A reddit user did an experiment. Let's discuss the results.
Comments
-
this is devolving really fast can we calm down?
2 -
A killer doing a 1K is a loss, so 1 survivors killed per match on average = killers lose 100% of their games on average, or in other words,
win 0% of their games on average
(no negative winrate, mate).You're not thinking about what you're writing here.
For any group to win 0% of their games on average, they would need to ALL win 0%, because, as you agree, there is no such thing as a negative winrate.
If you have five killers with the following winrates
1 - 50%
2 - 20%
3 - 10%
4 - 0%
5 - 25%
Then the average winrate would be (50 + 20 + 10 + 0 + 25)/5 = 21%.That is how you calculate an average.
Now, let's see a set of five killers that have a 0% winrate on average:
1 - 0%
2 - 0%
3 - 0%
4 - 0%
5 - 0%
Then the average winrate would be (0 + 0 + 0 + 0 + 0)/5 = 0%.Let's do the absolute minimum to deviate from that set to see what that does to the average winrate!
1 - 1%
2 - 0%
3 - 0%
4 - 0%
5 - 0%Now the average winrate would be (1 + 0 + 0 + 0 + 0)/5 = 0.2% =/= 0%.
So when you claim that on a 50% KR, the killers would win, on average, 0% of their matches, you do indeed claim, with absolute certainty, that NO killer in the game has managed to win a singular match.
Because if they had won a single match, their winrate would not be 0%, and as a result, the average winrate could not be 0% either.
Conclusion: Yes, we have a winrate of 50%, a loss rate of 50%, and a draw rate of 0%. Which, still,
means a draw rate of 100% on average.
Legitimately, dude, you are getting confused. You're talking about the outcome of the average match, NOT average draws here. If someone has a draw rate of 0%, you CAN'T have 100% draw rate on average.
Just please think about the distinction between these two things:
The outcome of the average match
vs.
The average of all win rates/draw rates
4 -
"Woosah, woosah" rub the ears - Martin Lawrence in Bad Boys
I love seeing stats like op posted. I just wish there was more we could see about the matches. Or even a couple to watch.
2 -
what I'm getting from this is that you, too, agree that your previous assertions that "50% kill rate means 0% win rate" was erronious, and that things are actually a whole lot more complicated. glad to see that you've seen sense.
absolutely. i love you, @totallynotamegmain you're the cutest non-meg-main i've ever met.
1 -
😊
1 -
If 2 survivors where killed per match on average, and according to the win conditions a 2K is a draw, then killers draw their matches on average.
No, we're outside of math and just talking about how words are used in English. That's a .500 win rate, an even split, etc. This is a game that has draws, saying that the result is a draw on average implies that most games result in a 2k.
Can't I?
You can't, and your examples are bad math.
Survivors win condition: Not get killed. Ergo, a 60% killrate on average means 40% of survivors win their games on average.
Killer win condition: Get a 3K. Ergo, if a 60% killrate on average means 2.4 survivors killed per match on average, killers are doing mostly 2K, drawing those matches, while doing 3K 40% of the time approx. Therefore, they win 40% of their games on average.
The second part remains incorrect because you refuse to engage with the idea of an uneven data set. 2k is the least common outcome in DbD.
This carries through to your other examples:
Survivors win condition: Not get killed. Ergo, a 75% killrate on average means 30% of survivors win their games on average.
Killer win condition: Get a 3K. Ergo, if a 75% killrate on average means 3 survivors killed per match on average, killers are doing 3K 100% of the time. Therefore, they win 100% of their games on average.
What does the phrase 100% of the time mean? It means always.
You seem to understand how the killrate is derived. The average is 75%, meaning there is a mix of game results that over many games result in an average of 3 killed in this hypothetical. You then turn around and use a totally different definition for what winrate would be.
In this hypothetical and the framework for wins, we'd take every game played, how many where 3k or greater, and then we'd have the average.
You could say the most common result would be a killer win, which is already true
Survivors win condition: Not get killed. Ergo, a 50% killrate on average means 50% of survivors win their games on average.
Killer win condition: Get a 3K. Ergo, if a 50% killrate on average means 2 survivors killed per match on average, killers are doing 2K 100% of the time. Ergo, they win 0% of their games on average, as they draw 100% of their matches on average.
Again, that's not what an average means. If we keep with your simple math, on average they win 50% of their matches that are not draws (i.e. the average winrate is 50%). More realistically, we'd need more information as you could have a radically different winrate from a 50% killrate.
