I'm noticing a bunch of comments with people having issues opening the HTML file to access their data as it will often crash browsers so try this instead if you're having issues. Open up a new sheet in Google Sheets, then click "File", "Import", and then the "Upload" tab. Drag and drop your HTML file there and it should populate the sheet with your data.
Thanks! I have a couple questions about what to do with the file though. Is it okay to mess with the file in ways that won't affect where the columns are located? Such as removing rows / stuff that doesn't provide any data? And are you asking that we remove data that isn't from BO6, or are you just saying that you're only going to collect that data and we don't need to bother removing the non-BO6 data?
Is there anyway to do this on android? I have my file, but I can't find anything helpful on getting the data on android other than I suppose specifically copying and pasting my match data bit by bit from the HTML. I can't get the URL to even try the import formula.
I just got mine and i sit at 340-380 skill rating and i have only played ranked for 2 months now. Been in crim for 2 weeks now and it is now around 360-390. I hope my data was a little helpfull
Did you notice that shipment and party game modes seem to have their own skill rating? I was going thru my mw3 data and realized all my shipment games had 375+ skill, but regular sized maps would have around 50-100. It was very consistent.. but I haven’t looked into bo6 maps:skill. Curious if this is the case
I hate how so many modern AAA games are made by psych majors who want to manipulate engagement and spending rather than just earning by making a good game.
@@cobaltcardsgaming5982 I concur. I get Activision is a business and at the end of the day all businesses need to make money, but their excessive greed has forced too many poor game designs on these devs and they're actively destroying the sense of community that CoD had in its glory days.
If you're in the United States it depends on the state you're in. like California, Virginia, Colorado, Massachusetts & others have laws requiring company's give this info. i think usually it's within a certain amount of time or they can face penalties.
This is really the only topic I'm still following cod content for at this point. I've moved on from playing but do have some interest in understanding from the hard data where the franchise might have gone wrong and managed to remove all the fun from the game.
Try to expand to other games. Battlefield players had to do this a while ago, and some like myself found temporal refuge in CoD 2019 and Warzone. Now I feel at home playing The Finals, developped by the same people from BF3-BF4 and Mirrors Edge. Even some cosmetics are actually worth it both in price an appearance
Exactly, this is why we had ranked in the first place I stopped playing a month back. It’s just not fun anymore, to top it off the servers and hackers are always a problem they don’t fix
100% agreed. This constant up and down of who I’m being matched with is so lame when you know it’s not organic, just an algorithm. If you’re gonna have this formulaic matchmaking, atleast flat out let me know “ok these guys are statistically way better and this is your punishment for doing too good in the last hour” by showing some kinda rank even in pubs
I requested the data 10 minutes after you released that initial video. I still have not received it. Can’t wait to send it to you. Thank you for everything you do for the community. This is really awesome work!
I remember playing bo3 domination and running into extremely casual people one game and doom members another game, never bothered me cause you can get slammed one lobby and dominate another lobby, that’s what made the game FUN!
you're by far the most empirical content creator in this community. I don't even play bo6 but your videos are so engaging and thorough i cant help but watch
You're a proper treasure, Ace. Observation and analysis without presumption or hyperbole. Someone whose goal is absolutely the acquisition and sharing of knowledge, and not just the view count. You have my utmost respect and admiration. All power to you, sir x
That's a huge issue like you said.. A player going negative and having a higher skill rating but feeling like they're awful at the game. Versus an average dude who goes off vs a bad lobby and has a lower skill rating and feels like they're the best. A bad player logging on usually has an easier time winning a game than an above avg player and that's not a sustainable model.
It doesn't work that way mate.good players know that they are on difficult lobbies or easy lobbies. You can't go negative and be a good player . A good player means that you are positive most of the times and you are top of the leaderboard. I have 2 kd ,I have nukes etc I don't manipulate anything I just get on and play and I play the objective consistently.
@@elg7365 ur literally the reason why they get away with this. You don't even understand what's happening and arguing against an idea with reasoning that just confirms it.
I believe that your implying that having a negative skill rating would indicate a bad in game experience. I would not get too hung up on this idea as it may not be true. Your game experience may great, but your skill level shows as negative. A great example of this outside of video games is in Chess. Magnus Carlsen has stated previously that his chess skill level is declining, but its declining because the other players are not good enough. And in fact its such an issue he has chosen to not defend his World Chess Champion title. Simply put it does not matter if Magnus Carlson wins or losses his skill level will decrease because none of his opponents are skilled high enough. Its possible that something similar is occurring with these negative skill values that you are seeing and the in game experience may not be as bad as it would imply.
This is worse than ranked play. This is outcome manipulation and pretty much fraud. COD is basically a game of chance now and should be disclosed as such to consumers
SBMM, I ignored it, but now i see it! One game i go good with my team cant pull over 4 kills (SnD only btw), next game i got japanese prestige 10 all PC on other team with my 78ping because the game decides to search 190+ping after a good game whenever i search and again and again.... like it wants me to suffer. Crossplay is always off! SICK & TIRED of PC cheaters!
@@RedheadReff I'll bet most of the PC players who discredit controller players by comparing aim assist to aimbot, probably would still lose to a good player with it turned off. Chances are, if they can beat you with it on, they'll still beat you with it off.
I'm actually glad that you would actually go out of your to study this. That is definitely why your 1 of the best call of duty youtubers. Good information that is well researched instead of assuming things and just seeing everything in the negatives like blame the truth or Nero does. I love your content.❤️❤️❤️ and I will be happy when your gonna break down the lmgs in your weapon guides
I’ve been following along these past few weeks, but Seriously, this is so scummy. You could have told me this was a video on ranked play and I wouldn’t have bat an eye. The fact that so much goes in to public matches is just gross. Mainly because it’s a game I used to love, and still do to an extent, being so manipulative with me. It’s like I’m in an abusive relationship and I’m just now seeing things for how they really are.
That's exactly how I've described my feelings towards Activision: like you're in an abusive relationship. On the one hand, we've grown up loving this series; on the other hand, we've become nothing more than pawns to be manipulated for the purposes of increasing Activision's returns to their shareholders. Apparently, in their zeal to go public and make the big bucks, they've forgotten who the games were originally made for. Therefore, that's why I tell everybody: #UninstallCoD #BoycottCallofDuty #BoycottWarzone #BoycottCoD
The reason why they don’t want us to see our skill rating is because it’s not a real number. It varies based on the lobbies you are in and not your overall skill. What I’m interested to see is if you buy skins, does it impact your lobbies? This is a more important test as it will either confirm or deny the theory that spending money equates to easier lobbies.
that is very easy to prove by people buying things, I can tell you I bought my first skin ever and the lobbies were not easier so, no, it doesn't affect your skill rating
Used to work that way back in mw2019, buy something from shop and it was easy ppl to kill. They really wanted me to show off the new stuff in the shop.
Correct. It's not a stable number that will predictably go up or down a bit depending on how "well" you've done past 5 matches or whatever (like your rank in Ranked Play would). It's just relative to your lobby in last game - if you get matched to a skilled lobby avg score +300 and you're at -400... your score will go up at end of match *even if you get zero kills* and get like 200 score total. For BO6 mine varies wildly from +120 to -600, often within a span of a few games (though on average it's around -300).
@@Yuuki_Asuna really depends from my experience. CW - MW22, buying a bundle and playing a match after felt incredibly sweaty compared to the matches before. MW23 on the other hand, did the opposite for me tbh. Didn't really buy too many bundles on MW23, but after purchasing a few (especially the Darryl skin) my lobbies have just been nothing but complete bot lobbies, with maybe 1 or 2 other players actually being able to do decent.
I haven’t played COD for two years and I’m still enthralled watching your videos. You’re doing some great content. I’m hoping COD gets to a game that I want to play again, but my hopes aren’t high for it.
I doubt Activision would show the average lobby skill rating at the start of a map because if they did people would lobby shop looking for the lowest average skill rating they can find. However, it might be interesting if they showed it at the end of the map with a comparison with the player's skill rating. This would allow a player to see how they did against the average e.g. did they do well against a lobby with an average higher than their own or did they perform badly against a lobby with an average skill rating lower than the player's.
That would be great, Id love to know if the games I get absolutely thumped in are because its put me in a god lobby or I wasn't paying attention in and it was actually average.
My scrub ass should be useful as my skill rating varied from -500 to -200. I've also been no lifing BO6 so I'm at Prestige Master, so there's a good sample size, and I do have a sub 1 KD while mostly playing Domination, so score per minute vs k/d will be easy to tell. I'm almost certainly bottom 5%, if not 1%, of your regular viewers, so you know, still coping about that, but happy to help.
