Chaos Shuffle stats

We didn't get stats for the previous two Chaos Shuffles, but we did get some stats for 2V8, so there is precedent that you are able and willing to share some stats on the modifiers.
I think getting stats from Chaos Shuffle would be uniquely interesting because as all perks are applied randomly for everyone, over large enough sample sizes insightful trends could show up. Any biases of certain players gravitating toward certain perks or standings certain perks hold in the community are nullified, at least at the selection stage. Of course, stats will still be skewed by various other things, most of all the fact that the different killers are differently lethal, and that the value of perks will also change with the killer being played as and against (therefore you would really want to look at these stats on a per-killer basis). But still, in an overall trend, there should be some interesting or otherwise curious insights to be gleaned from this.
I think the most interesting stat to get would be kill and escape rate by perks. So all perks sorted by their associated kill and escape rates. Or at least the top 10-20 most lethal and bottom 10-20 least lethal killer perks, and top/bottom 10-20 most and least "survivingest" survivor perks. Ideally this would be separated by killers, and further distinguishing between personal and team survival rate per perk, but I know that would be a lot of info to dump.
I just thought I would drop this suggestion here to signal that there's interest, as I'm quite sure a ton of people would likewise love to see these stats. I have actually been taking some of my own stats from myself and streamers, and there's already really rather interesting things showing up in that small sample size.
Comments
-
I would be very interested in this as well.
0 -
Here's some stats I collected from a few sessions of KnightLight's solo Chaos Shuffle escape streak. Keep in mind that this is really a tiny sample size (ranging from 5 to 11 appearances per perk, over a few dozen matches - a lot of perks showed up even fewer than 5 times, which I decided to exclude entirely), so all the things that skew these stats will weigh even heavier. This is really not very representative at all, but I still wanted to share it just because the one or other thing might still be interesting, and I think in general this way of looking at both personal and group survival by perks has merit:
Perks by personal escape rate (so the escape rate of the users of the respective perk themselves):
deliverance - 100%
boil over - 87.5%
overcome - 85.7%
parental guidance - 85.7%
saboteur - 85.7%
unbreakable - 83.3%
plunderer's instinct - 83.3%
streetwise - 83.3% adrenaline - 80%
leader - 80%
we'll make it - 80%
residual manifest - 80%
resurgence - 80%
scene partner - 80%
sole survivor - 80%
moment of glory - 72.7%
resilience - 71.4%
self-care - 66.6%
self-preservation - 66.6%
hyperfocus - 62.5%
rookie spirit - 62.5%
eyes of belmont - 62.5%
lightweight - 62.5%
boon: shadow step - 60%
bardic inspiration - 60%
calm spirit - 60%
bite the bullet - 60%
bond - 60%
alert - 60%
open-handed - 60%
power struggle - 60%
teamwork: collective stealth - 60%
invocation: treacherous crows - 60%
head on - 60%
empathic connection - 60%
wicked - 60%
mettle of man - 57.1%
boon: exponential - 57.1%
reactive healing - 57.1%
invocation: weaving spiders - 57.1%
made for this - 57.1%
dance with me - 57.1%
breakout - 50%
lucky break - 50%
deception - 50%
desperate measures - 50%
hardened - 50%
sprint burst - 50%
slippery meat - 42.8%
no one left behind - 42.8%
aftercare - 40%
dead hard - 40%
dark sense - 40%
diversion - 40%
counterforce - 40%
prove thyself - 40%
hope - 40%
urban evasion - 40%
mirrored illusion - 40%
blood pact - 40%
this is not happening - 37.5%
camaraderie - 37.5%
friendly competition - 33.3%
specialist - 33.3%
chemical trap - 33.3%
wiretap - 30%
strength in shadows - 28.5%
lithe - 20% fogwise - 16.6% reassurance - 16.6%
troubleshooter - 14.2%
windows of opportunity - 14.2%And the same perks sorted by their group escape rate (how many survivors in a group escaped on average when these perks were in play):
leader - 80%
unbreakable - 79.1%
overcome - 78.5%
deliverance - 75%
we'll make it - 75%
residual manifest - 75%
boon: shadow step - 75%
parental guidance - 71.4%
resurgence - 70%
aftercare - 70%
mettle of man - 67.8%
plunderer's instinct - 66.6%
hyperfocus - 65.6%
scene partner - 65%
bardic inspiration - 65%
calm spirit - 65%
boil over - 62.5%
rookie spirit - 62.5%
