61% Kill Rate. Sure, But...
See, here's the thing.
BHVR wanted kill rates higher. That might be a bit too high, but I guess this was aimed at the killer/SWF matchup.
Solos need help, but that's a different discussion.
The problem I want to address is survivor rewards and scoring, which seem to be designed around the old '50%' paradigm.
In a nutshell, survivor 'pipping' is based heavily off escaping. One you start hitting Gold or higher, it's entirely possible to depip in games where you die and quite challenging to pip without escaping. And if you get a camper/knockout slugger/Twins etc. it can be nearly impossible.
What the meta shakeup did, in actuality, is make it quite a bit harder to get to Iri 1 on survivor.
I'd prefer to see:
- Credit given to survivors for activities done whilst on hook, both in terms of BP and emblems, much as in the same way as you get credit for stuff done during chase. This would at least encourage people not to suicide.
- Some numbers shifted away from escaping and into other activities, such as healing, doing gens and evasion.
- The breakpoints decreased a bit to make losing games less punishing.
Comments
-
The 61% number means basically nothing anyways.
2 -
It's inflated by newer players, sure. But it still does represent a pretty hefty increase to the median kill rate.
I'm okay with the killer changes, but I do want to see some love given to survivors too - and it's very frustrating to bring a flan or streamers and then find yourself getting less than 15k BP on a 75% queue bonus (and also depipped) because you got Bubba'd.
10 -
tbh yeah, all-around more BP economy would be sweet
0 -
"only the numbers that fit my bias/narrative mean something." all other numbers mean nothing.
10 -
Not so much 'more', more like 'structured differently'.
I really like the idea of rewarding people for sticking it out on hook, because holy crap do people just suicide or DC on a dime.
Be nice.
What he's saying is that the median kill stats BHVR gives us can't be taken as gospel because a lot of that comes from extremely new players (killer has an easy time of it at low MMRs). This is 100% correct.
1 -
Or just remove depip as a whole. Its all just a play time isnt it?
No need to make people more frustrated when they depip for something like teammate DC at the start and making it 1v3
And for people saying it will make easier to reach rank 1, so what? Its just 1m BP and you earn several millions just to getting there by playing normally.
I know you had some good ideas but this would be the most simple one without need of messing with emblems and scoring events.
5 -
I do think the depipping is the worse part about the grades system on both sides of the fence, because it just encourages a try-hard run until you reach the plateau and frankly I get bored after 2 or 3 games of try hard killer or survivor sweat. Makes me not wanna try fun things or new perks/killers i haven't put much time into.
5 -
It’s called "stats that result from combining every single possible factor are next to meaningless". Every single skill level, all regional playstyle differences, every killer, solo/SWF/any mix thereof, all builds, all maps, all tile/gen/etc spawn RNG, etc. Literally one of the most fundamental principles of statistics is that you need to control for other factors so they don’t impact your results when testing any given factor.
But yes, go ahead and make it about "us vs them" stupidity I guess.
1 -
With a caveat, of course..new players were playing before the patch, and new players are playing after the patch. The increase will undoubtedly show an increase at all mmr levels. Like, without a doubt.
It was more a dig at that poster, though - he spends a lot of time making false narratives.
2 -
I started using NightLight.gg recently. Granted my personal escapes be below the 39%, but the overall was 61/39. Those stats are for overall amongst all skill levels.
2 -
Please take a statistics class if you’re going to try and act like a smartass.
- Nightlight is a self-volunteer-submission aggregate site, which is susceptible to bias of only really being used by people who know of and care about stuff like that in the first place.
- It merging all skill levels (as well as every single other factor ever) is exactly why the data means next to nothing.
0 -
Didn't mention or even imply "us vs them."
I'm disappointed you say that, too. I've been very consistent in my posts about balance for survivors and killers.
1 -
I think, reading between the lines that 61% number is very troubling. Since at high MMR we still see killers complaining about the strength of survivors, that must mean that at low MMRs, survivors are getting killed at very high rates to more than make up for the higher mmr experiences. This is the only way that all the feedback makes sense, if you assume that people aren't just fabricating feedback to get their way.
As someone that lives on the below average spectrum of survivors, I can say with certainty that my escape rates are way lower than 39%. And this is pretty dissappointing, after a lot of matches, the game just feels futile now. That lack of hope / despair, is not what I am looking for when I play games. I want the expecation that if I play well I have a reasonable chance of success. The game just doesn't feel that way anymore to me.
1 -
Nightlight has a bias toward above average MMR and English speaking countries, but it also does seem to correlate pretty well with actual overall stats in a lot of cases. For instance, its perk and character pick rates are based on stats from opposing players, not the uploading player, so for example its global stats page is tracking how often opposing survivors pick, say, Dead Hard from all the matches killers uploaded or how often killers played Pig in all the matches survivors uploaded. That eliminates a good chunk of the potential bias you might otherwise get from, say, a Pig main uploading 50 matches every night because its tracking what that player’s opponents used and not what the Pig main used.