I’m in the exact same boat as you, prestige mater, play dom/hard point and k/d is around 1 and my skill rating floats around -200 to -500 so I would what makes playing obj so weak
Same as I only really play HC domination and skill rating is around 10 to - 120. KD is 107 and I do well against sweats and usually top of the board 40 or 50 kills. I was pissed when I saw my rating but figure is different for domination and lower skill ratings are common there. Hopefully we find out more soon. Sending my data this evening cheers op. 👍
I was camo grinding the model L how the hell my skill level go up and I’m feeding the other team kills tryna get these headshots with a weapon I am terrible with but once I switch to a weapon I’m good with my skill level plummets shii don’t make sense. I think they taking in to account the gun you use as well as every thing else and this is why sbmm sucks cuz it makes no sense
Two Ideas (from a mathematician): 1. The SR may not even be the number you’re looking for. This is because it may to be related to the derivative, or even second derivative of your lobby skill. Clearly if a high KD player can have a near zero SR and a low KD player can have a significantly higher SR, maybe that means SR itself is not related to your absolute lobby skill, but rather the direction your lobby skill is moving (or accelerating). Therefore, I would take the integral of each SR and compare it to KD. Then I would integrate it again to see if SR is really the acceleration. This would also have the effect of recreating the “lobby skill” to some extent. 2. This theory, that SR is really the first or second derivative of the actually important skill number, explains two-box SR graph. If SR determines matchmaking, then it actually makes perfect sense that the twobox SR was low. The twoboxing would keep his SR artificially low in order to pad stats. This is sort of counterintuitive because you would expect your SR to be higher against bot lobbies, but it makes sense that it is lower because that is the way the game knows how good you are-so if you are manipulating matchmaking, then it would manipulate SR. What we can infer from this is that SR is not exactly “skill rating” and should not at all be used to compare the quality of players.
Hey Ace, just be aware that one issue you might have with the data coming in is people that play with friends. It could have a similar effect to the 2 boxing players data and skew the lobby data and overall skill rating of the player relative to their true K/D. Also, I suspect your recent W/L is a big component into what your lobby skill rating is going to be. It seems like your are not allowed to win 3 games in a row in modern COD without the 4th game getting totally stomped by high skill rating players.
@ yes I back out all the time to skew SBMM in my favor. If it gives me a crappy match I instantly back out. I play the ones that are fun and competitive but I’m not playing when people are smashing me because the system said it was time for it
loved the content, great graphs representing the statistics. truly shows the volatility of the market. you unequivocally state the intricate yet fundamentally enigmatic ways of 3arc. (youtube algorithm loves this)
It looks like SBMM (for B06 at least) wants you to average 0. When you do well it records that you have a positive skill value and it increases difficulty to force you back to 0, but it ALWAYS overcorrects which sends you negative, and that makes everything easier. I have a wild rollercoaster where I keep peaking at 300, then it crashes me down to -100 to -200 before working back up to 300. I do also need to consider that I've been camo grinding and I don't do it "in order" so I'll randomly work on a launcher, pistol, or melee until I get board then swap to an AR or SMG. But it feels and looks to me like it wants to force people into the 0 range, and it looks like once you get above a certain skill it can't push down (so ace for instance keeps getting pushed to the 200 range). Just my thoughts.
@@CAPSLOCKS0N you cant even spell the word “bored” and you dont know the difference between bored and board. Says everything anyone would ever need to know about you. But u wanna talk About data and algorithms clearly stuff thats way too advanced for you
@@CAPSLOCKS0N the zero range means a 1 kd or for every kill u take a death. That has always been tne goal of sbmm this is not groundbreaking news or a brand new revelation
Asked for the Data one or two days after your video on it and still nothing in sight, but will share this Data with you so it can help you decorticate a bit all this information! Very nice from you as always, thanks!
Finally I got to request my data! I'm so exited to see what the data looks like considering I've been try-harding this game since mw2019. Just took 5 attempts to get the request to go through.
Really insightful analysis! I think there are some fascinating opportunities to dig deeper into the temporal aspects of this data that weren't covered. Since we have timestamp data for each match, we could analyse skill rating volatility across different times of day and days of the week. This could reveal whether SBMM behaves differently during peak vs off-peak hours when player pools are different sizes. Additionally, I'd suggest implementing a time-series decomposition to separate the 'roller coaster' effect into its trend, seasonal, and residual components. This could help quantify how much of the skill rating variation is systematic vs random. Using something like an autocorrelation function (ACF) could also reveal if there are any consistent patterns in how long it takes for skill ratings to cycle between peaks and troughs. For the correlation analysis between KD ratio and skill rating, I'd recommend using Spearman's rank correlation instead of Pearson's, since the relationship appears non-linear at higher skill levels. You might also want to look at conditional probability distributions - i.e., given a player's current skill rating, what's the probability distribution of their next match's rating? This could help characterise the 'rubber-banding' effect more precisely. I've done similar analyses on other games' matchmaking systems, and one thing that often proves insightful is looking at the first derivative (rate of change) of skill ratings rather than just the absolute values. This can reveal whether the system adjusts ratings more aggressively after certain types of matches.
I'm glad K/D isn't the most important because my K/D is only 1.4, but I'm an objective player and I play Ranked. I die a lot just trying to capture points because teammates won't.
how come nobody is talking about win/loss? As someone who cares more about my W/L than my K/D and plays to win each match (unless I’m grinding a specific camo on a crap gun or something which will be relevant in a moment), the thing that ruins my fun the most is when I can tell the sbmm/eomm/whatever is constantly putting me on teams with pylons who completely ignore the objective, won’t even walk through a hardpoint by accident, won’t even throw grenades at the guys capping B or give me any assistance at all when we’re spawntrapped on Nuketown and I’m fighting against the most objective-horny dudes in the world on their team that swarm B like locusts the second I start capping, you get the point. I’ve noticed that if I win too many games in a row, I’ll be punished for it by being put in the most ludicrously one-sided lobbies where I’ll play my heart out and still lose a game of hardpoint like 17-250 while still being the best player on my team by a country mile. And inversely, I’ve noticed that if I bite the bullet and ignore the objective for a few games and focus on trying to get double-kills with the combat knife or whatever it may be, I’ll start getting put lobbies that are more evenly matched. And if I’ve *really* been ignoring the objective for a few games and letting whatever happens happen score-wise, I’ll start getting put in lobbies where someone on my team carries the objective and I’m filling the dead-weight role instead haha. I’ve even had games of kill confirmed for example, where we’ll start off crushing it and be up like 40-15 and then suddenly two players on their team and a player or two on my team will disconnect and be replaced by new players, only for the score to shortly flip in their favor and SO OFTEN they’ll end up winning by 1 which drives me bonkers. But TLDR/point being, it’s very obvious to me that my games are being manipulated behind the scenes to try and account for my objective play. It’s very much not just tied to score per minute and K/D, whether you win or lose and how much you try for the objective plays a big factor too and I hope you’ll include a section on that in your breakdown video because that’s the angle I’m most interested in by far.
I play mostly hardpoint and I live on the point and am usually at the top of the scoreboard with a W/L ratio >2 and pretty much a 1.0 K/D ratio (+ E/D ratio) but my skill rating per this file is negative. I don't see anywhere on the file where it takes into account time on objective. I agree that the matchmaking must also largely be taking into account W/L cause when I win a few in a row the game tries to humble me by putting me with a team of pacifists who cannot shoot the enemy and actively run away from the objective.
One thing that annoys me a lot is that in game it shows elimination death ratio. But assists count as elims - so I never know if I’ve actually got a positive KD during a match ..
Got my data after a couple weeks wait. it's been interesting looking through it. I'm not at all surprised they collect it and use it to match make. It's no secret COD lobbies are 'skill' based, it's what makes it addictive to play. As a low level casual player I have no interest in getting randomly added to lobbies with die hard players, your team might still win rounds but if there's one or two players dominating, it'd just be frustrating. My skill rating for BO6 has been consistently around minus 320 (low -450, high -200), KDR is on average 1.07, score per min is 422. I looked into correlations of 'change in skill rating' with things like KDR and SPM. On the full time base of the data set there's not much correlation, but within a single session you can see them move together as you warm up and then get tired.
@shari0nplaying slow or methodically is generally better for a good K/D than playing fast and being constantly rushing around to new places (even though streamers will have you think otherwise) So I’m not sure that will be the case
Hey Ace, Just got my skill data, gonna send it to you. One thing that stood out to me is my skill rating in black ops six is quite low as of lately 50-100 & often times in the negative wih only a small patch of games in the 200s. Yet other games like MWII and MWIII my skill rating is in consistently higher in the 300s & 400s? This is odd to me because all of the games i play are literally sweat fests. I also played the same way across all those games & the same modes. (90% S&D) All I do is snipe at a very high level 95% of the time. I Often get matched against FaZe members, YTers TikTokers, TTVs etc & have proof of that in my YT vids & tiktoks. Another thing to note is Ive never reverse boosted or 2box on any of these games. My KD is hovering around the same between all three games (1.1 - 1.4) sniping only mind you. I think a huge missing part is definitely the skill rating of the lobby and players we are going against this has to tell another story on how our skill rating is calculated because something isn’t adding up to me between games. My skill rating particularly in bo6 has such large fluctuations in it yet almost all my lobbies feel the same.
