn't look at Steam as my thought was that it would has being also deleted from there).

But in conclusion, the only proof that such a system has ever being implemented is the Wiki, as they data mine things from the code of the game itself. And according to the Wiki, this system only affects survivors grouped in a SWF squad, not when you are SoloQing:

And about my math, seriously, you guys should look what averages and percentage are and represents. It's not oversimplifying, it's pure math (60% of 4 = 2.4. 100 matches = 400 survivors, 60% of 400 = 240 survivors killed = 2.4 survivors killed per match in 100 games).

And you can't balance a game based on the skill floor of the players but the skill ceiling for multiple reasons, starting with if you balance the game based on the players with low skill that doesn't know how to play the game effectively, those who do know would crush everybody else.

And as I said, what Nightlight is showing is that even there survivors are winning more than killers on average right now for a small percentage.

You did not explain anything. You keep refusing to answer. You keep refusing to consider the math.

Do you? Because you seem to be confusing it for the median.

Switching tracks again. Not that it fixes anything, because whether we're talking about an individual killer or killers as a whole makes not a lick of difference for the purposes of this conversation. Which is probably why you bring it up now to begin with.

I'm going to ask this question for the third time, because you keep dodging it. I know why you're dodging it, but I just want to make sure it's clear to you why you're dodging it. That you don't have any more pretence about having a proper understanding of this subject.

How do you intend to raise the kill rate, while lowering the killer win rate?

Can you have a 40% win rate and a 60% kill rate? Yes, it is theoretically possible.

The problem is there is a natural inverted bell curve in the % of games that fall within the 5 outcomes for the killer. The highest amounts being 0k and 4k while the lowest is a 2k with 1k and 3k being in-between. This isn't an issue of balance, it's inherent within the current game design. There is no way that I've been able to work out to make a bell curve with 40% of the match results falling in 3k and 4k, while still maintaining a 60% kill rate.

The reality is what many have been saying forever - kill rates are a bad metric.

I already answered all his real questions. The ones that I didn't answer are his mental gymnastics and tries to distort what I said. For example:

I say: We get all killers across the board to a point where that kill rate on average is achieved by either general buffs or nerfs to killers and/or survivors

His answer: Please explain to me how you intend to raise the kill rate via killer nerfs or survivor buffs. Or conversely please explain to me how you intend to lower the killer win rate via killer buffs or survivor nerfs.

When have I said that? Nowhere. But he interpret it like that, because the other way around won't fit his narrative as obviously you get kill rate average higher by buffing all killers or nerfing survivors, and vice-versa if what you want is lowering them. And when you already got the 60% kill rate, you make every killer have a 2.4 average of survivors killed per match by buffing or nerfing every killer individually, as every killer has it's own powers and stats so they need individual balancing. That's what I said, and that's what he refuses to see.

So, again, I'm not going to answer any of his fallacies and mental gymnastics because I already responded to his other questions with the logic for it and the math behind it. I'm not going in any circle, I'm sticking be the math and facts. If he still insist in not accepting it, it is on him.

First, it's 40%, not 50%, and as I has already explained in other post, a win for a killer is a 3K according to MMR. So, 60 kill rate = 40% win rate for survivors and for killers (even if for this to be true, every single killer need to be as close as that 60% kill rate as possible).

That's why BHVR wants the kill rate around 60%, as that is half the job to balance the game done. Maintaining that 60% they only need to balance each killer so his individual killrate is around the 60% and then every single one of them and the survivors would have a 40% win rate ON AVERAGE. Meaning, no every single survivor would just win 40% or every single killer kill 60% in reality, there would be people that do more or less than that, but ON AVERAGE there would be 2.4 survivors killed per match.

And last time I explain it.

See? I explain, in detail, how you get to a 60% kill rate and that would be the only way that lead to an equal 40% win rate for everybody, and his response is:

Should I really need to explain it to him again? Should I make him an sketch to see if he understand it better? Or should I just assume that however I explain it to him he would still not understand for the simple reason that his bias makes him not wanting to understand and move on?