So the stats aren’t a perfect match to the overall global numbers in the game but they’re almost certainly better correlated to the overall game than typical anecdotal stories and personal data you might see posted.
2 -
You are exemplifying why education is so important. You claim dumbassery about cherry-picking when you don’t even understand controlling for factors is important in statistics. Rate of education in the US have gone up over time. So did the number of people in the US dying. Your logic is like saying the increase in education is killing people.
Please take a stats class before you potentially endanger someone in the future with your ignorance.
0 -
Per BHVR, few killers/players are at high MMR; most players are in the average (middle) zone. It wouldn’t make sense that higher players have enough of a presence to skew any statistical data, as their placement is generally an outlier—and even more insignificant than players at low MMR.
2 -
I don't think you actually read people's posts. You just need a sounding board to bounce whatever ideas you are thinking off of.
You seem pretty hung up on defending killers (and nurse) no matter what, but like for the strangest things. Killers were undeniably buffed in many ways. Survivors weren't. The kill rate went up. My experience at high mmr shows a higher kill rate on average, as do the posts of basically everyone on this forum (except you, apparently). You don't need a stats class to understand what that means..?
Really hoped you wouldn't push it with the name-calling, though. Now the forum mods will come and shut down this post.
Nice example, though! Next time try to come up with one that is actually accurate and/or relevant.
3 -
You forget one thing. 61%, across all ranks. Including new players + the incredible high amount of survivors that throw the game by suiciding on first hook + the amount of matches where people try to complete challenges like escaping through the hatch, or only using SC, or hiding next to the killer for 2 hours, etc. All those things affect the kill rate negatively.
Once that calms down (it will calm down, right?), the kill rate will be much lower than its now.
0 -
ITT: Why a statistics class should be required in education
0 -
I have statistics. 61% is VERY worrisome. It means most of the survivors are really suffering. Taking into account (bad, but still present) MMR, this means that killers are climbing higher and higher with abnormal rate - suggesting something is really wrong with the game for 80% (bell curve) of players.
There could have been an alternative theory why it isn't so bad - and that would be if small number of people could pull the data very significantly. But that is not possible in DBD. There is no match result with -50 killed players. Or -50 rescued. If MMR is (loosly) balanced around 50% (kill 2 and your MMR does not change) but you still get 61% killrate, that means that (on avarage - different MMR can have different experience, but it still averages out somewhere else) killers get more than 1 free win that they shouldn't have been given every 10 games they play.
Taking into consideration that BHVR does not count statistics when someone DC's (as they stated before) - and people complain about DC's all the time... That's total terror and I truly believe some people that say the game is no longer playable.
2 -
Why would people ragequitting be factored in? Of course that will throw the match result into killer favor. The issue is the nonstop immature suiciding on hook is factored in because there’s no easy way to count that out. Which also throws a game.
0 -
I am pointing out that 61% is very worrisome. It's global stats inside MMR system 61% should not be sustainable in "balanced" MMR (where under slight simplification 1 win and 1 loose results in 0 MMR change). It would require constant and large stream of new players.
61% means the game is currently broken
1 -
'Most' is misleading here.
See - the way this game works makes killer very intuitive initially whilst the survivor role takes a bit longer to understand, as it's a team role versus a solo role.
This means that, for very new players (who will make up a lot of this stat), kill rates are going to be massively inflated. It's the same reason that Cenobite and Pig have such high kill rates and Nurse is, by these same numbers, the weakest killer in the game.
It's only 'worrisome' if you think this game should be balanced around brand new survivors, which would be absurd.
0 -
But that is false argument as well. Most gamed copy bell curve - there are very few experts, but also very few beginners. So the avarage player creates all the statistics... If you get killrate somewhere within 50%+-5%, you can say X extreme MMR is pulling all the stats. 11% is way too much - it means stats are normed around killers winning the games - which might not be the case in some specific MMR. That's how statistics work
0 -
...What?
How on earth is that a false argument?
This isn't a bell curve situation. We aren't evaluating a group who are all starting on the same level. If I were to map this out on a graph, it would likely look more like a gradually descending slope, with kill rates being excessively high and then slowly evening out.
If you wanted actually useful stats here, you'd need to remove the steep initial part of that slope, otherwise - by your logic - Nurse would need a buff.
0 -
1, I am talking about player's MMR distribution. You are talking about killrate/MMR. If you say that it should be balanced around far edge on your graph, I am saying you want to have fair matches for 1% of player base and 99% of people to have unfair matches (in this specific case OP killers). You can't balance only for 1% of best players. Most people will never belong to 1% of best players at any time of DND history.