Yep, I’ve made an alt account this week. Played pubs until I got 50 wins, absolutely smashed it. Went into my first game of ranked and had to play against pretty good G3 /P3 players..
@@TheXclusiveAce The funny thing is, they base your ranked SR goal, and therefore your supposed rank division, on your hidden MMR. So, I’d say the hidden MMR is more valuable to know than the rank division you’re in. It would also be interesting to study the correlation between your MMR and your SR goal
joining a lobby of a new/intentionally low skill account in order to get easy lobbies. Back in MW3 you could basically feed the opponents for 30-40 matches and then get 10 people that can't shoot, and 1 player that was also doing the same. As this is mostly unsuccessful for veteran players, they started creating fresh account, joining a match of that account while it's in progress and killing every 8 year old Timmy they could aim at.
It occurs to be that a big part of the problem with the discussion around matchmaking is that SBMM/EOMM do not accurately describe what's going on. I prescribe that the most accurate name would be Performance Based Matchmaking.
if activision allowed players to see the overall skill rating in public lobbies, then I imagine people would just start backing out if they see a level that they consider too high for them. And activision will never allow that.
@@spartanx169x Nobody cares about how they do or how much their challenged its casual people are there to fuck around with friends and grind camos. This a system that belongs in ranked/competitive games.
@@MC_605 You may be right, you may be completely wrong. Unless you have a source that says otherwise. Other games(not COD) had this information and its always beneficial to the player to be able to see it.
Can you include the users’ RANKED rating as well? That will help us understand how the skill ratings correlate with actual player skill and lobby skill
Ace, I got my data today and looking through it, I think the skill rating might be the *lobby* skill rating. There does not seem to be a correlation between round performance and skill rating. If skill was an individual rating, you should see a rough correlation between performance and skill, but it doesn’t seem to impact skill rating much at all.
I appreciate what you're doing, Ace, trying to shed some light on Activision's opaque SBMM algorithm. I want to help you out and share my stats, which go all the way back to MW2019 alpha (right?) and beta tests and every release since then, up to and including BO6 and every version of Warzone. (Until I uninstalled in frustration for the first time ever, about two weeks ago now.) Unfortunately, I'm getting the bullshit "We've encountered an error. Please try again." message when I click on [ Submit ] after logging in and seeing that they do indeed have all my linked platforms, etc. I'm sure they noticed the uptrend in people requesting their data and decided an "Error" message was the simplest way to put a stop to that. Fuck this company, fuck their games (both literal and figurative games). I'm done with them until I hear they've made significant changes to how they do business. Knowing Activision, that simply means I'm done with them; they won't change their business practices, no matter how vocal their veteran community is. #UninstallCoD #BoycottCallofDuty #BoycottWarzone #BoycottCoD
Low skill account searches for a match. Low skill account finds match. Your actual account instantly joins the half filled lobby your low skill account just found. Voila. Everyone in your lobby has now the same skill as your low skill account, instead of the skill your actual account, which joined the lobby later, has.
If I was implementing this system, I would have 1 KD as 0 skill on that graph. So < 1 KD is negative, and > 1 is positive. Of course this is just a rough, simplified way to describe the system. We would have to know how they are valuing certain things in game. But I would say that 1 KD is a starting point.
Not a bad starting point, but this would only work for kill-based modes and not objective. In MW22 (the last game I played that had actual K/D’s), my K/D was only .85ish, but I only ever played objective modes and farmed points for score streaks, so my W/L was like 1.4 and my team would routinely steamroll pubs.
@ yes but you wouldn’t see a skill increase if you played against better players over time because your kd would average out to a 1 over time. So your skill would be the same if you were comparing the stats of pro players against each other or Christmas noobs against each other.
I don’t play COD anymore but I do find this really interesting… my skill PLUMMETED after MW2019. To a degree that made me *so* confused as to what happened between games…My experience between 2019, and II, and III was night and day. The only thing I could think of was either some tweak to the damage values or the SBMM system that I wasn’t privy to… and the data is leaning towards the latter.
This is why I have been loving the 4 player Co-Op games. With no multi player. After work I just want to chill and game. Not be playing like if I’m on rank play
I don't know why you have 3 but I think it's fine, if one is larger than the other that's when you might need to be worried because it could be testicular cancer, but could also just be a hydroceal (can't remember the spelling) that's a build up of water in the testicle which isn't as serious but still should get it looked at If you're from America I wouldn't worry but if you're from a place with actual good Healthcare I'd recommend just to be safe getting checked out by a Doctor especially to see why you have 3 testicles
Great job dude. Im very interested to see how this turns out. I wish they'd just lighten up a bit on sbmm. I dont want EVERY game to be my application video for the CDL.😂😂 just chill mostly.
I’m a data scientist in my day job and ran some Python data analytics on my skill data (sent it to XclusiveAce and Nero). I was able to back into the SBMM algorithm with an XGBoost regression model and find the biggest correlations to skill, and it is mainly your skill in your last 8 matches, last match score, lifetime wins, and time played. I started BO6 with a fresh account and my skill was -320 for a long time. Then they started to set it to random values and the algorithm kicked in, shooting up to 75, before hovering around 0. What threw me was the #1 skill indicator was lifetime wallbangs! Perhaps to catch people walling, so it jacks up your skill to the max. Either that or they assume wallbangs = skill… 🤓
Feature Importance Lifetime Wall Bangs 0.33780137 Skill_1 0.23941274 Skill_2 0.114405714 Skill_3 0.037927497 Lifetime Games Played 0.032032818 Skill_8 0.02712443 Skill_6 0.025060102 Score_2 0.018084485 Lifetime Wins 0.013437924 KD_Ratio_1 0.006047078 Skill_10 0.005968014 Score XP_5 0.005213942 Shots_6 0.004903351 Skill_4 0.004834562 Kills_1 0.00427607 Rank at End_4 0.00380922 Rank at End_3 0.00364872 XP at End_6 0.002988484 XP at End_7 0.002617068 XP at End_5 0.002531297 Score XP_1 0.002427384 Score_10 0.002400091 Hits_2 0.002327372 Weapon XP_5 0.002243329 My regression model feature importances (weightings) that go into skill. Showing the ones that account for 90% of my skill variance.
Hey Ace, if I could make a request, when you do the video can you take a look how K/D ratio correlates with W/L ratio? It seemed when I first started that everyone who was above a 1.0 K/D had a negative W/L, and a lot of people with a negative K/D had a positive W/L. Would be interesting to see if that is true when you get the data. Thanks for taking the time to do work on all this skill rating stuff and keeping the community informed.
I could pretty much guarantee that K/D ratio and W/L ratio have an inverse correlation amongst the community. And it’s not because Activision/COD is conspiring against “good” players with high K/D. It’s because people that value K/D over all else pretty much screw over their objective playing teammates every match by not making any effort whatsoever to try and contribute to winning the match and thus the low W/L.
I think the problem with BO6 is that when you do well, it puts you in a lobby where your entire team is much weaker than the entire enemy team. MW2019, BO:CW and MW2022 they'd put you in a lobby of strong players on both teams and a weaker one here or there on each team.
That is exactly what I'm experiencing in BO6. None of my team will be able to break a 1.0 KD and the entire other team is like 2.0 or better and it happens way too often. I'm inclined to believe that they are not making any effort to balance the teams.
@@Calma123_ Hard disagree, the mechanics are great, but many of the maps are too small for the fast movement speed, even more so than MW3 2023's remade maps. MW2019 remains the best CoD to this date.
@@spartanx169x yup, that's what sucks. Because it's alright if 1-2 plays on the team are weaker, but if the entire team is weaker it compounds/snowballs and makes it even worse.
I wish these stats could explain how solo queueing in Domination is the worst experience in COD history. Either puts me in an existing game or gives me 5 teammates who lack opposable thumbs.
521 SPM and 125 OBJ SPM and constantly queued with guys who have no idea what's going on. If am not going as hard as possible, my team gets steamrolled. Thanks, COD
I’m a crimson player (prob could be iridescent but cheaters and lack of time). Have a 2.25 pub kd. Average skill is 415.44. Highest skill game was 622. Sent the data over.
And in the data I received. Dozens of items are saved that prove that the game is rigged at all levels: Operator skin, operator bundle, the execution chosen for the operator, AND THEY TELL US THAT PURCHASES IN THE STORE DO NOT COUNT FOR THE EOMM 🤔🤔, the percentage of time moving per game, the time without dying, shots through walls... etc, etc, I UNDERSTAND BETTER AND MORE AND MORE WHY I HAVE THE IMPRESSION THAT IN EVERY GAME I AM A BOXER WHO TRAINS IN FRONT OF A MIRROR. THE GAME IS REALLY COMPLETELY RIGGED.
