Ok, so as an (intermediate) competitive chess player and casual math nerd, I have a few comments. 1. Awesome video! Despite what I'm about to say, I completely agree with all of your conclusions. I've been following you for a while as well and I just really like how thorough your analysis always is. Keep up the good work :) 2. So /technically/ the Elo system is kinda terrible, and was quickly replaced by the Glicko and then Glicko-2 systems. Long story short, if Elo assumes that a player's performance is a bell curve (which is the nerd way of expressing that d20 metaphor), Glicko and Glicko-2 basically give variables to how wide the bell curve is. This allows the system to account for some players being more volatile than others, and just makes for a better system overall. Based of your explanation of the Overwatch rating system (I've never played Overwatch in my life, I don't even own the game!), I would be shocked if they weren't using some version of the Glicko-2 system. 3. Even more um-actually-ly, the Glicko rating system is known to be slightly deflationary, benefiting lower rated players ever so slightly. It's actually been a problem for the International Chess Federation, and they are currently considering increasing the skill floor needed to get a rating and just adding a bunch of elo to people rating below/around that range (which is insane, but also super interesting). In theory, this means that it's quite likely that pure implementations of the Glicko system would benefit lower rated players slightly, which is another interesting thing to note about the whole "Elo Hell" thing. However... 4. ...Most games don't implement a pure version of the Glicko system, which can lead to unexpected behavior. A good example of this is Splatoon 3's implementation (X Power), which guarantees a certain number of rating points if you can win 5 games out of a set of 7. This leads to effectively unlimited ratings if you meet a certain skill requirement and are willing to grind, and as a result the whole system has to be reset every month. This makes it a little difficult to make the argument of "Elo Hell doesn't exist because the system doesn't allow it!", because we don't really /know/ if the system allows it. Unless I'm missing something, Overwatch doesn't show people's Elo's, nor does it make the equations public. For all we know, the Overwatch team could've implemented some tweaks to make the game "less frustrating" that result in Elo Hell being a totally plausible thing. Now, do I think that is the case? No. However, does it have precedent in other games and is technically possible? Yes. 5. Last one I promise. So, from my experience, most people who get deep into intellectualizing "Elo Hell" eventually reach a singular argument as their endpoint: There statistically almost certainly exists at least one person stuck in "Elo Hell" to some degree. What do I mean by that. Well, let's say we're playing a game where we flip coins. On heads, we gain $10, and on tails, we lose $9. Statistically, we would expect to gain money playing this game, and most likely that's what'll happen! However, if we get a couple million people to play this game, it's highly likely that at least a couple people are going to be deep in the hole, and some very deep. Theoretically, it's entirely possible that, in a game with random attributes, somebody is getting murdered by lady luck. Elo Hell is a good cope because you can almost guarantee that somebody exists there, and so you can just say you're that person. Now, the chances of that being the case go down pretty quick as you play more games and as the actual skill gap increases, but they're never quite zero. Of course, there are a number of things you can do to get yourself out of Elo Hell even if that statistical version of it does exist, but explaining that to a person with a defeatist mental attitude is quite difficult I've found, and I'm fairly certain that there's a strong correlation between people with a defeatist mental attitude and people who believe in Elo Hell. (Looks like you basically said this last sentence yourself in the vid, good to see I'm not hallucinating!) Thanks for reading my horribly rambly essay, and have an amazing day!
Well said. I had a conversation on an Overwatch stream today about how Elo hell isn't really a place, but a timeframe. If you are truly below your actual rank, if you play enough games you'll get out of Elo hell. But it's definitely possible to get an incredibly bad string of games and get stuck below your rank. There was a season where I dropped from diamond 1 to plat 2 and got stuck there because of an incredibly unlucky bout of leavers and throwers, but the very next season I was able to peak masters 5 after the matchmaker started being nice again. And while I cannot say I'm a perfect player and didn't lose any games that I could have won if I had performed better, it's incredibly demotivating to keep playing when you get an unwinnable game after an unwinnable game in a row.
I like the concept of following someone because of their personality/interesting insights even if you don't play the game. It's the same for me with WarOwl, even though I've not played Counter-Strike in more than a decade, save for some nostalgic days here and there.
