39% Escape Rate is a Lie
Comments
-
I actually think there are two other points that back up your 39% is inaccurate claim
1: If a player disconnects early the game is tossed for data purposes (to my understanding). But if you play it out it still feels like you lost.
2: While disconnects are tossed, AFK killers probably aren't. Mercies probably aren't. So while there is a system in place to catch and throw out kills because of survivor behavior, I don't think there is one to do the same thing for killer (or put another way: the system tosses kills that happened because of something happening outside what the game intends, it doesn't have a mechanism to catch the same about escapes).\
-
I'm a relatively new player and am escaping maybe 50% of the time. But yeah, there was a period early where it was just non stop massacres.
-
On your chart, shouldn't the bottom 25% of killers get matched against the bottom 25% of players (or what ever the groupings are) - so maybe I'm not getting something but I don't see how that is a problem.
0 -
statistics is just numbers and has no real meaning, I doubt BHVR meant to convince anyone with that.
0 -
If they were balancing for the top, mediocre perks like Ruin and Dead Man's Switch wouldn't have been nerfed. Those are the types of changes for mid and low level play, because instead of playing around those perks, they complain about them til they get nerfed. And yet, those some people still lose to those perks.
I get that solo sucks because of matchmaking, but that's because of the matchmaking. That is no indication that the game isn't balanced for solo or that killer is OP. Fix the MMR system, and the rest will follow.
1 -
I don't think 39% is a lie. I'm a mid level survivor and my most recent run from Ash to Iri showed 45% but that was after also running meme builds fun for about 2/3 of it. That's higher than but within the ballpark of 39%.
0 -
...What? Ideally, all mmr ranges should average 50% imo if the game was well balanced. However, I don't think that will ever occur. Realistically, the best you could probably ever hope for, even then remote, would be 40% to 60% survival rates for different mmrs. It might be survivor sided at high mmr, but that is the issue/point of this thread. BHVR won't post that data because it likely is very ugly. Imagine the uproar if BHVR posted that high MMR survivors had 70% and low mmr survivors got 15%? Killers would be ballistic along with survivors.
If what BHVR posted is accurate, Solo vs SWF really doesn't matter much. Someone posted the BHVR quotes. The escape rate difference between SWF and solo at same mmr is within a couple percents. Now there is a lot that could be going on there to make those statistics lie as well, but not a topic I'm trying to explore here.
I suspect your first few games you were in the kiddie pool still. They usually group brand spanking new players (killers and survivors) before dropping them into the mmr pool. I don't know what the cut off is for this game.
The problem with the chart is an ELO mmr match making system gets long queue times if the mmr distributions don't match up forcing them to loosen match making criteria. If all the killers are lumped together at high mmr ratings, it can be hard to accurately judge weakest from mediocre killers. Most games there is just one pool (ie call of duty there is only 1 player type), but this has killer and survivor pools. So they have to pick from each, you want those curves to at least be similar shapes to match easily between.
Statistics can have plenty of meaning. Modern medicine depends entirely on proving the effectiveness of treatments through statistical measurement for instance. Statistics are used throughout all of engineering. They can also be used to distort and lie. BHVR wouldn't have released this unless they were trying to appease people that everything is "fine".
I haven't dug it up, but I remember BHVR posting showing Ruin was pretty dominate at all tiers of MMR. I don't recall DMS usage rate off top of my head.
Uhm, matchmaking isn't magic. It can't take unbalanced mechanics and make it balanced, unless you are ok whichever side is favored at high mmr having insanely long queue times. BHVR instead throws lower mmr players into the meat grinder to keep those high mmr players with someone to play against. IE: Imagine one side is far weaker than the other, mmr could balance for a while by matching pro players on weak side vs weak players on the strong side, but that is going to cause issues in the long term (see my graphic in the first post).
If you ran up to Iri 1 with meme builds, you are not mid level. You are at least good if not very good. I would expect 50-60% of the player base sees well under 35% escape rate and it is skewed high to 39% by pro players that play tons. I spent 300 games trying to push survivor to Iridescent 1 last season and never made it past gold 4, about 20% swf and 80% soloq. Before the 6.x patches I got Iridecsent 1 for 3 seasons in a row without much trouble.
0 -
The thing is SWF death squads are not common at all. You can turn on any sweaty high MMR killer main streamer to easily verify this. High MMR killers do not play against good players every game. I doubt the extremely small percentage of teams this good are common enough to have any major effect on the escape rate. Like how many SWFs do you think are actually above the softcap? Maybe 1-2%?
0 -
I disagree but, honestly, I'm basing that on my opinion. If you would like to see what my games look like I have a list of my games and builds from Sept 13 to 18th here. If you do have the time to look at it and say why you think our perspectives are different it might give me a better understanding of your argument.
Post edited by TheSubstitute on0 -
Because people don't seem to understand how easy it is to produce a 39% global escape rate while at the same time having terrible escape rates at lower mmr tiers here is some example data of the kind I would love to see from BHVR.
In this case, look how easy it is to get a global 39% rate while having vastly different escape rates for different mmr groups. This does not indicate a healthy balance and 39% hides a very ugly reality.
Now that first example is assuming equal play rates. As no surprise to anyone, pros play a lot more than beginners.
Now in this case I'm assuming the top 10% of players play up to 5 times (2nd top tier playing 4 times as much) as many games as the bottom 25% of players (big assumption on my part although conservative I suspect). I have also towned WAY back the escape rates on the upper end, yet I still end up with just over the 39% global escape rate they are reporting. If we had kept my first assumptions on the top tier escape rates that would require even worse escape rates for the bottom tiers to reach the 39% when weight for game frequency.