Right? Like why keep track of which operator you used, unless its used to adjust the games damage. All these people who bought stuff from the store are getting a pay to win game.
Ace one thing I just noticed that I’m sure you will notice is that something is up with the Hardcore game modes when it comes to skill rating. It seems to me like the ratings are inflated for hardcore modes
Ran-snack definitely had no SBMM, I haven't played as much Cod in years as I did when that playlist was out, it genuinely felt like old cod and my EKIA/D for the prestiges I was doing it was up to a full point higher than in the other lobbies.
Okay Ace. Highest skill rating I've had was 300. Its fluctuated between that down to the negatives and back up later on. This is all Warzone data I'm giving you from Black Ops 6. 718 games played. 3360 kills 2007 deaths. 1.76 KD. My elimination death ratio is 2.60. 5428 eliminations 2091 deaths compiled in a total of 782 games. So for 64 games, I'm missing match data that would truly show what my warzone kd is.
Freedom is a great thing. Give the players the choice. Those that want to play general lobbies/modes with regular SBMM and one where it's extremely loose or non-existent. Let the players decide so we can end people whining about being stuck in one or the other. Problem solved right?
In the last CoD, I always played the s2 map pack, s3 map pack, s4 map pack. I found there was often seemingly minimal sbmm and I could regularly dominate with 3kd and 4kd. When I went into the regular general playlist, I didn't really enjoy the game at all.
7:48 lmao... I've said it before as a member of communities in other games, just give us our ELOs and of the lobbies we play against. Just like Ace says, I think people's gripes with SBMM would go way down if they knew where they stood in everything. "But that's what ranked is for." Sort of... I think it's very different to have a "ranked ELO" on top of a regular "skill rating." Going into ranked is like an agreement that everyone is putting their ELO on the line as the objective of the mode. Getting matched based on skill in a more casual mode makes the ELO only a byproduct of playing however players want to play. Finally, the speculation that limited-time modes have not affected "skill ratings" in the past and that perhaps SBMM is not used as much or at all shows, in my opinion, that games would be better off just being clear about which modes have SBMM and which don't. These changes would clear up any illusions people have about how good they are or why they're struggling, etc. As a noob, you would join a non-SBMM mode at your peril, but you'd know that going in. As a veteran player, you wouldn't necessarily take so much pride in stomping a lobby with a weak skill rating, in part because everyone you would be playing against would know the implication, as well (perhaps that it would actually be more impressive if you went and did that against some tougher competition, whether in ranked or just another, more casual SBMM mode). Anyway, this is great info! Thx! A not insignificant chance that these revelations actually change competitive online shooters for the better in the future.
It looks like sbmm is weighted the same as ranked, in ranked if your gold 1 and you play against gold 1s and hit top of the leader board and you go positive then let’s say you get 100 SR then if you play a lobby of gold 3s and hit top leaderboard and go positive you’ll get 135 SR , I’ve done some very limited testing of this and it seems to work that way , so in pubs your skill level against the same skill level brackets then you’ll barely go up but if you go wayyyy up in skill level brackets then your skill ranking will go up more after a good performance
So, I'm assuming in ranked MP, they are mostly putting you in games at your own skill level and not in a random lobby of your rank. So if you are a decent Plat player, all the golds, plats and diamonds in your lobby would basically be the same. Which is SBMM essentially.
I'm noticing a bunch of comments with people having issues opening the HTML file to access their data as it will often crash browsers so try this instead if you're having issues. Open up a new sheet in Google Sheets, then click "File", "Import", and then the "Upload" tab. Drag and drop your HTML file there and it should populate the sheet with your data.
Thanks! I have a couple questions about what to do with the file though. Is it okay to mess with the file in ways that won't affect where the columns are located? Such as removing rows / stuff that doesn't provide any data? And are you asking that we remove data that isn't from BO6, or are you just saying that you're only going to collect that data and we don't need to bother removing the non-BO6 data?
can we get those macros? lol
Is there anyway to do this on android?
I have my file, but I can't find anything helpful on getting the data on android other than I suppose specifically copying and pasting my match data bit by bit from the HTML. I can't get the URL to even try the import formula.
I just got mine and i sit at 340-380 skill rating and i have only played ranked for 2 months now. Been in crim for 2 weeks now and it is now around 360-390.
I hope my data was a little helpfull
Did you notice that shipment and party game modes seem to have their own skill rating? I was going thru my mw3 data and realized all my shipment games had 375+ skill, but regular sized maps would have around 50-100. It was very consistent.. but I haven’t looked into bo6 maps:skill. Curious if this is the case
Remember when you could just play a game without having to think about all of this bullshit?
You think about this while playing?
I hate how so many modern AAA games are made by psych majors who want to manipulate engagement and spending rather than just earning by making a good game.
@@codyo7384 I don’t play it.
@@cobaltcardsgaming5982 I concur. I get Activision is a business and at the end of the day all businesses need to make money, but their excessive greed has forced too many poor game designs on these devs and they're actively destroying the sense of community that CoD had in its glory days.
@@cobaltcardsgaming5982 its made by greedy corporate executives who hire psych majors who need a job.
Thank you EU for having actual laws for this kinda stuff
If you're in the United States it depends on the state you're in. like California, Virginia, Colorado, Massachusetts & others have laws requiring company's give this info. i think usually it's within a certain amount of time or they can face penalties.
Common EU W, this is coming from an American
@@iddqdgodmode if you’re from another state do you think they still need to give it?
@@iddqdgodmode i asked about 1 week ago and havent received anything. WTH activision being weird.
Im eu is there a law to get out of a shadowbann?
This is really the only topic I'm still following cod content for at this point. I've moved on from playing but do have some interest in understanding from the hard data where the franchise might have gone wrong and managed to remove all the fun from the game.
I stopped about 4 years ago and this is the only content I keep up with as well. I moved on to CounterStrike
Same, stopped partway through mw22. I rarely play fps games period now.
Mw19 and its strict sbmm. If that game wasn’t during the pandemic and didn’t look so good it would have been hated.
@@i_amChrisJxmesThat’s where I quit playing. Garbage game
@@DanKamYouKnowdidn't expect to see you here!!
Great to see the COD community developing their Microsoft Excel skills
Libreoffice, its free, screw microsoft.
Skill based excel making wil be the next topic 😏
As a career long intermediate-level excel user (at best), can attest to the benefits of this skill. 😆
@@craciunator99 😂👌💯
@@craciunator99he said while playing a game owned by Microsoft 🤣
Man, I really wish there was a competitor to call of duty
Try to expand to other games.
Battlefield players had to do this a while ago, and some like myself found temporal refuge in CoD 2019 and Warzone.
Now I feel at home playing The Finals, developped by the same people from BF3-BF4 and Mirrors Edge. Even some cosmetics are actually worth it both in price an appearance
RIP XDefiant
@reg1990
Xdef was not a cod competitor.
@@kunimitsune177xdefiant was created with the intention of being a cod competitor
marvel
It's crazy that pubs are now essentially ranked without being able to see your actual rank. And they wonder why COD is bleeding player numbers.
Sincerely agree. I don't mind a ranked system, but not knowing the level of your opponents is just aggravating. Just give us a rank.
Exactly, this is why we had ranked in the first place I stopped playing a month back. It’s just not fun anymore, to top it off the servers and hackers are always a problem they don’t fix
100% agreed. This constant up and down of who I’m being matched with is so lame when you know it’s not organic, just an algorithm.
If you’re gonna have this formulaic matchmaking, atleast flat out let me know “ok these guys are statistically way better and this is your punishment for doing too good in the last hour” by showing some kinda rank even in pubs
I requested the data 10 minutes after you released that initial video. I still have not received it. Can’t wait to send it to you. Thank you for everything you do for the community. This is really awesome work!
Same
Where do u go to rewuest it?
Same
Where did you go to request it?
Go to his earlier video on how to.
I remember playing bo3 domination and running into extremely casual people one game and doom members another game, never bothered me cause you can get slammed one lobby and dominate another lobby, that’s what made the game FUN!
agreed. that variety kept every lobby fun and kept me addicted. there has been nothing like bo3
You could also use when you were getting better
@@Shotgun-qv6xn bo3 happened 10
Years ago. The landscape of gaming has completely changed. Make sure u mention that part
@@Shotgun-qv6xn thats all you people know how to do is complain, moan, victimize yourselves and look back into the past with rose tinted glasses
@@Shotgun-qv6xn every cod felt like that back then. It wasnt just bo3. And it was because cod was at an earlier stage in its life
In Soviet Activision, if the product is not free, you still are the product
Thats just capitalism as a whole
In Soviet Activision the game plays you.