While I agree that using elo hell as an excuse for a person not ranking up over time is an absolute cope, I actually think that people in lower ranks DO experience a form of elo hell. Its just that elo hell doesn't drag your rank down, it instead takes away some of your agency in affecting the outcome of matches. What do I mean by that? Well I’d like to compare my experience in two competitive seasons, one where I went from low silver to mid gold, and another where I went from low/mid gold to low diamond. For both these seasons I started at a mid gold skill level. I know this because I had been in mid gold for like the past 7 seasons of competitive, and I hadn't really been trying to improve until that climb to diamond. In my climb back to gold, it felt like many of my matches were decided before they even started. Despite my gold level play, I lost a ton of matches due to teammates that seemed to lack a basic understanding of the game. I also of course won matches that were steam rolls against those types of players. But generally, win or lose, it felt like I had less control over the outcome of my matches, my level of play didn't really matter if it was gonna be a steamroll one way or the other dependent on who got those 5/20 players (to use your dice roll analogy). Matches that were incredibly close and both teams felt mostly even, ones where I felt my impact was expressed in the outcome, felt few and far between. My climb to diamond on the other hand, was a much more fun experience. While actually more difficult personally, as I actually had to improve my play to earn the higher ranks instead of just climbing back to where I belonged, I felt like I had more impact in every match. Even when I had teammates that were like a 5/20, in platinum they still had a baseline competence that meant it was easier to play around their shortcomings. Those close matches, where the teams were largely even in skill, felt like they were happening far more often. In my experience, elo hell does exist in the sense that the variance in skill level at lower ranks has a much larger impact on the outcome of matches, with many resulting in one sided experiences, whereas in higher ranks the baseline competence of players ensures most matches are at least somewhat close. You will still drift towards your correct rank no matter where you are on the ladder, but I think that the removal of agency in lower ranks results in a more frustrating climb, and makes it easier to fall back on the elo hell cope. It feels like your rank is being dragged around at the whim of the matchmaking system rather than your actual level of play. All of this is of course subjective experience, hence why I overwhelmingly use "felt" to describe stuff, so I accept that is could be a bunch of baloney, but that's my two cents.
I can understand that. Sometimes the disparity of both teams seems so big, that one or another just get rolled. Some wins and loses just feel free. I've won so many games being absolutely bad, and also lost a little doing the most as I could. Now that I'm on plat 3, this seems to be more rare to happen.
something I've noticed about my friends in gold and silver a bunch is that ones who dont play comp sessions over a hourish in length or play in anyway consistently, believe they are stuck in elo hell alot more than the ones who play regularly. ive even vod reviewed their games, and i cant deny the game session they do have when they hop on are very streaky. it leads to nights where they stop playing after 4 matches as they didnt win a single one and a player a rank higher would not have made a difference. and other nights they play and they win 80% of thier matches mostly due to the other team just being substanically worse.
Definitely agree on some things. i think bronze is the hardest rank to get out of and i WILL die on this hill. i’m high plat/low dia now but i was hardstuck bronze for like 2-3 seasons in ow1 and man it was so tough to climb. nobody knows what theyre supposed to do down there
I have been hard stuck silver for many seasons now and it feels like every time I’m 1-2 wins away from gold, I go on a losing streak and end up right back where I started. It’s frustrating because I see things like tanks repeatedly running in to 1v5 matches and dying the complaining they didn’t get healed to me as a support having higher dps than one of my dps while still maintaining heals on my teammates yet I somehow I am solely responsible for my rank being what it is.
I think actually elo hell does exist, just not in a way that you think. You think elo helll is just someone complaining about them being hard stuck in a rank that they deserve, but think that they don't. When in reality, elo hell is the 51/50. I like to call it where if you win 51% of your games. You will climb, but you will climb so slowly that it will feel like you're stuck in the rank because you will win 5 game, Lose 4 games, win 5 games, lose 4 games. And that is what elo really is. Now, if someone doesn't play enough games to see their rank eventually slowly climb and go up, it will Feel like they're Stuck at a certain rank. So instead of it taking 10 games to get to their True rank of maybe plat 5. It takes 500 games to go from silver to Plat 5 because of all the winning and losing, even though technically they deserve plat 5, but because of the 51/50, they climb so slow. It feels like their hard stuck in silver/gold. and that is real elo hell, and it does exist. imagine someone who doesn't have the tenacity to see their games through, tilts and gets angry ,upset. Then what happens? Is they end up throwing games, which now even lowers thier rank even more, and so even though they could be placed right into a plat game and hold their own. And in fact, even win a lot of games. But because of them Not having the tenacity to see their games through because of how frustrating it is to win 5 loose 4. They just don't ever see their true rank. And that is real Elo hell
Yes elo hell is when someone is focused on their rank instead of actually trying to get better. That's not elo hell though, it's the hell of not being able to hold oneself accountable and that hell is almost never isolated to a game. This is a deeper personality problem that people write thesis about and it will forever fascinate me
elo hell was truly truly real the first 8 seasons. I spent a few hundred hours to figure this out (doing multiple unranked to GMs, bronze to GM, climbing on my main, etc.) and it's because it was nearly impossible to actually rank up. It took me like 50 straight wins to get out of bronze which is frankly ridiculous. I'm just hoping season 9 fixes this issue now that there's more direct feedback.