My ultimate point is that reporting a 39% escape rate is masking a huge issue in the lower 50% of survivors. Public data on the lower 50% of players is non existent outside our own records. High tier play stats is more accessible and wecan weigh in the smaller population size into the data.
Yes, my data is all hypothesis. I'm making a point that we need more data from BHVR because what they have provided is a extremely misleading and deceptive in attempt to mask a severe balance issue for a large percentage of the player base.
2 -
That's interesting but is still completely hypothetical. If there is an issue with the lower skill levels what would you suggest to fix it?
Personally, I think Reassurance needs a buff as the skill floor difference between camping and countering camping is quite high and survivors need either icons or a basekit Kindred type of effect.
What would you suggest?
0 -
Solutions are going to be complicated and multi pronged, but also outside the scope of what I want to discuss here. I'm trying to draw attention to the need for more data so we can first prove the existence of a problem. Proposing fixes without knowing where the problem resides is a pointless mental exercise. We need to identify the problem, determine the cause, then we can discuss fixes.
That said, a lot of this exploded from the 6.x patch changes, such as gen times. Problems were there before, but not nearly as bad. I think that would be a first place to look for causation, but solution could be something else.
0 -
Duplicate post.
Post edited by Ciabe on0 -
More data does nothing really, we'll see what we want to see in those things anyway.
Just like what we are doing today.
0 -
More datas would help to build & justify our opinions. Right now we're just making assumptions.
0 -
No matter what data is there, people always assume something is hidden and use it as justification.
Definitely will not change anything.
0 -
Nothing's hidden if everything is transparent. On the contrary, what we got right here is blurry enough to create confusion and unfounded discussions.
0 -
Showing literally everything is just impossible, we can't have data like that.
0 -
I disagree. That's what already happens. You get team mates which arent very good so low mmr and it's impossible to carry.
0 -
Why that ? Just add variables to these percentages : filter by MMR / Filter by killer / Filter by Solo or SWF. Boom.
Here you have precise statistics for perk usage, Killrate & escape rate. EZPZ
0 -
And still there is no data of how ACTUAL match goes, it's all results and nothing else.
You can't record every matches and make it a big data.
0 -
Yes thats what statistics are. And thats all I want to see.
0 -
I'm just stating many people can't accept or properly understand "statistics", I won't oppose whether you want to see it or not.
0 -
Yeah that's true ! But I have hope that it might help a little bit to avoid chaotic statements haha
0 -
You can hardly get any useful data with all the cheaters the game has anyway
0 -
Your position does not have much strength behind it no matter how much you repeat it without explanation. Our world is built upon data. Yes, you can have excess data that provides no benefit (looking at you Apple Watch sleep data), but in this case we have almost none of the data. Do we need complete match history publically available? Of course not (although some games provide that). However, I have laid out what data we do need that would prove or disprove the existence of a significant problem. If the data were to match my two example data sets, would you disagree that it shows a problem?
While cheating is a problem, appears to be heavily on the high mmr side and targeted at streamers. Different discussion topic for many other threads. My 2cents is people should stop live streaming and just upload videos after. Then cheaters would never get their air time. However, I know some like the live interaction with viewers.
0 -
Im not high mmr and i see cheaters very often, latest one were able to blind me from the sky, and teleport around the map.
2 days before, i was playing Wraith with the addon so i could see gen progress, there was a survivor that soloed a full gen in 25 about secs, with no toolbox and no perks to boost speed. He did attempt to hide it by joining the match with a toolbox, but i "disarmed him" with franklins.
And there are a lot of cheaters like that guy, where you barely notice it, another example was a 4 man SWF, i managed to down them all but when i picked one of them up, he escaped as if he was using DS, but I had not downed the guy before, and i did not get the DS debuff icon. But he managed to save the rest. Had i been a new player, then i might not even have noticed it. (and he didnt even have the DS Perk)
Most cheaters do it so that you barely notice it, for example a little faster run speed, or like the gen repair speed, i would not have noticed with out that addon in my Wraith game.
0 -
Two types of cheating, the streamer target and the subtle cheater that wants an edge. I obviously never witness the streamer type. The subtle one has been very rare for me, maybe 1 in 50 games I've actually noticed. It is obviously a problem, but there are cheats on both sides so I think we can still generate better data than the blanket "39% escape rate" they have provided so far. Nothing will ever be perfect and no game is ever free of cheaters. If anything, cheaters is usually a sign of a games success. People don't tend to cheat unpopular games unless it is trivially easy.
0 -
Oh yeah, but killers already have certain ways of guaranteeing kills. Like camping with certain addons(or bubba), tunneling against SoloQ(which guarantees an early kill with practically no risk), slugging for a 4k etc.
Those matter right now, because there is no survivor equivalent to it. Give something to survivors that is relatively equal, and killers who camp and tunnel have the risk of giving survivors a free escape, survivors are given the incentive to not go for hooks and remain on a gen to unlock their escape, and killers can counterplay that by keeping them off gens. Essentially, the more killers intervene with survivors, the more likely they are to get more kills, where if killers dont intervene with survivors, the less likely they get more kills.
Currently, this isnt the case. Killers are incentivised to intervene with 1-2 survivors(out of the 4) to try and rush their first kill, and because it can be hard to cooperate as survivors, they can often get a kill with 3 generators remaining against soloq with a second kill relatively close.
Personally, the only thing that killers need, is basekit devour speed buff. Once they hooked a survivor, give them 20 seconds of 10% increased movementspeed to find a new target ASAP, and you can prevent pretty much anyone from escaping early on without finishing all gens.
In the end, if you both play optimally you have an EGC where only 1 survivor could have a guaranteed escape(if they didnt die already), and the killer could camp for a guaranteed kill. Giving both sides an equal option to escape. And killers rushing kills would be met with survivors rushing their personal escape.
0
