Please stop nerfing killer perks
Comments
-
We know from previous stats compared to NL, that NL stats are not accurate. In reality, killrates are probably 5% higher.
0 -
I completly agree. STBFL is only really a good perk on M1 killers and they could need some extra help in chase compared to M2's.
3 -
They should stop nerfing, buffing, reworking, and releasing everything for at least 6 months and let the game breathe. There's no way to tell if anything is actually working as intended when there's a new chapter every quarter and PTB every 3 weeks.
1 -
It does not favor killers when they are nerfed 24/7. A 5th perk slot for killers would be wonderful, but then Because someone might lose a game or two, of course there going to yap, huge early example right here
2 -
Bhvr always listening to survivors and making unesassary changes, but killers say something and everybody yaps. Smh
2 -
If survivors are dying it's because there probably new since there's Alot of new people and older players leaving.
1 -
You never explained it at all.
A multi-page post where I, again, explained it to you in all ways possible says the contrary. And yet, you still doesn't know what average means.
And with that, I don't have anything more to tell you here. Yet again, go read that post where you had all the explanation you need and if you still doesn't understand it, as it was then: It's on you.
3 -
No, as usual: you state what your perception is. And then I explain why you are incorrect.
As demonstrated once again here.
6 -
Prove it.
Pull up some stats, last stats we had from BHVR show that Killers are fine.
8 -
Yes, 2.4 kill per match on average to be exact. Which in terms of MMR means killers wins 40% of their games (3K), while survivors also win 40% of their games (Escapes in 40% of matches).
The key here is that just as the game, MMR is not symmetrical. Killers need a 3K to win, survivors only need to escape themselves to win. The only exception is the "draw" condition, as even if a survivor draw with the killer (in other words, escape through the hatch) in game counts as a full escape even if their MMR doesn't move.
Some people doesn't understand that, just as what happens in fighting games (a genre DbD have a lot in common in terms of balancing, among other things) even if you get to a perfect 5/5 winrate between two characters at high skill play in terms of balancing, there is no possible way of physically guarantee that everytime those two characters play against each other they would always grant their users a 50% win rate because a lot of factors (including the skill of such users).
That's why you have to center your attention on the mathematical average, which again, in DbD is a 60% kill rate on average, which would mean a 40% winrate on average for everybody, as for the killer the MMR works as a "1v4" where you have to kill at least 3 rivals to wins, but for survivors the MMR works as a "1v1", where you only need to escape by yourself to win.
In other words, if in high MMR there is a 39% escape rate that means that yes, the game is killer sided but just by a 1% more kill rate on average than it should.
4 -
I think it's more nuanced than that, as you keep saying wins and losses which is an entirely personal perception.
I will agree with your larger point, although I feel the game should trend towards 50 instead of 60, as you need to account for Hatch escapes, which is hardly a win.
1 -
It's not a personal perception:
It's literally how the MMR works.
2 -
No, I explain to you things in detail and then you proceed to simply ignore 90% of what I said and start twisting the other 10% with mental gymnastics and fallacies.
That's what usually happens with you, as was demonstrated in that other post (and any other were you have posted, especially the one where I pointed every single fallacy you said) were I already explained to you everything I had to explain about this topic.
3 -
Nah. Killers have some pretty nutty perks at the moment.
Pop is pretty decent, Surge is still good, Pain Res and Deadlock are no brainers, Corrupt Intervention is arguably the strongest perk in the game (although I personally prefer Lethal Pursuer), UW hasn't been nerfed (which is genuinely surprising considering the mass of complaints), Grim Embrace is getting a buff... I really don't know what more you could ask for.
There is at least 1 good perk for every situation. Anti healing? Sloppy. Info? UW. Slowdown? Yeah, we've got plenty of that. Chase? STBFL is pretty good (although this one gets the wrong nerf).
2 -
I think it's more nuanced than that
Just as a quick explanation:
Yes, 2.4 kill per match on average to be exact. Which in terms of MMR means killers wins 40% of their games (3K), while survivors also win 40% of their games (Escapes in 40% of matches).
