Massive discrepancy between BHVR stats and reality? Solo queue seems to be ~30% winrate.
So recently there has been a few topics discussing how miserable solo queue is, with many people claiming their average winrate being around 30% or even less.
This matches my experience but as BHVR didn't give any stats to confirm, I started to look at solo queue streamers on Twitch and record how many times they win to figure out if we can make our own statistics.
RESULTS
I took the top 4 english streamers who are doing solo queue currently:
ANGRYPUG : 2 wins / 9 games - 22% escape rate (vod link)
PROBZ : 4 wins / 16 games - 25% escape rate (vod)
DOWSEY 5.5 wins / 12 games (counting hatch as 0.5) - 45% escape rate (vod)
Otzdarva : 0 win / 2 games - 0% escape rate (vod)
TOTAL: 11.5 WINS / 39 GAMES = 29.4% WINRATE
I will continue updating the results later and redo the same experience exactly next Monday with the same streamers to avoid selection bias.
Now assuming this is true, is it known/intended that the solo queue winrates are so abysmally bad? Wasn't there a statement that 50% killrate is the ideal balance? And isn't MMR supposed to bring players close to 50% regardless of balance?
Now, if even these players who have thousands of hours on DBD and probably amongst the highest MMR, can not achieve a decent winrate, how are casual players supposed to?
Comments
-
Honestly, I wish mmr were tighter so that I would get either slightly better teammates or slightly worse killers. Whatever the case, most of my games are imbalanced. As killer too, but I care less about that because I am actually good at killer.
27 -
Isn't it weird how some people on here claim to have a 70% escape rate in solo q? I wish they'd show us their games.
38 -
Every killer main have an escape rate over 75%
23 -
Most streamers don't play seriously really, and are the worst sample to test this with, as they're either actually trying, or are just chilling with their community.
5 -
26 games divided among four streamers, all of whom are presumably at the highest echelons of MMR is not a useful sample size.
Take a hint from the people on this forum who post the results of their last 100 games and get a volume of entries that actually reflect moving up or down in MMR to balance out the artificially high placements many people were at thanks to Dead Hard and the like.
0 -
Killer streamers have over 90% 4k for sure
19 -
So, let me get this straight.
The measurements using all the matches data must be wrong because some anecdotal data sampled on a specific population (entertainers) says otherwise?
Oh yeah, that makes sense. 😂
12 -
Never trust data from a sample of FOUR people
10 -
It's not anecdotal. What is anecdotal are the people on here saying they have 70% escape rate in solo q and then their post history is just them talking about playing killer.
16 -
I'd be all for both the 30% crowd and the 70% crowd showing their VODS. But to be completely clear - It IS anecdotal.
4 -
Sure, meanwhile most streamers are streaming 10+ win streaks with killer, any explanation for that? Angrypug has been playing survivor for 2 hours and not won a SINGLE game, getting people who DC every 2 games with 1500 people watching (and he DC'd himself after being slugged). PROBZZ has 3000+ hours and an incredible looper, far from a "carried" guy, currently sits at 3 wins / 14 games.
10 -
Kill streamers had an extremely high win rate prepatch. I think Otz was 80% or something for example. That’s understandable because top players don’t exclusively face top opponents, there’s only so many top players total to go around so people like Otz have to face above average and average players a lot or they’d never get to play at all.
2 -
If you had a data set of more than 4 people and a larger match count overall, this might mean something.
But we're looking at less than 40 matches overall between 4 players. Hardly enough to make any sort of conclusion about solo queue.
I can tell you from personal experience, I've had nights where it was a struggle to get a single escape. Then I had nights where I would escape multiple times in a row with a couple deaths. (all after 6.1.0).
EDIT: I should also mention, sometimes the nights were I have trouble getting an escape. The match result is still a 3-man escape because I would die in the end game after the generators are already done.
4 -
I have thought about that as well. Their stats do not match reality at all. No offense to Angrypug but he hardly ever wins and it is not his fault, skill is meaningless for survivors. This man has more survivor hours than the entire playerbase combined and his winrate is embarrassingly low. Angrypug constantly complains about his teammates but it is because so much is expected from them. Casuals simply can’t win.
Meanwhile, casual killers are winning constantly. This is evidence the game is extremely unbalanced in favor of killer.