Your point about players at the bottom being incentivized to tilt is what I would define as elo hell. Not saying its the games fault that they’re at the bottom but the feeling of hopelessness when you win more games but don’t see any visual progress due to the huge gap in bronze, makes it feel like hell. So I believe in the feeling of elo hell but not the theory of elo hell because I remember what it was like to start at Bronze 5 and see no indication of climbing after hours of grinding. Took a while but I forced my way out lol
I was kind of on the fence about this at the beginning of the video. I don't play comp but from an outsider's perspective, the idea of elo hell seemed plausible. Your explanation is excellent. You're right, it just took a little bit of patience to humor it for long enough for you to change my mind. Good video, man. Good video.
21:39 i'm exactly the type of person you're talking about here. i was DEEP in the bronze 5 SR hole, maybe because of my performance in ow1 or something, hard to know. it took me almost a year to get out of bronze 5 but once i got out of bronze 5 i was able to hit silver 5 in a matter of just a month or two. that massive SR pit at the bottom of bronze 5 is the closest thing to elo hell that i believe in, only because when the SR stat is hidden it can feel like you're making zero progress no matter how much you're actually improving. i think you hit the nail on the head about this topic in the "what do" section. it's really hard to accept you're bad at something you like doing and want to be able to be proud of. the self esteem hit from this realization has definitely, at times, made me feel almost as much lack of control over my ability as the elo hell mindset brought in the first place. but it's possible to overcome that and being able to see the light at the end of the long *long* bronze tunnel definitely helped
My biggest problem with my own personal "Elo hell" was just my pure, unadulterated, HORRIBLE luck. More than 70% of my games had literal trolls (Throwing themselves off cliffs, running around being unhelpful hiding in corners (Note, i'm not talking about flankers or sombra here. Just literally people squirreled off doing jack all.), or verbally abusing everyone on the team. (Often more than one of these.) After a while, I gave up even bothering TRYING to rank up. Then I ended up quitting shortly after 2 came out. I probably was somewhat close to my actual rank, but I definitely knew I could play better than I did when playing with randoms in queue, because I did play with higher ranked players in various tournaments and I invariably performed much better with teammates of a higher rank who understood strategy and comms. It's a team game. If the team can't team, then the game don't game. Like I said, I gave up. It wasn't worth the frustration anymore.
This will always exist as an outlier type deal, statistical improbabilities that disprove the existence of certain phenomena doesn't mean it's not possible for that phenomena to manifest in certain cases for certain periods. I also have some pretty abysmal luck, such that it does actively play a part in slowing my rank climbing in a noticeable manner. However, after a certain point it still does boil down to a skill issue. That's why I'll generally believe someone that says they're, say, competent enough to compete well in low diamond but are stuck in midplat because of throwers and lack of coordination, but they're almost certainly doing something wrong if they claim to be masters or something. Over time, it is almost certain that you will skew towards your real rank, and it just isn't feasible to say that your skill isn't holding you back. Either you aren't good enough to play consistently and don't deserve that rank, or you're good enough but unlucky so you're slightly below where you should be, thus making it not that big of an issue. With that said, there is one nuance to this in that you could be in a situation of the likes of being GM of one specific hero but don't perform nearly as well in any of the other heroes, so you end up stagnating as a result. Or, more commonly, not being able to play hitscan well or something. That is most certainly still a skill issue, but a more understandable one since nobody wants to actively play unfun heroes just to gain rank. Still though, in terms of ranking for the role, that's definitely a massive skill issue.
@@DragonEdge10 yup you're spot on with this one. Given enough time and a decent amount of games, everyone is at their intended rank within a relatively small margin of error. The outliers can certainly be a few divisions off, but by no means are there masters players stuck in plat for long.