So I'm not going in circles, I'm just taking option 3. He has the answer to his question posted multiple times already in this thread, he can re-read them the times he needs. In fact, the post above his has the answer simplified.

At this point, yes. Especially since we've only seen anything remotely resembling an actual answer to his question way after the fact.

"I explain, in detail, how you get to a 60% kill rate and that would be the only way that lead to an equal 40% win rate for everybody"

No, you don't. You assert it, but you don't explain it. Because you cannot explain it. You keep dodging the question, asserting that you don't need to answer it, because you know what'll happen if you try.

Let me try and make an analogy for what is happening here, and hopefully you'll get it.

There's a group of ten people. One of them has no arms. This means that, on average, the people in this group have 1.8 arms.

-I- assert that, given only the average number of arms, we -cannot- deduce what percentage of people have fewer than 2 arms. It could be one person with no arms, or two people with one arm.

-You- assert that you know, for a fact, that the 1.8 average -must- mean there are two people in the group with one arm, and that there could not possibly be one person with no arms.

Curtain rises, we see the group. Clearly, one person has no arms, while the other nine have two arms.

You -see- the line-up, and turn to me, and tell me, to my face, that you are still correct about your assertion that there -must- be two people with one arm each, and that my insistence that this group with an average arm count of 1.8 does -not- have two guys with one arm each means that I simply do not understand it.

...Honestly we don't know why we bothered trying as this is the point where both have lost the point of the discussion.

For the sake of [BEEPING] closure. From our understanding:

Batusalen: You are arguing that increasing the kill rate for killers will give both sides a 40% win rate average eventually. Your saying that first buffs to the kill rate then nerfs accordingly on an individual basis to effectively try and hammer the win rate to 40%.

Firellus: You wanted an detailed answer on how increasing the kill rate will lower the win rate, probably in a more immediate time frame. At the moment, we not honestly sure what you want. You and the OP have shown that a 60% kill rate does not always equal to a 40% win rate and want to know how Batusalen's method would work.

Are we understanding this correctly?

Four our own 4 cents to at least be semi on topic: We think how sided it is depends on how many sided maps there are as most our harder matches are determined more by map/killer combination than survivors themselves. (Barring occasionals)

"No, in that case, what I would be saying is "the average shows that between all of them, 2 arms are missing""

No, that'd be the equivalent of -only- talking about kills, but the issue is specifically about your assertions about wins.

Your problem is that you do not understand the distribution of outcomes.

Your 'explanation' is built in a system that has only 2Ks and 3Ks. You think that 'average' is a magical word that will, by itself, eliminate the weighted distribution that DbD has, even though Nightlight shows that this is not the case.

You seem to be under the misapprehension that, on average, the kill distribution will fall into a neat, even distribution of outcomes, but we already know it won't. It doesn't.

The question I asked is meant to illustrate that, but you keep dodging it. You keep saying 'I answered it!' by showing you stated again that you'd just set everything to 60% kill rate, but you didn't think this through:

If the current win rate at 53% killrate is 46%, what do you think is going to happen if you add in more kills?

If you buff killers to make their kill rate increase, do you not think that this will make them win more, too?

Let me try and illustrate it another way.

You state the following:

60% kill rate is 240 survivors killed over 100 matches. This is, on average, 40 3Ks, with the rest being 2Ks, and thus a 40% winrate for the killer.

This is what it would average out to, correct?

Now I'm going to try to run this with the current Nightlight stats:

53% kill rate is 212 survivors killed over 100 matches. This is, on average, 12 3Ks, with the rest being 2Ks, and thus a 12% winrate for the killer.

This is what it would average out to, correct?

To address a couple points. You CAN increase the kill rate and have the win rate drop. You can see it in the spread sheet i added above. The thing is you can only do it a few percentage points before the natural distribution of kills gets so out of wack that you simply will not see it in normal gameplay.