One of two things is happening at BHVR. Either they are being dishonest with their stats to make the community happy in regards to stats or their stats are spoiled by hackers. There are games where a hacker will light all the gens and prance out the gate and the killer is standing mesmerized. I think that happens a lot at high MMR and BHVR thinks their data is accurate.
14 -
Professional Killer players (that is to say, their full-time job is to play DbD killers) have always had a much, much higher than normal win rate. Look at any of Oh Tofu's videos from the last few years. He has actually been losing MORE now, than he had been historically.
Like I said: come back when you have enough games to actually make a point.
3 -
Not surprised.
2 -
A forumer on here boasted to have escaped 30 Solo queue matches in a row after the patch dropped when Solo queue was made into total Clown world lol.
I think Ayrun's escape streak is something like 34 in a row (pre-patch). Apparently we're schmoozing with God Tier solo survivors..../s
5 -
It isnt though. From the stats they shown before solo players are more than swf.
8 -
This. If survivor is meant to be a team game, I'm fine with that. But teammate matching has to be much closer than what it is now.
0 -
And only a few games from each to add.
0 -
Excluding the games where one of my teammates DC's or gives up on first hook (which is a fair amount still) my escape rate is pretty good, at around 50-60%. Though I am usually the last remaining survivor, playing the "open the exit in 2 stages between killer patrols" game.
Pretty sure if you're legitimately escaping <30% then you're the problem, you're probably not adapting to the changes from the patch. Survivors, especially solo, need to play more cautiously now. No more bumrushing the killer with bodyblocks and clicky flashlights because they will catch up to you before you reach the next loop. But if you adapt and try to stay out of sight where you can, while focusing on gens, then you'll do well. This is how the game is meant to be played, not zerging the killer like a bunch of superheroes taking pot shots at the alien menace.
And yes, you exclude all the games where a survivors DC or gives up, because in those cases you lost because of other survivors sabotaging the game. Direct your anger in the right direction, at the survivors having tantrums.
5 -
Sounds like the escape rates are where they should be.
0 -
I play exclusively SoloQ and my win rate varied, but is more consistently a sacrifice than an escape.
Yesterday I played about 10 matches and escaped two, one match another survivor was a dick and blocked me from exiting and at least 5 of them the killer went about the what is becoming usual method of 'slug player 3, chase player 4 and hook both to get the 4K' routine.
I tend not to let losing get get to me, but it is undeniably grating when you get to that point and end up losing the game to that method. A lot of losses were due to team mates acting like idiots though.
1 -
Solo queue is getting increasingly hard to play for me. I get teamed with either noobs or just bad players and most killers will resort to any tactic to get their 4k. Why should I keep logging on and spend what little free time I have on this train wreck of a game?
4 -
A 26 game sample means virtually nothing. A two game sample is nothing at all.
Some days I'll escape in ~half my games, but have a stretch where I die in 4-5 in a row. If you were to tune in for those, it'd look brutal. I'm sure those guys had win streaks in there too. Sample size is a thing.
Also, defining a "win" is a subjective exercise, but that's another discussion.
3 -
OP did you filter out any games that had survivors DC/kill themselves on hook? Or were there any that even had that in the games you observed
0 -
While I agree collecting data like this incurs a lot of problems, I heavily disagree about sample count being low (btw: now it's 39 games, I added a few new ones).
Despite the sample count being apparently low to our intuition, it's far enough for statistical laws to apply. MMR is supposed to make you win 50% of your games, but if that was true, then the probability of getting only 11.5 wins in a sample of 39 would be 0.4% according to Binomial calculator, so while we can't tell for sure what is the probability of surviving, we can already say that it's not 50%, and sample count won't change anything.
That said, my data is affected by the following problems:
- Low diversity of players (only 4, not representative of what other streamers can maybe experience)
- Time of day, mood of players
Fixing 1) would be tedious but doable, fixing 2) is unlikely to be doable. If you can name a few solo queue players (say 5, ideally with different MMRs) and an approximate time when they play then I'll redo the study with these players. You choose the parameters so I avoid selection bias.
I defined a "win" based on the MMR rules: Escape = win, Death = loss, Hatch = Draw.
1 -
39 games is still virtually nothing.
0 -
No I didn't filter that out, why would I? Didn't count how many times that happened as I didn't watch all the games in full, but it seemed to happen a lot, in particular: PROBZZ had one person disconnect at 5 gens which effectively ruined the game, and ANGRYPUG disconnected himself twice (but the game was in a very bad shape).