eyes of belmont - 62.5%
friendly competition - 62.5%
streetwise - 62.5%
moment of glory - 61.3%
resilience - 60.7%
slippery meat - 60.7%
sole survivor - 60%
bite the bullet - 60%
bond - 60%
alert - 60%
dead hard - 60%
dark sense - 60%
diversion - 60%
counterforce - 60%
lithe - 60%
wiretap - 60%
self-care - 58.3%
specialist - 58.3%
adrenaline - 57.5%
wicked - 57.5%
saboteur - 57.1%
open-handed - 55%
power struggle - 55%
prove thyself - 55%
self-preservation - 54.1%
breakout - 54.1%
boon: exponential - 53.5%
troubleshooter - 53.5%
lightweight - 53.1%
fogwise - 50%
lucky break - 50%
deception - 50%
reactive healing - 50%
hope - 50%
urban evasion - 50%
this is not happening - 50%
desperate - 45.8%
mirrored illusion - 45%
teamwork: collective stealth - 45%
invocation: treacherous crows - 45%
head on - 45%
no one left behind - 42.8%
strength in shadows - 42.8%
hardened - 41.6%
sprint burst - 41.6%
reassurance - 41.6%
invocation: weaving spiders - 39.2%
made for this - 39.2%
dance with me - 35.7%
empathic connection - 35%
blood pact - 35%
camaraderie - 34.3%
chemical trap - 33.3%
windows of opportunity - 25%There is always a certain correlation between personal and group survival rates (the likelier it is to escape oneself, the more likely it is for others in the group to also survive, and vice versa - this holds true as a general concept both mathematically and game-logically), but some interesting discrepancies can still be observed, such as Strength In Shadows for example having a considerably lower personal than group survival rate, which may be explained by the circumstance that its user would end up caught and camped in the basement, from where the rest of the group would be reluctant to rescue them, instead just doing gens and escaping. Or more obvious ones like Sole Survivor having a noticeably higher personal than group survival rate.
Although it bears mentioning again that a ton of things will skew these stats, and that's all the more so true with this limited a sample size. First of all, this is taken from the escape streak of a top player, so the escape rate (both personal and group) will skew higher than would be the case on a global average (to note though: "personal" escape rate does not refer to KnightLight only - I looked at all the random players too, so it's their combined "personal" average). Then also just think of something like Treacherous Crows being regarded as so bad that its users regularly may not even think about ever using it, whereas Weaving Spiders may seem more attractive to more people - this could lead to people then wanting to actually activate it, which in turn could backfire in a multitude of ways and explain the lower escape rates for Spiders than Crows: it would simply be because activating Crows (or trying to…) is more detrimental than only having 3 perks and never even thinking about invocating. Or consider that some perks may be "boosted" in their success rates simply because they happened to show up alongside actually good perks that carried them as it were. Or perk synergies. Or that they appeared more often against particular killers (and killer perks) that they are particularly good/bad against than other perks did. And so on.
While many of these things that skew the stats would matter less and less with an increasing sample size, others would always play some role (such as killer played against, which is why I think these should additionally be separated by killers), and the small sample size of my stats basically render them useless. Consider them more so a fun way of "bumping" my thread than anything else. Although if anyone does find anything interesting or worth pointing out or asking about in them, they should of course feel right free to.
I'll also say that if BHVR lets us down again and does not provide any stats on Chaos Shuffle, I might either contact NightLight.gg to inquire about whether they would be open to collecting Modifier data as well, or I'll start a "community stat call" thread next time around where I'll ask people to play any number of Chaos Shuffle solo matches and post match result screenshots for me to collect and run stats on.
Just for fun, here's the top 3 and bottom 3 killer perks as per KnightLight's sessions (obviously an even tinier sample size since there's only one killer player per match, and I also limited the perks to those that had at least 3 instances, so again, doesn't mean much of anything):
Top: Dark Arrogance, Floods Of Rage, Darkness Revealed
Bottom: Shadowborn, Plaything, Agitation
0