Look at Onryo for example. She has the closest numbers to a 60% kill rate and 2.4 avg kills, but her winrate is around 57%. You'd have to bring down the win rate around 16% without dropping her kill rate more than a percentage point and that simply isn't going to happen.

The other thing is "wins". I'm a big supporter of making up your own win conditions, but the reality is it comes down to what feels like a win to the average player in normal gameplay. For killer that's a 3k or better. Survivor is a little different depending on whether they play for the team or themselves. The game is aimed at them playing for themselves so that's the best approach to this, and what feels like a personal win to the average player is escaping via gate or hatch.

So what I listed at the end of my last post is correct, yes? That is your claim? That is how you believe the game to function?

"With this cleared, the answer to your question is "ON AVERAGE, yes.""

If the math that I outlined is indeed how you calculate average winrates from killrates, think about it: What happens to your 'average winrate' projection, using your calculation of it, if the killrate is lower than 50%?

And what would this value require, unconditionally, in order to be attained?

You're the one who said that MMR determines what is a win or loss for both killer and survivor, right up until you realised that the hatch not being a win could drop survivor winrate under that of killers.

Then you were all "a win is whatever I feel like".

Hey thanks for at least looking through the work. Yes in the screenshot, the numbers in the "Ideal" are incorrect because I had set the kill rate to 60% to test some things elsewhere in the sheet. The screenshot was not to prove any numbers, just to show I'm not casually doing napkin math on this. The numbers have long been reverted to true calculations, and this had no effect on my conclusions.

As far as rounding, it does not matter if the AK per match was rounded to 2 decimal places for the sake of keeping it easily accessible visually. Largely because AK per match doesn't matter. It is simply another way of expressing kill rate and is a redundant stat that mostly serves to satisfy curiosity. That's also why it currently sits at the end of all the hidden calculation cells in the sheet.

When I get home from work I'll post functional shots of the whole sheet and go into depth a bit on my points on how this all works.

Yeah, you explained that later when I pointed out that the hatch means survivor win rate isn't what you said it is.

But that's -after-:

"I mean that it is the MMR what dictates what is a win or a lost for both killers and survivors."

And you also didn't argue why this only goes for the hatch and not for a killer's 2K.

Also still curious about your killer winrate projection with <50% kill rate.

"because a killer doesn't receive the same rewards for getting a draw (2K) than winning (3-4K) in any aspect while survivors do?"

That's not true at all! You can even score more points with some 2Ks than with some 3K+.

"And it still requires the same skill to kill survivors while getting hatch can be as easy as it spawning at your feet even if no gen was done?"

How is this an argument for hatch being a -win-, though? If anything, this would be an argument for the hatch -not- being a win. If it negates the entire match, it's not a win. You can't win if you don't play.

"And I really have to explain to you what having less than 50% kill rate would mean? Can't you just pick the percentage you want, calculate what that percentage of 4 is and see how much survivors would be killed ON AVERAGE per game?"

Alright, buddy, let's talk about this.

If a 60% killrate MUST mean that on every 100 matches, 240 survivors are killed along a '40 3Ks, 60 2Ks' distribution, thus producing an 'average winrate' of 40%...

Then your model also claims that on a 49% killrate, 196 survivors are killed along a '4 1Ks, 96 2Ks' distribution, on average.

This would, according to you, produce an average winrate of 0%.

Not 0.12% or 0.8% or anything of the sort.

0.0000%.

The ONLY way to get an average of true 0, without the inclusion of negative values (which don't apply here, no one has a negative winrate), is by never including any value higher than 0.

If ANY trial, anywhere in the dataset, were to ever produce a 3K or higher, you already could not possibly produce a winrate average of 0%.

According to Nightlight, there are -multiple- killers with a sub-50% kill rate.

And they ALL have recorded wins.