@@tylerhill40 😂😂😂
The fact that it took this long for someone to find out how to get access to our data is kinda crazy
^^^ THIS
you're by far the most empirical content creator in this community. I don't even play bo6 but your videos are so engaging and thorough i cant help but watch
You're a proper treasure, Ace. Observation and analysis without presumption or hyperbole. Someone whose goal is absolutely the acquisition and sharing of knowledge, and not just the view count. You have my utmost respect and admiration. All power to you, sir x
About the gayest comment I’ve seen zip him up when your done bud
@@TeaRApolloGaming /slurp
That's a huge issue like you said.. A player going negative and having a higher skill rating but feeling like they're awful at the game. Versus an average dude who goes off vs a bad lobby and has a lower skill rating and feels like they're the best. A bad player logging on usually has an easier time winning a game than an above avg player and that's not a sustainable model.
It doesn't work that way mate.good players know that they are on difficult lobbies or easy lobbies. You can't go negative and be a good player . A good player means that you are positive most of the times and you are top of the leaderboard.
I have 2 kd ,I have nukes etc I don't manipulate anything I just get on and play and I play the objective consistently.
@@elg7365 ur literally the reason why they get away with this. You don't even understand what's happening and arguing against an idea with reasoning that just confirms it.
I believe that your implying that having a negative skill rating would indicate a bad in game experience. I would not get too hung up on this idea as it may not be true. Your game experience may great, but your skill level shows as negative. A great example of this outside of video games is in Chess. Magnus Carlsen has stated previously that his chess skill level is declining, but its declining because the other players are not good enough. And in fact its such an issue he has chosen to not defend his World Chess Champion title. Simply put it does not matter if Magnus Carlson wins or losses his skill level will decrease because none of his opponents are skilled high enough. Its possible that something similar is occurring with these negative skill values that you are seeing and the in game experience may not be as bad as it would imply.
@@elg7365 I have a 1.8 KD and a 3 win/loss..
Winning means nothing anyways since there's no rewards for winning
What’s the point of ranked play modes if we are being ranked in casual play?
This is worse than ranked play. This is outcome manipulation and pretty much fraud. COD is basically a game of chance now and should be disclosed as such to consumers
They can never disclose the “skill” rating because “skill” is a misnomer for what they’re really doing.
SBMM, I ignored it, but now i see it! One game i go good with my team cant pull over 4 kills (SnD only btw), next game i got japanese prestige 10 all PC on other team with my 78ping because the game decides to search 190+ping after a good game whenever i search and again and again.... like it wants me to suffer.
Crossplay is always off! SICK & TIRED of PC cheaters!
@@Beautyofanime1 just keep playing it. It'll get better 🤣
@@Beautyofanime1pc cheaters? Every console has built in aim-hacks gtfo
@@RedheadReff I'll bet most of the PC players who discredit controller players by comparing aim assist to aimbot, probably would still lose to a good player with it turned off. Chances are, if they can beat you with it on, they'll still beat you with it off.
@@mattwells1036nah, it's been done and controllers are just shit without the AA.
I'm actually glad that you would actually go out of your to study this. That is definitely why your 1 of the best call of duty youtubers. Good information that is well researched instead of assuming things and just seeing everything in the negatives like blame the truth or Nero does. I love your content.❤️❤️❤️ and I will be happy when your gonna break down the lmgs in your weapon guides
I’ve been following along these past few weeks, but Seriously, this is so scummy. You could have told me this was a video on ranked play and I wouldn’t have bat an eye. The fact that so much goes in to public matches is just gross. Mainly because it’s a game I used to love, and still do to an extent, being so manipulative with me. It’s like I’m in an abusive relationship and I’m just now seeing things for how they really are.
That's exactly how I've described my feelings towards Activision: like you're in an abusive relationship. On the one hand, we've grown up loving this series; on the other hand, we've become nothing more than pawns to be manipulated for the purposes of increasing Activision's returns to their shareholders. Apparently, in their zeal to go public and make the big bucks, they've forgotten who the games were originally made for. Therefore, that's why I tell everybody:
#UninstallCoD
#BoycottCallofDuty
#BoycottWarzone
#BoycottCoD
The reason why they don’t want us to see our skill rating is because it’s not a real number. It varies based on the lobbies you are in and not your overall skill. What I’m interested to see is if you buy skins, does it impact your lobbies? This is a more important test as it will either confirm or deny the theory that spending money equates to easier lobbies.
that is very easy to prove by people buying things, I can tell you I bought my first skin ever and the lobbies were not easier so, no, it doesn't affect your skill rating
@@Yuuki_Asunalmaooo “I bought something and it wasn’t easier, therefore it never is.” This is called anecdotal evidence. Not fact.
Used to work that way back in mw2019, buy something from shop and it was easy ppl to kill. They really wanted me to show off the new stuff in the shop.
Correct. It's not a stable number that will predictably go up or down a bit depending on how "well" you've done past 5 matches or whatever (like your rank in Ranked Play would). It's just relative to your lobby in last game - if you get matched to a skilled lobby avg score +300 and you're at -400... your score will go up at end of match *even if you get zero kills* and get like 200 score total.
For BO6 mine varies wildly from +120 to -600, often within a span of a few games (though on average it's around -300).
@@Yuuki_Asuna really depends from my experience. CW - MW22, buying a bundle and playing a match after felt incredibly sweaty compared to the matches before. MW23 on the other hand, did the opposite for me tbh. Didn't really buy too many bundles on MW23, but after purchasing a few (especially the Darryl skin) my lobbies have just been nothing but complete bot lobbies, with maybe 1 or 2 other players actually being able to do decent.
I haven’t played COD for two years and I’m still enthralled watching your videos. You’re doing some great content. I’m hoping COD gets to a game that I want to play again, but my hopes aren’t high for it.
I doubt Activision would show the average lobby skill rating at the start of a map because if they did people would lobby shop looking for the lowest average skill rating they can find. However, it might be interesting if they showed it at the end of the map with a comparison with the player's skill rating. This would allow a player to see how they did against the average e.g. did they do well against a lobby with an average higher than their own or did they perform badly against a lobby with an average skill rating lower than the player's.
I lobby surfed last night. Found a lobby where i dropped 100 kills. I dont blame people who do that. Results in more fun.
Reverse boosting already has that effect it’s just more powerful but less convenient
@@ChrisDaAce27 More fun for whom?
That would be great, Id love to know if the games I get absolutely thumped in are because its put me in a god lobby or I wasn't paying attention in and it was actually average.
@@georgereston me obviously
My scrub ass should be useful as my skill rating varied from -500 to -200. I've also been no lifing BO6 so I'm at Prestige Master, so there's a good sample size, and I do have a sub 1 KD while mostly playing Domination, so score per minute vs k/d will be easy to tell. I'm almost certainly bottom 5%, if not 1%, of your regular viewers, so you know, still coping about that, but happy to help.
I’m in the exact same boat as you, prestige mater, play dom/hard point and k/d is around 1 and my skill rating floats around -200 to -500 so I would what makes playing obj so weak
I love to play chess but can't for the life of me get above average rating. So I understand what you mean. COD on the other hand - I dominate. :D
Same as I only really play HC domination and skill rating is around 10 to - 120. KD is 107 and I do well against sweats and usually top of the board 40 or 50 kills. I was pissed when I saw my rating but figure is different for domination and lower skill ratings are common there. Hopefully we find out more soon. Sending my data this evening cheers op. 👍
I just did my averages for bo6 and mw 3 and my average is -356 and -244 and yeah I don’t feel like I’m that bad of a player.
I was camo grinding the model L how the hell my skill level go up and I’m feeding the other team kills tryna get these headshots with a weapon I am terrible with but once I switch to a weapon I’m good with my skill level plummets shii don’t make sense. I think they taking in to account the gun you use as well as every thing else and this is why sbmm sucks cuz it makes no sense
I'm getting "We've encountered an error. Please try again." when I tried requesting my personal info from Activision.
You need to log in with linked accounts like psn/steam/battlenet along with Activision
@@rami9225 Thanks
That doesn't work either
@SteveV1231 Did you log in to all of the linked profiles? Fixed it for me
I tried a different browser and it worked
Two Ideas (from a mathematician):
1. The SR may not even be the number you’re looking for. This is because it may to be related to the derivative, or even second derivative of your lobby skill. Clearly if a high KD player can have a near zero SR and a low KD player can have a significantly higher SR, maybe that means SR itself is not related to your absolute lobby skill, but rather the direction your lobby skill is moving (or accelerating). Therefore, I would take the integral of each SR and compare it to KD. Then I would integrate it again to see if SR is really the acceleration. This would also have the effect of recreating the “lobby skill” to some extent.