This calculation only works if the only possible outcomes are 2Ks and 3Ks. But killers can get five possible outcomes, which are 0K, 1K, 2K, 3K and 4K. The distribution of these outcomes isn't even or logical either, which is why you can't take a specific kill rate and calculate what the resulting winrate will be.
Using Batusalen's model would, for instance, result in an obviously extremely improbably scenario on a lot of other kill rates. For example, a kill rate of 40%, or 1.6 kills per match, would produce only 2Ks or lower, resulting in an average winrate of 0%. And the only way you can get an average of 0, barring the inclusion of negative values (And as far as I know, no killer has ever won -1 match) is by having no value other than 0.
IE: a killrate below 50% would mandate absolutely zero matches ever being won.
So yeah, Batusalen's model is lacking nuance. A ton of it. To the point where it's not applicable. Nightlight kind of proves that, with much higher winrates on lower kill rates. The only role kill rates play in determining balance is that they indirectly indicate the winrate for survivors. If killers kill 55% of survivors, then the other 45% of survivors escape.
Killer winrate can only be taken from the percentage of 3Ks and 4Ks, NOT from the kill rate. As I mentioned above, Wesker's kill rate is higher than Nurse's, but his winrate is lower. The distribution of outcomes varies from killer to killer, and there's way, way too many factors to include, many of which are simply not mathematically predictable.
Let alone by a model that only considers two of the five possible outcomes of a trial at a time.
6 -
Then you cannot count Hatch Escapes as a win.
2 -
To the point where it's not applicable.
So, a model that has being used in the fighting game genre for years to balance their characters is not applicable? Damn! Call Capcom and Namco, they had being doing balancing wrong all this time.
And just with that, just because you still doesn't understand and don't know what it's being talked about but still want to prove me wrong somehow, it's how you start saying things that doesn't make any sense.
3 -
"The only exception is the "draw" condition, as even if a survivor draw with the killer (in other words, escape through the hatch) in game counts as a full escape even if their MMR doesn't move."
I never said that the system is perfect. In fact, in my opinion the hatch should be removed from the game and survivors only able to either win or lose.
2 -
I think the Hatch does a LOT to keeping the Escape Rate over 30%. I don't know that you'd want to remove that.
6 -
Me: *Explains how DBD killers have five possible match outcomes, and that's why a model that only considers two outcomes is not applicable*
You: "So a model used in the fighting game genre for years is not applicable?"
Also you:
But yes, bluntly, yes. A system used in fighting games with binary outcomes cannot be applied to a game with five outcomes. If you'd read what I wrote, you'd have figured that out.
9 -
I just had 2 rounds back to back against a Nurse with
- Dead man's switch
- Ultimate Weapon
- Pain res
- X
I DONT THINK IT IS A GOOD IDEA TO STOP NERFING KILLER PERKS! NOT AT ALL!
2 -
Exactly.
I hope they cancel the change before the patch hits live, it is not needed.
1 -
A fighting game doesn't have a binary outcome, it has 3: Win, Draw, Lose.
How many outcomes the MMR system have, again? Oh yes, 3: Win, draw and lose!
That's the only outcomes that you are interested about. If you want to make mental gymnastics about the 0K to the 4K and how funnily enough the mathematical truth and fact that is "a kill rate of 40%, or 1.6 kills per match [...] resulting in an average winrate of 0%" is "an obviously extremely improbably scenario" somehow when that's exactly what an average means, just because you want to prove me wrong and you don't know how, yet again: It's on you.
Here is the abridged explanation:
Everyone understood it, you are still saying things like "Killer winrate can only be taken from the percentage of 3Ks and 4Ks, NOT from the kill rate." without obviously knowing what AVERAGE means yet, even after 2 pages of a post in the day where I explained it to you in detail.
3 -
How many outcomes the MMR system have, again? Oh yes, 3: Win, draw and lose!