Ever since OW2, Blizzard has added an insidiuous element to ranked: grading on a curve. They're not hiding your sr, they've genuinely stopped tracking it. They only measure your rank relative to the position of other people. This means that if you go 5 wins in a row, depending on how long it took you to do so (say you play 1 game a day) you could actually *drop* in rank. It's why I quit. Edit: I forgot to mention why I thought this was relevant: believing in elo hell is a sign of tilt, and either way you're screwed here: you either doomqueue and end up losing a bunch of games due to losing focus/tilt, or you try and take it slow and limit your games per day and get kneecapped by the system. I got lucky and caught it red-handed, but the diminished returns from not playing often enough could mean the difference between going positive or negative in a 5-3 cycle, for example. The player would never know they're being screwed over, only knowing they dropped in rank and probably blaming it on elo hell.
Like, I genuinely cannot see how my own biases can in any way influence the fact I went 5 wins 0 losses and went from top 28% to top 35%. That's just a drop. Either my theory is true, that I'm being punished for not playing often enough, which honestly; what the fuck Blizzard, or I'm being made to play matches that have only decreased my rank despite me winning, and then why would I play?
Dude, thank you so much for this video. I’ve genuinely thought ive been stuck in elo hell, but this video kinda made me take an objective look at things, and ive noticed that ive been coping hard.
Honestly, the best counter-argument to the concept of Elo Hell I've heard comes from Flats: the enemy team is just as likely to have a bad player as your team. If you can guarantee for yourself that you are a good player, your team will have 4 potential bad players. The enemy will have 5. Even with 'elo hell', the odds are in your favor if you really are a competent player. Edit: oh you literally just mentioned that at 12 minutes haha
I don't play anymore, and I accept now that the average player's skill increases at a rate faster than my own, so I accept whatever rank is given to me. I used to cope hard in OW1. I could peak plat, but my average match would be silver, and I was left holding the bag. And, unfortunately, the first 2-3 seasons of OW2 were the devs saying "oh we oopsie'd the ranked system on our end", which really made it easier to cope. I now believe that any...frustrations, or beliefs, that the system is rigged against you might still exist, with varying levels of merit, but it is not ELO specifically causing the problem. Could be a bad patch, could be time of day, region (look at how OCE/AUS/NZ players talk about their ranked experiences, for example); could be a myriad number of things. And, I feel like for a lot of people, especially lately, there's more underlying, negative factors, that come together consistently enough to give off the impression it's an ELO thing; which is just not the case
as someone who personally took 100 wins at a 65% wr to get from b5 to b4 it feels like your going in circles for a long time im gold now but that frustration i wont forget and i agree with how dumb the lack of transparency is
I would like to propose an anecdote and ask for an explanation: My friends and I started the game around gold. Left and came back years later and placed immediately into diamond and stayed there. Following season we got bored and threw back to gold (no we don’t throw anymore). Following seasons couldn’t climb out of gold at all. Recently came back after they reset the mmr and am back to high diamond/masters, but literally couldn’t climb out of gold before the reset.
nah brother, i didnt believe in elo hell til i got a game where both my dps had literally 0 kills. Not a single one. tank is trying to create space and supports were on par. Get those shitters out my game bruh
I started playing Overwatch again recently and that led me to your channel and I fell in LOVE. Your content is great! I love the way you meld Overwatch and more deep topics together, keep up the good work!!!!!!
I've liked a saying Eskay said on one of her Lucio guides. about 30% of games you can't win, 30% you can't lose, and 40% your individual skill can decide the difference
The problem with the ELO system is that it isnt well designed for a hero-shooter like OW but for games like Halo where everyone has the same kit. Overwatch clearly has hard counters; the bronze moira will shutdown the Gold Genji with ease but still can be less skilled overall. An amazing mercy isnt going to win against a winston diving her. The widow with a mercy pocket and Shield tank will always come out on top against the better widow if they dont have the same mirror composition. That right there makes ranks inconsistent and not reflective of anyones skill. In ranked they expect you to win games not taking into account which heroes your teammates are picking and as a DPS player it is near impossible to carry bad hero picks especially when it's your tanks picking wrong and also positiioning poorly. I woulnd't mind the grind if they made the game less team-work dependedent and give more agency to individual skill. The game is so counter-pick meta that how can any system guage skill by just looking at wins? I've won games without breaking a sweat just because my HERO countered theirs, not because I out-shot/out-smarted them....I simply outpicked them. Going Sombra against a widow isn't exactly a brilliant play just the right one. On the other hand I've busted my ass on 76 or Ashe and dropped 60 elims just to lose because Tank decided to chase and die while on onvertime and they rushed us. Another whole issue with ranked is that noone is actually their rank, people have smurf accounts, this makes each game inconsitent among skill levels. SO many smurf accounts I know can barely climb out out plat on one account but hit GM with ease on another....that right there is all you need to realize the system is quite all over the place and not a reliable source of determining skill.