2. This theory, that SR is really the first or second derivative of the actually important skill number, explains two-box SR graph. If SR determines matchmaking, then it actually makes perfect sense that the twobox SR was low. The twoboxing would keep his SR artificially low in order to pad stats. This is sort of counterintuitive because you would expect your SR to be higher against bot lobbies, but it makes sense that it is lower because that is the way the game knows how good you are-so if you are manipulating matchmaking, then it would manipulate SR. What we can infer from this is that SR is not exactly “skill rating” and should not at all be used to compare the quality of players.
Making general observations and initial assessments of potential trends IS scientific. This is the fuck around stage of the scientific method.
The best video I've seen so far where the SBMM is really studied and dissected. Thank you for taking us all with you. best man 💪🏻❤
Hey Ace, just be aware that one issue you might have with the data coming in is people that play with friends. It could have a similar effect to the 2 boxing players data and skew the lobby data and overall skill rating of the player relative to their true K/D. Also, I suspect your recent W/L is a big component into what your lobby skill rating is going to be. It seems like your are not allowed to win 3 games in a row in modern COD without the 4th game getting totally stomped by high skill rating players.
You get 3 game streaks? LoL
Exactly why I no longer care about W/L or K/D. My K/D is 1.8 but my W/L is 0.2. I'm over it and won't be touching COD again.
@ yes I back out all the time to skew SBMM in my favor. If it gives me a crappy match I instantly back out. I play the ones that are fun and competitive but I’m not playing when people are smashing me because the system said it was time for it
@@emwolsi leave at 198 on dom and 70 in tdm
@@davidc1961utube I do, but when I do I will get put on to teams where nobody can get above a 1.0 KD for at least another 3 to 5 matches, if not more.
loved the content, great graphs representing the statistics. truly shows the volatility of the market. you unequivocally state the intricate yet fundamentally enigmatic ways of 3arc. (youtube algorithm loves this)
It looks like SBMM (for B06 at least) wants you to average 0.
When you do well it records that you have a positive skill value and it increases difficulty to force you back to 0, but it ALWAYS overcorrects which sends you negative, and that makes everything easier. I have a wild rollercoaster where I keep peaking at 300, then it crashes me down to -100 to -200 before working back up to 300.
I do also need to consider that I've been camo grinding and I don't do it "in order" so I'll randomly work on a launcher, pistol, or melee until I get board then swap to an AR or SMG.
But it feels and looks to me like it wants to force people into the 0 range, and it looks like once you get above a certain skill it can't push down (so ace for instance keeps getting pushed to the 200 range).
Just my thoughts.
@@CAPSLOCKS0N you cant even spell the word “bored” and you dont know the difference between bored and board. Says everything anyone would ever need to know about you. But u wanna talk
About data and algorithms clearly stuff thats way too advanced for you
@@CAPSLOCKS0N the zero range means a 1 kd or for every kill u take a death. That has always been tne goal of sbmm this is not groundbreaking news or a brand new revelation
@@CAPSLOCKS0N go cry about it
Asked for the Data one or two days after your video on it and still nothing in sight, but will share this Data with you so it can help you decorticate a bit all this information!
Very nice from you as always, thanks!
Finally I got to request my data! I'm so exited to see what the data looks like considering I've been try-harding this game since mw2019. Just took 5 attempts to get the request to go through.
was it giving you an error too?
@@sstrahann Yeah, just had to keep on trying untill I got it
Really insightful analysis! I think there are some fascinating opportunities to dig deeper into the temporal aspects of this data that weren't covered. Since we have timestamp data for each match, we could analyse skill rating volatility across different times of day and days of the week. This could reveal whether SBMM behaves differently during peak vs off-peak hours when player pools are different sizes.
Additionally, I'd suggest implementing a time-series decomposition to separate the 'roller coaster' effect into its trend, seasonal, and residual components. This could help quantify how much of the skill rating variation is systematic vs random. Using something like an autocorrelation function (ACF) could also reveal if there are any consistent patterns in how long it takes for skill ratings to cycle between peaks and troughs.
For the correlation analysis between KD ratio and skill rating, I'd recommend using Spearman's rank correlation instead of Pearson's, since the relationship appears non-linear at higher skill levels. You might also want to look at conditional probability distributions - i.e., given a player's current skill rating, what's the probability distribution of their next match's rating? This could help characterise the 'rubber-banding' effect more precisely.
I've done similar analyses on other games' matchmaking systems, and one thing that often proves insightful is looking at the first derivative (rate of change) of skill ratings rather than just the absolute values. This can reveal whether the system adjusts ratings more aggressively after certain types of matches.
I'm glad K/D isn't the most important because my K/D is only 1.4, but I'm an objective player and I play Ranked. I die a lot just trying to capture points because teammates won't.
Def plan to submit info as soon as it...hopefully comes through lol.
Love seeing statistical analysis like this so very excited for more results!
Thats why i loved playing infinite warfare because you see the kd of the lobby before going in game
how come nobody is talking about win/loss?
As someone who cares more about my W/L than my K/D and plays to win each match (unless I’m grinding a specific camo on a crap gun or something which will be relevant in a moment), the thing that ruins my fun the most is when I can tell the sbmm/eomm/whatever is constantly putting me on teams with pylons who completely ignore the objective, won’t even walk through a hardpoint by accident, won’t even throw grenades at the guys capping B or give me any assistance at all when we’re spawntrapped on Nuketown and I’m fighting against the most objective-horny dudes in the world on their team that swarm B like locusts the second I start capping, you get the point.
I’ve noticed that if I win too many games in a row, I’ll be punished for it by being put in the most ludicrously one-sided lobbies where I’ll play my heart out and still lose a game of hardpoint like 17-250 while still being the best player on my team by a country mile.
And inversely, I’ve noticed that if I bite the bullet and ignore the objective for a few games and focus on trying to get double-kills with the combat knife or whatever it may be, I’ll start getting put lobbies that are more evenly matched. And if I’ve *really* been ignoring the objective for a few games and letting whatever happens happen score-wise, I’ll start getting put in lobbies where someone on my team carries the objective and I’m filling the dead-weight role instead haha.
I’ve even had games of kill confirmed for example, where we’ll start off crushing it and be up like 40-15 and then suddenly two players on their team and a player or two on my team will disconnect and be replaced by new players, only for the score to shortly flip in their favor and SO OFTEN they’ll end up winning by 1 which drives me bonkers.
But TLDR/point being, it’s very obvious to me that my games are being manipulated behind the scenes to try and account for my objective play. It’s very much not just tied to score per minute and K/D, whether you win or lose and how much you try for the objective plays a big factor too and I hope you’ll include a section on that in your breakdown video because that’s the angle I’m most interested in by far.
I play mostly hardpoint and I live on the point and am usually at the top of the scoreboard with a W/L ratio >2 and pretty much a 1.0 K/D ratio (+ E/D ratio) but my skill rating per this file is negative. I don't see anywhere on the file where it takes into account time on objective. I agree that the matchmaking must also largely be taking into account W/L cause when I win a few in a row the game tries to humble me by putting me with a team of pacifists who cannot shoot the enemy and actively run away from the objective.
I'm glad my 2-box data was interesting - super excited to see what you find once you're able to collect all of this data!
Regardless of what people feel about 2 boxing, your data feels like it holds some real sigfingance in this conversation. It was really interesting.
2 box ng is for weak cowards. No 2 ways about it
To be a coward in a video game is just embarrassing
Have to pick on weak players thats the only way u can feel good about yourself LMAO
@@JfkSush I decided that I didn't want an algorithm manufacturing my experience in the game, sorry that upsets you!
One thing that annoys me a lot is that in game it shows elimination death ratio. But assists count as elims - so I never know if I’ve actually got a positive KD during a match ..
still don't have mine. did it the day of the first video on it.
Mine came in literally 45 mins ago, I submitted the request 3 hours after the initial video was posted
Still waiting myself. Fingers crossed will come soon be interesting to see it all
mine came 2 days ago, must be low on the list or something
Same
Same here, looked for updates and kept saying there was none. Now it says I've never requested anything lol
Got my data after a couple weeks wait. it's been interesting looking through it. I'm not at all surprised they collect it and use it to match make. It's no secret COD lobbies are 'skill' based, it's what makes it addictive to play. As a low level casual player I have no interest in getting randomly added to lobbies with die hard players, your team might still win rounds but if there's one or two players dominating, it'd just be frustrating. My skill rating for BO6 has been consistently around minus 320 (low -450, high -200), KDR is on average 1.07, score per min is 422. I looked into correlations of 'change in skill rating' with things like KDR and SPM. On the full time base of the data set there's not much correlation, but within a single session you can see them move together as you warm up and then get tired.
make sure you do a regression on skill and percentage of time moving. would love to see if the game favors rushers or campers
I am willing to bet on the game favoring camping. My guess is that more time moving increases your skill rating giving you sweatier lobbies.