And if you've played DBD for more than a week, you'll probably have witnessed one of those wins yourself.

This proves your model to be incorrect, and your assertion that certain killrates guarantee specific winrate averages is also incorrect.

Which is what Ratcoffee explained to you from the start.

I knew you'd just pull out the same dodge.

Just handwave it by saying 'but averages!', even though I explicitly explained why that cop-out doesn't work here. You have no idea what an average is, or how it works. Whenever your model comes up short, whenever there's any kind of fault in your math, you simply say 'you don't know what an average is' as if there's some missing datapoint that's magically going to drag the average over to where you want it to be.

And when I show you, explain to you a situation where that missing datapoint -cannot- exist, you just retreat back to not thinking about it and just saying 'but muh average'.

I present you with your logic, I run the math, you don't like the outcome so you just say 'average' like it's a magic spell, and bail out.

I'd respect you more if you just admitted that your calculation is wrong, instead of this thoughtless deflection.

Listen, dude...

You presented your model. You've explained how you -think- these statistics work.

Ratcoffee has explained that your thinking is incorrect. You have made a mistake in your calculations.

I have explained and shown that your model does not work.

At this point, it's blatantly clear your model does not match reality. It does not match its practice, or its theory.

Your response is to reassert that your model is correct, thereby claiming reality is wrong, as opposed to -you- being wrong.

At this point, you are either thoroughly dishonest or thoroughly delusional. Take your pick.

Here's the full spread sheet as promised, sorry it took a bit longer that anticipated, work sucks.

So anyway, you don't need a 60% kill rate to achieve a balanced game. In fact, you realistically CAN'T achieve a reasonable balance with that as that sets the individual survivor's win rate at 40% which of course to be balanced, needs the killers win rate at 40%, which because of weighting of each outcome, means that 1) The survivor team win rate would be so low that it would be an awful experience and 2) getting anything in real world that looks like the curve you'd have to have would be impossible.

This is basically what the 60/40 curve looks like:

That is just not going to happen.

Even perfect balance isn't going to happen with just nerfs and buffs as you basically cannot have any 2k outcomes:

My initial thoughts on balance was just to balance the outcomes, so every outcome has the same probability, but you ened up with a pretty significant gap between the killer win rate, and the individual survivor win rate. so while it's balanced against the survivor team, playing survivor ends up being considerably easier to get an individual win than playing killer.

Really the Twins data here provides for one of the most balanced realistic curves I've seen, even beyond my own tinkering of trying find the most realistic balanced outcomes.

It's "where did the survivors spawn" and "how many skill checks did the gen jockey randomly get" sided.

This post just screams skill issue to me im sorry but im saying it how it is.

"The game is so killer sided it is not even a question any more."

Throughout the lifespan of this game for an average killer in terms of strength, it has been survivor-sided until recently when it has become arguable (which is the most balanced this game has ever been). The previous metas has been circle of healing until 2023 combined with old Dead Hard making every chase against a competent survivor a 3 hit chase so pressure evaporated instantly, small pp build before that (DS+Unbreakable) where survivors could consistently take 2 hits for eachother when combined with old dead hard mid chase after being unhooked. and before that we are entering the times where DS was an extra life and juggling was the only way to deal with it when i started playing dbd just before instablind flashlights were removed. This is not even mentioning the fact i have seen games where 4 gens have popped off the start of a game when BNP use to compete gens instantly (without that skillcheck minigame) and toolboxes had so many charges they may of well have been a base gen speed buff.

Contrary to popular belief this game has NEVER swung in either side's favour from year to year, it just use to be broken.

I think the game has gotten less fun on survivor I will give you that and I have made posts about this and the reason i think this is the case.

I want less killers like the ones we are getting — BHVR

Declining quality of killers and the survivor experience — BHVR

But if you honestly think anything other than this game is in the most balanced state it has ever been compared to its history you are just wrong and looking through some rose-tinted glasses.