Whats's worse is, that even that 1% that could result in best players (to get fair match) will not invest 5000 hours just to start have fun/balanced matches. As time goes on, there will be less and less survivors, because people tend to slowly leave even best games - and there won't be any new survivors if they will get stomped 100% of time.
2, I am very positive, that nurse already has positive winrate. They buffed her twice recently - patch 5.5 and 6.1. I would be very surprised if her killrate stayed under 50%
0 -
I'm not speaking about the Otz's of the world. If you draw a line through a population of data, with half on one side and half on the other this is a median. I think the upper half are experiencing lower kill rates and higher escape rates, and people in the lower half are seeing the opposite. I don't care how many "high" MMR's there are, there will always be an upper half and a lower half.
0 -
- You can't talk about MMR distribution because A. we don't know how this is actually structured and B. you'd have to shave off the back end, as there will be more people at lower MMRs than higher. I'm not sure what you are on about here. Nobody is saying balance only around the top 1% of players, but you can't balance the game around newbies either.
- She's still the lowest. My point is that this is why you can't balance around that data, because it's completely thrown off by new players.
0 -
Not true. Good games balance around both newbs AND top AND avarage player. You just buff/nerf hard (and rewarding) vs easy (but not as strong) perks according to which party inside MMR you want to buff.
Also. We generally do know how the game is structured across MMR. 1, most game have bell-curve distribution. 2, devs already stated that this is so for high MMR. We also know that there are not that many new players so....
0 -
No, they don't.
Name me one game of a similar complexity/nature to DbD that managed to actually do this.
The way this game operates means that the 'team' role, as it requires coordination, will be harder on newbies than the solo role, but you can't just buff the team role because then the solo role gets destroyed against coordinated groups.
Also - no, we don't.
Also also - yeah, I'm going to need a source on that one. 'Not that many new players'...where are you getting this from, exactly?
0 -
In example LoL is much more complex and they are able to do it. All you need to do is identify hard stuff to do vs easy stuff to do. When you want to buff newbs, you buff easy things and nerf hard things. When you want to buff experts you do the opposite
0 -
LoL...just no.
Firstly, LoL is a MOBA. Not an APVP game. Both sides have identical goals, so it's just a matter of tweaking heroes and items.
Secondly, LoL is a full on Esport with something likely in the region of 50x the resources DbD has.
Thirdly, LoL is notoriously unfriendly to new players, which is why it has a full on botmatch mode.
Fourthly, still waiting on your 'there aren't many new players' source. Really curious about this one.
0 -
I don't play as often as I used to but it literally takes me about 4 or 5 days after reset to hit Iri 1 on survivor every reset.
0 -
Solo or in an SWF?
That said, 4 days is dubious. Even on killer, playing quite a lot it takes me about 6 or 7 to hit Iri 1.
Do you stream or something, because clearly you're doing something differently.
EDIT: Yeah. Now that I think about it, even Otz and his mates doing the HCS, which involved completely fresh accounts (and thus being against rather new killers for half the time), in a full SWF made of some of the best survivors around, playing marathon sessions every day - it took them 5 days to hit Iri 1.
0 -
Considering 1 kill is basically baseline and still a loss for the killer (Killers can almost always get at least one kill no matter what), you have to adjust it for the 2-4 number where things matter. Which is much closer to a 50/50 where the killer either loses or draws.
0 -
Kill rate counts only when it's low, when it's high it's insignificant and it doesnt mean anything
-dearly your average forum discussion
0 -
I agree with you about the truth of what you said about the 61% figure not making sense.
It is important to look at the numbers alone and to consider what is behind the numbers and why they are the way they are. It may be difficult to think about it, though.
Nurse has a kill rate below 50%, do you think she should be buffed? I believe she should be nerfed.
It's the same thing.
0 -
I do stream (just finished one actually) but idk. i jsut survive a lot and get lucky with hatch when my team is bad
0 -
Mind putting your handle in this thread? I'm really curious how you are racking up numbers like that, as even the HCS team didn't pull it off in 4 days, and they were playing as a coordinated group, on smurfs, with their #1 priority after surviving being score.
0 -
My twitch name is the same as it is here. And I play a mix of solo and SWF although I primarily prefer to play SWF because randos are ass most of the time.
On the days I play, I tend to go for like 6-8 hours depending on the day so that probably also has something to do with it
0 -
I'll give it a look. '4 days' is what I'm most curious about, as whatever you're doing, it's something that even Otzdarva and his group didn't think of (it took them 5, under fairly optimal conditions). That's a 20% difference.
0 -
It may not have actually been 4 days (I think I initially said 4-5 days in my initial comment anyway), but it only takes me a few days of gameplay to get there.
0 -
Kill rates and win rates are entirely different anyway, since securing 1 kill is not difficult even in losing matches it's expected to have slightly higher than 50% kill rates when win rates are equal.
And yes, emblems are extremely outdated stupid mechanis, it has to be fixed for both.
0