I would love to see that data too
@shari0nplaying slow or methodically is generally better for a good K/D than playing fast and being constantly rushing around to new places (even though streamers will have you think otherwise) So I’m not sure that will be the case
@@emwols that may be true. the annoying thing is that no one wants to play like that
@@blikthepro972 I play like that when defending an objective or getting close to higher streaks. Otherwise yes I totally agree
Hey Ace, Just got my skill data, gonna send it to you. One thing that stood out to me is my skill rating in black ops six is quite low as of lately 50-100 & often times in the negative wih only a small patch of games in the 200s. Yet other games like MWII and MWIII my skill rating is in consistently higher in the 300s & 400s?
This is odd to me because all of the games i play are literally sweat fests. I also played the same way across all those games & the same modes. (90% S&D) All I do is snipe at a very high level 95% of the time.
I Often get matched against FaZe members, YTers TikTokers, TTVs etc & have proof of that in my YT vids & tiktoks.
Another thing to note is Ive never reverse boosted or 2box on any of these games.
My KD is hovering around the same between all three games (1.1 - 1.4) sniping only mind you.
I think a huge missing part is definitely the skill rating of the lobby and players we are going against this has to tell another story on how our skill rating is calculated because something isn’t adding up to me between games. My skill rating particularly in bo6 has such large fluctuations in it yet almost all my lobbies feel the same.
Also there is ranked mode for skill ratings and ranks, keep Casual, you know... Casual...
Unfortunately it seems like the ranks don’t mean anything since it’s all skill rating based, which defeats the whole purpose of ranked
@@HaidDaSalami1 Damn, I haven't played ranked, I just don't care for it, but it's sad to know that it's not properly implemented.
Yep, I’ve made an alt account this week. Played pubs until I got 50 wins, absolutely smashed it. Went into my first game of ranked and had to play against pretty good G3 /P3 players..
I remember when cod didn’t collect all this data…
When lobbies were made based off connection
8:12 arn’t you just describing ranked mode?
Keep watching!
Lol
@@TheXclusiveAce The funny thing is, they base your ranked SR goal, and therefore your supposed rank division, on your hidden MMR. So, I’d say the hidden MMR is more valuable to know than the rank division you’re in. It would also be interesting to study the correlation between your MMR and your SR goal
I hope CDL pros share their data that would be interesting to compare against low and average players.
What 5:23 is 2 boxing ?
joining a lobby of a new/intentionally low skill account in order to get easy lobbies.
Back in MW3 you could basically feed the opponents for 30-40 matches and then get 10 people that can't shoot, and 1 player that was also doing the same.
As this is mostly unsuccessful for veteran players, they started creating fresh account, joining a match of that account while it's in progress and killing every 8 year old Timmy they could aim at.
It occurs to be that a big part of the problem with the discussion around matchmaking is that SBMM/EOMM do not accurately describe what's going on.
I prescribe that the most accurate name would be Performance Based Matchmaking.
The skill Rating absolutely need to be seen by the player prematch and see who you are playing against at least afterwards.
Everyone is gonna leave
Leads to abuse easy
Why?
You would end up with half a team in the match as people try to avoid good players.
Stacks will shop around for easy lobbies
Way to go, Ace! Very cool and interesting stuff!
if activision allowed players to see the overall skill rating in public lobbies, then I imagine people would just start backing out if they see a level that they consider too high for them. And activision will never allow that.
Yep. But we should be able to see our own skill-rating at minimum.
People are already quitting in droves. If anything they may feel challenged to do better.
I imagine they would let you see you're overall skill in your personal stats. Lobby ratings could be displayed after the fact in a recent stats area.
@@spartanx169x Nobody cares about how they do or how much their challenged its casual people are there to fuck around with friends and grind camos. This a system that belongs in ranked/competitive games.
@@MC_605 You may be right, you may be completely wrong. Unless you have a source that says otherwise. Other games(not COD) had this information and its always beneficial to the player to be able to see it.
Can't wait to see what insights you can pull from this study.
Sent my data in. Hopefully I didn't mess up any formatting lol
Can you include the users’ RANKED rating as well? That will help us understand how the skill ratings correlate with actual player skill and lobby skill
Ace, I got my data today and looking through it, I think the skill rating might be the *lobby* skill rating. There does not seem to be a correlation between round performance and skill rating. If skill was an individual rating, you should see a rough correlation between performance and skill, but it doesn’t seem to impact skill rating much at all.
I appreciate what you're doing, Ace, trying to shed some light on Activision's opaque SBMM algorithm. I want to help you out and share my stats, which go all the way back to MW2019 alpha (right?) and beta tests and every release since then, up to and including BO6 and every version of Warzone. (Until I uninstalled in frustration for the first time ever, about two weeks ago now.) Unfortunately, I'm getting the bullshit "We've encountered an error. Please try again." message when I click on [ Submit ] after logging in and seeing that they do indeed have all my linked platforms, etc. I'm sure they noticed the uptrend in people requesting their data and decided an "Error" message was the simplest way to put a stop to that. Fuck this company, fuck their games (both literal and figurative games). I'm done with them until I hear they've made significant changes to how they do business. Knowing Activision, that simply means I'm done with them; they won't change their business practices, no matter how vocal their veteran community is.
#UninstallCoD
#BoycottCallofDuty
#BoycottWarzone
#BoycottCoD
Love seeing this info. Excited to see more on this!
What is 2-boxing?
Low skill account searches for a match.
Low skill account finds match.
Your actual account instantly joins the half filled lobby your low skill account just found.
Voila. Everyone in your lobby has now the same skill as your low skill account, instead of the skill your actual account, which joined the lobby later, has.
@@Chrissy717thank you for the explanation, I had no clue either.
i miss when the lobby leaderboard was a standard thing in cod
If I was implementing this system, I would have 1 KD as 0 skill on that graph.
So < 1 KD is negative, and > 1 is positive.
Of course this is just a rough, simplified way to describe the system.
We would have to know how they are valuing certain things in game. But I would say that 1 KD is a starting point.
Not a bad starting point, but this would only work for kill-based modes and not objective. In MW22 (the last game I played that had actual K/D’s), my K/D was only .85ish, but I only ever played objective modes and farmed points for score streaks, so my W/L was like 1.4 and my team would routinely steamroll pubs.
@@qpeghnl97x yeah, that is why I said, it would be interesting to see what goes into equation and how.
The problem is that it doesn’t take into account the skill of others in your lobby. A 1kd against bots and a 1 kd against pros aren’t the same thing.
@mrdjwess645 Your KD is an average of all games that you play. So it is not a problem.
@ yes but you wouldn’t see a skill increase if you played against better players over time because your kd would average out to a 1 over time. So your skill would be the same if you were comparing the stats of pro players against each other or Christmas noobs against each other.
I don’t play COD anymore but I do find this really interesting… my skill PLUMMETED after MW2019. To a degree that made me *so* confused as to what happened between games…My experience between 2019, and II, and III was night and day. The only thing I could think of was either some tweak to the damage values or the SBMM system that I wasn’t privy to… and the data is leaning towards the latter.
MW2019 had very loud footsteps. Was a very campy game. People are much more confident to rush in BO6.
I'd come back to COD after years off if they released one without SBMM
Guess you’ll never come back then
They did, it was called xdefiant and nobody played it
@ xdefiant is not an equivalent to the Call of Duty IP and development
@@MGrimm1226guess the other CODs prior to mw19 don't exist then 😂
@@MGrimm1226reached over 1 m players. People left cuz the servers were shit
This is why I have been loving the 4 player Co-Op games. With no multi player.
After work I just want to chill and game. Not be playing like if I’m on rank play
As someone who just got shadow banned for nothing, they can go to hell with their systems. It's all nonsense anyway, evidently.
Data sent. Looking forward to the detailed analysis!
Is it weird if your left ball hangs lower than your other two balls?
Other two?
If but yeah they hang differently for temp control
I don't know why you have 3 but I think it's fine, if one is larger than the other that's when you might need to be worried because it could be testicular cancer, but could also just be a hydroceal (can't remember the spelling) that's a build up of water in the testicle which isn't as serious but still should get it looked at
If you're from America I wouldn't worry but if you're from a place with actual good Healthcare I'd recommend just to be safe getting checked out by a Doctor especially to see why you have 3 testicles
Great job dude. Im very interested to see how this turns out. I wish they'd just lighten up a bit on sbmm. I dont want EVERY game to be my application video for the CDL.😂😂 just chill mostly.
doesnt condone two boxing but condones the COD alhgorithm
I’ve been speaking out against their matachmaking system for many years
4:30. I know that had to suck. Not even 5 seconds in Double Agent. Also, love the info you are sharing, thank you.