But we're not talking about the MMR system, are we? We're talking about the kill rate.
Again: Your model only incorporates TWO outcomes. There are FIVE. Which is why your model diverges from reality to such an extreme degree. The MMR does not distinguish between a 0K and a 1K, or a 3K and a 4K, but the killrate does. Thus, the 4K, even though to the MMR it is not a separate outcome, still influences the relationship between killrate and winrate. This is the part that you are missing. This is your blind spot. As illustrated by:
That's the only outcomes that you are interested about. If you want to make mental gymnastics about the 0K to the 4K and how funnily enough the mathematical truth and fact that is "a kill rate of 40%, or 1.6 kills per match [...] resulting in an average winrate of 0%" is "an obviously extremely improbably scenario" somehow when that's exactly what an average means, just because you want to prove me wrong and you don't know how, yet again: It's on you.
You still do not grasp that your model mandates 0 wins on any kill rate below 50%. Not as an incidental thing, but as a hard-lined rule.
Your model states that in any dataset where the killer has a killrate below 50%, there cannot be a single trial recorded with 3 or more kills.
Otherwise the average would not be 0%.
Take a dataset of 1000 matches, with a total kill count of 1999. How high do you consider the probability that in those 1000 matches, not even -one- had more than 2 kills?
4 -
My kill rate around 80% , my survivor escape rate around 30%
Yeah such a survivorside game lol. Plus i am just doing fine with 4 perks, 5 perks would break the game.
2 -
Let’s talk about the massive amount of suicides in very winnable games first.
7 -
This game isn't survivor sided. How could you even think that w/ common sense, when the matches require 4 teammates to play well together & a killer just relies on themself? When have you ever gone on a survivor winstreak by yourself in PUGs? Have you seen any actual winstreaks by anyone not running coordinated SWFs like streamers? The craziest winstreaks ALL come from the killer side. Why is that? If they have such a hard time winning, you'd think people hitting 1000+, let alone 300+ winstreaks wouldn't be a thing.
I genuinely don't understand how anyone can objectively say this is survivor sided. Reminds me of Pre-nerf Eruption & some of the killer mains trying to defend the 25second incap on a 30second CD.
2 -
Me: "Yes, 2.4 kill per match on average to be exact. Which in terms of MMR means killers wins 40% of their games (3K), while survivors also win 40% of their games (Escapes in 40% of matches)."
You: "But we're not talking about the MMR system, are we?" followed by yet more mental gymnastics.
Have I to say anything more? That's how you work. Again, ignore 90% of what has being said, twist the last 10% with wrong logic until you can form an strawman fallacy to use as an argument. That's it. That's why I simply stop answering to your nonsense most of the time, because my only options are start pointing out each fallacy and faulty logic you use until you start being coherent or give up or simply ignore you and let you say whatever you want.
I'm going to just comment on one thing:
You still do not grasp that your model mandates 0 wins on any kill rate below 50%. Not as an incidental thing, but as a hard-lined rule.
It's not my model, I don't mandate anything, that's how AVERAGES WORKS. It's a mathematical fact! If a win is 3K or up and a killer does ON AVERAGE only 2K, it means that ON AVERAGE he doesn't win any game! ON AVERAGE! A simple concept yet so far from your grasp that you still says nonsense because of it. That, or you twist it on purpose so yet again, you can make some kind of strawman + ad hominem against me because "It's you who doesn't understand my twisted logic, so you are at fault here".
Either way, time to chose option 2.
3 -
idk what all the nerd stuff going on here is but it's almost a factual statement that this game is swf sided if they are above a certain skill level and then killer sided; any opinion stating otherwise should be disregarded.
3 -
It's not my model
It IS your model. It is what YOU use to come to the conclusions you do.
that's how AVERAGES WORKS
It is not. It is very clearly not. For all the reasons I have specified. You fail to grasp this.
If a win is 3K or up and a killer does ON AVERAGE only 2K, it means that ON AVERAGE he doesn't win any game! ON AVERAGE!