0 -
Did you read the paragraph you quoted? Here is a question for you, imagine you roll a dice 39 times, and you get only 6s. Would you need to throw it 1000 more times before claiming the dice is rigged?
0 -
As I've already explained.
If a game has a survivor DC or give up early, then the survivors haven't lost because of any kind of balance issues. They've lost because a survivor sabotaged the game. Anyone paying attention would realise there's a rampant DC issue going on right now.
BHVR don't factor games with DC's into their data.
6 -
You are trying to extrapolate 39 games between 4 players as an overview of the current state of the game for solo queue. 39 games is a drop in the bucket for the comparison you are trying to make. You need significantly more games before you can even make any sort of statement about the escape rate for solo queue survivors.
1 -
Especially when you have Otz there with 2 games. He certainly plays a lot more than that, and he's not an awful survivor. He recently did a survivor challenge which involved no more than about 8 deaths over about 50 games. It's almost not worth including that data, it's heavily cherrypicked.
1 -
Indeed. I often play with friends, but I also play solo, probably about twice as much. We're not a highly coordinated SWF, so my SWF games are about as successful as my solo games, which is about 50% escape rate.
0 -
Both are anecdotal, by definition.
To clarify:
2 -
Because a game that is balanced around a team of 4 working together that has someone quit instantly is not a fair metric value, at least in terms of gleaning 'is the game too killer sided?' There's for sure still value in seeing what the DC/suicide rate is and if it has risen drastically, it surely is indicative of something.
0 -
When you read the aforementioned statistics, you see that about 10% matches are full solos (and that's being generous) and .. I'd have to see it again to get the other numbers. Some people read "50%" and don't think what it actually means for lobbies.
edit: The following post is in a thread where the old statistics were posted. And the same post puts it into perspective.
0 -
You mean hardcore survivor, where he teams up with Ayrun, JRM and Dowsee?
2 -
Looking forward to see more results on this!
0 -
I was going to argue with your claim in the title by saying the changes are really only felt by low and mid tier survivors whereas skilled ones won't see much of a difference in wins.
Although you have a small sample size, it made me hold back a bit since some if not all of those are very skilled.
From my games, I go 8 games in a row as soloQ where most end in a 4k or 3k loss. I think once in an entire evening was I part of a 4 player escape. On the other hand, when I play killer, I am usually winning with 3k or 4k. Occasionally I will get stomped, but killer is a cake walk most of the time and am starting to feel bad for people playing survivor.
I think the big hit is making killer stun times less and also increasing the speed to blood lust up. They should have just picked one of those, not both.
1 -
A survivor is only as good as his team. If the killer is decent, he's not gonna chase probzz for very long. He'll get him when he saves his teammate and he's out in the open. He'll go for a significantly easier target
1 -
and not just the fact that it's 4 people but 2 of those 4 don't even reach a 10 game data set from them.
0 -
A bit silly to call that "less than 10 games", when Angrypug's has 9 games, and includes a 7 games loss streak...
Otz's is only 2 games because I counted the games as they were live, and I didn't know he would stop after only 2 (he switched to killer). It's rare enough to see Otz play solo.
0 -
Less than 10 games is less than 10 games. "It's a bit silly to call something less than 10 games when someone did less than 10 games". That's legit what you just told me.
Also so what.
If AngryPug played a 10th game and escaped, now all of a sudden instead of his escape rate being 22% it's now 30%. That's a big difference.
4 -
They don't know it's him when he plays with a different name / skin than usual. Actually many games he looped the killer well but his potato teammates would just not do gens and he ends up losing anyway. I would say it was a pretty normal solo queue experience. In one of the games he even got camped + tunneled:
Edit: timestamp doesn't seem to work, but it's at 1:38:30 in the video above
1 -
Have they even released recent stats?
1 -
We haven't gotten any stats since before mmr went live.
0 -
Must be it. I watched a lot of that on JRM's Twitch.
4-man SWF, of course they had headsets and comms. They all created brand new accounts, so versed absolute Baby Killers with no experience for a large number of games until their win rate moved them up in MMR where the games got more challenging.
6 -
Well I can't speak for anyone else, but that was pretty much my winrate prior to the update anyway. It's undoubtedly dropped a bit since then but not by a huge amount
0 -
I didn't mean they literally know it's him, I mean that a decent killer can tell right away if a player is a good looper and just bail without too much time investment to find an easier target
0