I hate first comments; but I accidentally tapped on the noti. Great to see yall 😊
I’m a data scientist in my day job and ran some Python data analytics on my skill data (sent it to XclusiveAce and Nero). I was able to back into the SBMM algorithm with an XGBoost regression model and find the biggest correlations to skill, and it is mainly your skill in your last 8 matches, last match score, lifetime wins, and time played. I started BO6 with a fresh account and my skill was -320 for a long time. Then they started to set it to random values and the algorithm kicked in, shooting up to 75, before hovering around 0. What threw me was the #1 skill indicator was lifetime wallbangs! Perhaps to catch people walling, so it jacks up your skill to the max. Either that or they assume wallbangs = skill… 🤓
Feature Importance
Lifetime Wall Bangs 0.33780137
Skill_1 0.23941274
Skill_2 0.114405714
Skill_3 0.037927497
Lifetime Games Played 0.032032818
Skill_8 0.02712443
Skill_6 0.025060102
Score_2 0.018084485
Lifetime Wins 0.013437924
KD_Ratio_1 0.006047078
Skill_10 0.005968014
Score XP_5 0.005213942
Shots_6 0.004903351
Skill_4 0.004834562
Kills_1 0.00427607
Rank at End_4 0.00380922
Rank at End_3 0.00364872
XP at End_6 0.002988484
XP at End_7 0.002617068
XP at End_5 0.002531297
Score XP_1 0.002427384
Score_10 0.002400091
Hits_2 0.002327372
Weapon XP_5 0.002243329
My regression model feature importances (weightings) that go into skill. Showing the ones that account for 90% of my skill variance.
The people who designed this is probly watching and laughing at us trying to figure it out they suck
Always love the videos brother ! Great job, keep up the good work!! 👏👏👏👏👏👏👏👏
MWIII >>>> Call of diddy blackmail 6
Hey Ace, if I could make a request, when you do the video can you take a look how K/D ratio correlates with W/L ratio? It seemed when I first started that everyone who was above a 1.0 K/D had a negative W/L, and a lot of people with a negative K/D had a positive W/L. Would be interesting to see if that is true when you get the data. Thanks for taking the time to do work on all this skill rating stuff and keeping the community informed.
That’s mostly a mindset people have , they value a win or a high kdr. My w/l is only in the .70 range but my kdr is 3x that.
I could pretty much guarantee that K/D ratio and W/L ratio have an inverse correlation amongst the community. And it’s not because Activision/COD is conspiring against “good” players with high K/D. It’s because people that value K/D over all else pretty much screw over their objective playing teammates every match by not making any effort whatsoever to try and contribute to winning the match and thus the low W/L.
I think the problem with BO6 is that when you do well, it puts you in a lobby where your entire team is much weaker than the entire enemy team. MW2019, BO:CW and MW2022 they'd put you in a lobby of strong players on both teams and a weaker one here or there on each team.
the real problem is, Bo6 is pure garbo
BO4 had this too
That is exactly what I'm experiencing in BO6. None of my team will be able to break a 1.0 KD and the entire other team is like 2.0 or better and it happens way too often. I'm inclined to believe that they are not making any effort to balance the teams.
@@Calma123_ Hard disagree, the mechanics are great, but many of the maps are too small for the fast movement speed, even more so than MW3 2023's remade maps. MW2019 remains the best CoD to this date.
@@spartanx169x yup, that's what sucks. Because it's alright if 1-2 plays on the team are weaker, but if the entire team is weaker it compounds/snowballs and makes it even worse.
If I ever get it Ill happily share it with you. Only been since you dropped the other video haha.
I can confirm switching between pc and ps5 has zero affect on the skill rating when looking at my data, even when ps5 typically is cross play off.
I wish these stats could explain how solo queueing in Domination is the worst experience in COD history. Either puts me in an existing game or gives me 5 teammates who lack opposable thumbs.
521 SPM and 125 OBJ SPM and constantly queued with guys who have no idea what's going on. If am not going as hard as possible, my team gets steamrolled. Thanks, COD
after hopping on BF4 player run servers with no SBMM, you get a wide range of skill across the 32 player teams
I will be sending my data when I receive it. I requested mine within an hour of your initial video.
And just received and sent!
New sub from me . In very intrested in this info and appreciate you taking the time to give it
I’m a crimson player (prob could be iridescent but cheaters and lack of time). Have a 2.25 pub kd. Average skill is 415.44. Highest skill game was 622. Sent the data over.
And in the data I received. Dozens of items are saved that prove that the game is rigged at all levels: Operator skin, operator bundle, the execution chosen for the operator, AND THEY TELL US THAT PURCHASES IN THE STORE DO NOT COUNT FOR THE EOMM 🤔🤔, the percentage of time moving per game, the time without dying, shots through walls... etc, etc, I UNDERSTAND BETTER AND MORE AND MORE WHY I HAVE THE IMPRESSION THAT IN EVERY GAME I AM A BOXER WHO TRAINS IN FRONT OF A MIRROR. THE GAME IS REALLY COMPLETELY RIGGED.
Right? Like why keep track of which operator you used, unless its used to adjust the games damage. All these people who bought stuff from the store are getting a pay to win game.
I play splitgate now. Amazing game with a portal twist. Splitgate 2 is coming out soon too!
“Sorry we encountered an error”
Can you do a video on how you take the data received to the graph you use in the videos?
Ace one thing I just noticed that I’m sure you will notice is that something is up with the Hardcore game modes when it comes to skill rating. It seems to me like the ratings are inflated for hardcore modes
I havent received mine yet and I requested it immediately when you posted the video about how to do so. Hopefully soon!
Ran-snack definitely had no SBMM, I haven't played as much Cod in years as I did when that playlist was out, it genuinely felt like old cod and my EKIA/D for the prestiges I was doing it was up to a full point higher than in the other lobbies.
Feel like every time I'm watching a video of yours you're playing on skyline lmaoo
You can organise and improve the layout with excel, like with conditional formatting.
Okay Ace. Highest skill rating I've had was 300. Its fluctuated between that down to the negatives and back up later on. This is all Warzone data I'm giving you from Black Ops 6. 718 games played. 3360 kills 2007 deaths. 1.76 KD. My elimination death ratio is 2.60. 5428 eliminations 2091 deaths compiled in a total of 782 games. So for 64 games, I'm missing match data that would truly show what my warzone kd is.
Freedom is a great thing. Give the players the choice. Those that want to play general lobbies/modes with regular SBMM and one where it's extremely loose or non-existent. Let the players decide so we can end people whining about being stuck in one or the other. Problem solved right?
In the last CoD, I always played the s2 map pack, s3 map pack, s4 map pack. I found there was often seemingly minimal sbmm and I could regularly dominate with 3kd and 4kd. When I went into the regular general playlist, I didn't really enjoy the game at all.
This is why I still play Sea of Thieves, Roller Coaster Tycoon, and single player story games 😂😂
7:48 lmao... I've said it before as a member of communities in other games, just give us our ELOs and of the lobbies we play against. Just like Ace says, I think people's gripes with SBMM would go way down if they knew where they stood in everything. "But that's what ranked is for." Sort of... I think it's very different to have a "ranked ELO" on top of a regular "skill rating." Going into ranked is like an agreement that everyone is putting their ELO on the line as the objective of the mode. Getting matched based on skill in a more casual mode makes the ELO only a byproduct of playing however players want to play. Finally, the speculation that limited-time modes have not affected "skill ratings" in the past and that perhaps SBMM is not used as much or at all shows, in my opinion, that games would be better off just being clear about which modes have SBMM and which don't. These changes would clear up any illusions people have about how good they are or why they're struggling, etc. As a noob, you would join a non-SBMM mode at your peril, but you'd know that going in. As a veteran player, you wouldn't necessarily take so much pride in stomping a lobby with a weak skill rating, in part because everyone you would be playing against would know the implication, as well (perhaps that it would actually be more impressive if you went and did that against some tougher competition, whether in ranked or just another, more casual SBMM mode). Anyway, this is great info! Thx! A not insignificant chance that these revelations actually change competitive online shooters for the better in the future.
Why aren’t they tracking info on server/lobby quality?
Like packet loss, latency, ping, or how often a game crashes ?
It looks like sbmm is weighted the same as ranked, in ranked if your gold 1 and you play against gold 1s and hit top of the leader board and you go positive then let’s say you get 100 SR then if you play a lobby of gold 3s and hit top leaderboard and go positive you’ll get 135 SR , I’ve done some very limited testing of this and it seems to work that way , so in pubs your skill level against the same skill level brackets then you’ll barely go up but if you go wayyyy up in skill level brackets then your skill ranking will go up more after a good performance
So, I'm assuming in ranked MP, they are mostly putting you in games at your own skill level and not in a random lobby of your rank. So if you are a decent Plat player, all the golds, plats and diamonds in your lobby would basically be the same. Which is SBMM essentially.
My games are between 300 and 400 skill rating and its borderline unplayable. Lots of cheaters and crackheads