And this is where the problem is screaming in your face and you don't understand it.
An average of zero in an environment without any negative values MUST mean that there is NO value above zero.
If they don't win any game 'on average', that MUST mean that they win NO game, EVER. That there is NO possibility for there to be a win, at all.
I ask you again to think about this scenario and what it means for your model:
Among a dataset of a thousand matches, the killer scores 1999 kills. What do you think the probability is that in those thousand matches, not a single one resulted in anything more than a 2K? Not -one-!
3 -
Does that change the FACT that Killers are still getting those Kills?
5 -
You must die to him all the time. Smh
1 -
Nah, i am acting like i am winning most of my killer games. And that's also correct most of killers as well. All rates are favoring killers.
All of your suggestions would just kill the game. Because all of them will make game unplayable for other side. And the question is why survivors will play your game when they have zero fun and zero chance to win?
Exception of top SWFs, i doubt most of people even would touch to play Survivor lol.
4 -
I don't play Killer too much anymore, as that's all I play with DBD but I haven't lost on a non-meme build since Alien.
Now, I've probably only played 250-400 games in that time, far less than what I used to average, but it's not insignificant.
That's mostly why I'm losing interest. Survivor is miserable, Killer is just not challenging FOR ME anymore and the designs are getting boring.
4 -
You have to keep in mind that BHVR counts the killer getting a single kill as being a win...which the community has generally completely disagreed on. The community has generally come to a consensus that we count a 3K as a killer win. Those "win rates" that BHVR posts are a 1K+
2 -
Playing killer honestly gets boring when you can't do fun builds cause they get nerfed
2 -
Playing survivor honestly gets boring when you can't do fun builds cause they get nerfed.
5 -
Hello! A reminder to please keep comments civil and respectful. Thank you!
2 -
BHVR never posted winrates to my knowledge. The winrates I'm referring to are the percentage of matches that result in a 3K or 4K, as presented by Nightlight, so we're talking about the same 'wins'.
3 -
I'm around there as well.
I dnt even play well, my perks play for me as killer.
1 -
Do we know that hatch escapes are considered escapes when it comes to kill rates? I really am not sure how this is counted because at least in terms of MMR it counts as neither a win or loss.
1 -
You know you can say the same for survivors, right? Lithe, WoO and Deja Vu come to mind.
I agree though, that some perks have too big of an impact and make up for a lack of skill. On one hand, we need these strong perks because they create this flawed yet existent balance. On the other, it would be nice not to feel like you lose against loadouts rather than players.
Edit: Thinking about it more, it might actually be good the way it is. A lot of the time our losses feel so crushing that we look for something to cope with. So in that regard it might be good that perks are so strong that we can blame them instead of our own mistakes.
5 -
Hatch is not counted for MMR, but I do think its influence is still considered when talking about kill rates. I wager Nightlight at least does not consider whether an escape was a hatch or not.
2 -
And what perks are those? Just curious, because with my main it's just sweaty after sweaty match so I would love to use those perks that allow me to do nothing like I'm in low MMR playing against SoloQs with no game sense at all...
1 -
Nightlight counts hatches as escapes. If you specify that you gave a survivor hatch, it counts as a kill.
0 -
I think it's very rare that only one person gets out through a gate, so yes, I think Hatch Escapes are counted in the kill rate.
1 -
I've seen a lot of very confusing and/or misleading math being posted here, so I thought I'd chime in. At a 60% kill rate, the lowest possible win rate (3k or 4k) is 20%. The highest possible win rate is 80%. The average win rate would be 50%.
6 -
So it distinguished between an earned hatch and a given hatch, huh? Interesting.
0 -
Sloppy, Infectious, Deer Stalker and Hysteria.
Bubba is doing well with it so far :)
0 -
I think you need to practice survivor more. I do agree, more perks could break the game, but again, for the killer, its a 4 vs 16 and for each survivor to have OP second chance perks its kinda annoying sometimes.
0