Is it clear now? Or shall we keep going with the fallacies and mental gymnastics?
It's clear that you are committing multiple mistakes. This starts with an incorrect definition of what average can mean and a misunderstanding and refusal to even engage on the concept data sets.
3 -
No, we're outside of math and just talking about how words are used in English.
The guy doesn't even know what a rule of 3 is, and even when having the dictionary definition written on top of an example still doesn't understand what an average means. Keep pretending to fight basic math while making language-ambiguity fallacies.
I will only say this:
and refusal to even engage on the concept data sets
Me two doritos earlier explaining to you how data sets are crucial for individual killer balancing, but not before having an average balance target because of the asymmetrical nature of DbD:
So, even if "uneven data sets" can help identify if an individual killer needs balancing, it doesn't work in terms of overall balancing and will not help balance killers against survivors.
[…]
When we reach that point, now we can use the kill distribution of each killer to balance those that are overperforming or underperforming until we get a balanced average […] But we can only do that once we have reached that target 60% killrate, and that is just what BHVR did.
So like I said dude, whatever makes you happy.
1 -
The guy doesn't even know what a rule of 3 is, and even when having the dictionary definition written on top of an example still doesn't understand what an average means.
I put two different definitions to show that the word average can mean different things. Mathematically it can be the mean, median, or mode, though it is generally the mean when people talk about it. I'm just asking you to choose one and stick to it.
Rule of 3: There are more variables than you are accounting for which is why you don't have enough data. But being you think you can use that to solve this does explain some things.
but not before having an average balance target because of the asymmetrical nature of DbD:
Who said they shouldn't have a balance target? It's simply that a 60% target is not necessary to achieve the 40% winrate you emphasized earlier and that is not the outcome it has in the game right now.
So like I said dude, whatever makes you happy.
You've also said multiple times that you're done with the thread. I'm not, I think you're wrong, and the more this discussion goes the more things you seem to be wrong on, and I'm still trying to get you to either explain why you are correct, and if you are I'll happily admit you are correct, or for you to see the error in your reasoning.
But why you're here:
Multiple killers have the exact same kill rates, do you think that means they have the same winrates?
4 -
The first paragraph was my entire point (on the topic of averages), that an average can have different results depending on how it is made up. Thank you for finally agreeing with me after all.
The second paragraph says we did none of the math…
What was this entire post then? Did you post a refuttal proving my methodology or arithmetic wrong? No? Then it would stand to be accurate, or you accept it as accurate, otherwise it would have been easily disproven. I mean if I made such a simple mistake as 1+3=25, I'm sure you would have jumped on that. That would mean our (or at the very least, my own, I didn't double check them so I take them with a grain of salt as well) point(s) is(are) supported by hard evidence.
To the last sentence I say 'every accusation is a confession'. Since you have nothing to fall back on, you have baseless accusations, against claims defended by math.
And again, I've said they already explained the why, "horror themes". It has to be close enough (for the game to be fun), but also lean Killer sided (to match the aesthetic/tone). I don't even disagree with that being the case. I do think a 60% killrate is fine, but I also acknowledge that makes the game Killer sided.
In general, what I care about more is people above and below the target balance rate, and more importantly, the why. If someone has sub-60% killrates, they likely aren't good, and above-60% killrates means they are likely better. Pre 6.1 I won the same as after 6.1 (because my playstyle didn't involve letting Survivors proc DS/UB), but the killrates on average raised. Since they deleted all meaningful anti-tunnel in 6.1 (and people slowly caught on to the 3-gen perks), weaker Killers started crutching on tunneling (and eventually 3-genning) to kill/win, and there was nothing soloq Survivors could do about it. The immediate 'why' was deletion of anti-tunnel (and eventually 3-genning).
4 -
The guy doesn't even know what a rule of 3 is, and even when having the dictionary definition written on top of an example still doesn't understand what an average means. Keep pretending to fight basic math while making language-ambiguity fallacies.
You've consistently been avoiding talking about this, though, Bats.
The way you try to calculate the 'winrate on average' is not how you calculate an average. What you have been talking about is not the average winrate, or the 'winrate on average', it's just the outcome of the average match.
These are not 'language-ambiguity fallacies'. No one is being ambiguous with their language.
You are getting your terminology and calculations confused.
Again, I urge you to think about the distinction between the following two things:
- The outcome of the average match
- The average winrate
Because I feel that that distinction is tripping you up.
4 -
Actually this deserves its own thread, will update in a bit when I have a break at work
Post edited by ratcoffee on0