Just got out of a Splatoon tournament and I’m feeling the whole high to low energy drain so I’m so excited to watch this. The Dunning Krueger video was why I subscribed so it’s nice to see this topic again
Thank you so much! Kind of a fun fact about this script, but I actually wrote this video and the upcoming one out of order - the members’ choice poll for this month was a dead tie right up until the end so I didn’t start writing this script until the 11th, and while I was waiting for the members to finish deciding I ended up writing a whole other video that’ll be coming out on March 1st. I had to burn the candle at both ends a little bit to get this one out on time, so I’m super glad that it still worked 🥰
10/10 no notes. A great deconstruction of "Elo Hell" and what you can do to change that mindset. I make this comment longer for the algorithm because I have nothing more to add to the conversation than what you have already said.
lol. i don't think that's how it works. if you play with someone who has a much better skill rating than you, you will likely start to win more games. it's once you stop playing with that duo, it becomes clear to everyone in the rank: you are boosted and you don't deserve that rank. elo hell, to me at least, is relying on the idea of "these events are out of my control," which is not even true. and being boosted is having your duo, imo, equalize your mistakes or making more positives plays (has a better understanding of how to control things ig?), and possibly having a better understanding of the mistakes they are making (in terms of, positioning, etc). as an example of elo hell, i lost some of my placements yesterday, i could whine about how it's my team mistakes/elo hell, but i also could've joined vc and called out when i was going to ult or when i saw a flanker. and an example of being boosted, is when i duo with someone who has a higher skill rating (or smurfing) than me, this would be extremely unfair to the enemy team as i already have an advantage (assuming they aren't being boosted). basically, these are two different things. being boosted does exist lol
@@kitenlovar699 people also miss that the higher ranked player in the duo is also boosted because they are always playing in lobbies at ranks below their ability level.
Elo Hell exists, that is to say in the mid ranks some of your games will be effected by other players fluctuating in their rank up and down and that will mean you will lose some games irrespective of your performance.. that happens in every rank but it’s more likely in a mid-rank like Plat. You also don’t factor in smurf accounts and grouped up players who are synergising certain heroes and utilising one of them being lower ranked to skew the matchmaker in their favor especially against solo queue players that cannot go in with a team strategy and coordination baked-in. then you have players that have been boosted, sometimes because they grouped-up and won games based off those advantages that they now no longer have, all these factors contribute to the sluggish feeling of what most people would call Elo Hell, it’s not fair to just say there is no validity in these issues that really do effect the game and the ranking system as just people trying to cope You could make the argument that this is all more of an Elo Limbo, you will overall break through the effect with enough games played, if you are a masters-grandmasters you’ll probably completely blast through it unless you are using a highly team reliant hero like lifeweaver, who has great healing yes, but can’t dictate the outcome of a game in the same way, say. an ana can, and yes some heroes are better at climbing than others given the skill tier and that heroes specific balance state. However, there is a more stark version of Elo Hell in Mid-ranks on console, why? because there are a lot of smurf/alt-account kbm xim players on console that will be in your games, they thusly have an artificial advantage in mechanical aim that drastically effects their team and your team in terms of this particular “skill” metric, and may also come attached with a pocket mercy to amplify the advantage of this cheating behaviour, but they often times won’t have comparable gamesense and other skills for the given rank or they would have risen from the rank they are in, so they can stomp some games off the back of this advantage but equally lose games when their advantage is minimised/countered and their pocket mercy isn’t there to bail them out or amplify this factor.. this creates an environment in which games are sometimes won and lost at the matchmaking level because the system wasn’t designed to factor in these ximming wild cards, it’s most noticeable in High Plat/Low Diamond on console, and varies with the region you play in, but effects console across the board across all ranks, by time you get to GM almost everyone is using Xim and it’s really not fair to say they are more skilled than people using controller when they are cheating.
Love the vid Viv. Theoreticals ahead: Early in OW1, you had 1 competetive score. Shared between all 4 types of role, but there was no role lock. So it was perfectly possible to have games where you could preform as a support and just not have the space to fit yourself into the team, causing you to (on average for mono role) lose more games. Being a plat support, but not being able to hit shots from hitscan dps or confident in positional tanking would lead you to being in a gold/silver position due to circumstances barely within your control. I don't think this is called ELO hell but it certainly is frustrating, even during the learning process, as it can often feel like the quickest (not best!!) way to improve is to find players who allow you to focus on strengths rather than shore up weaknesses. I do wonder if ELO hell is more of a cope in games without role lock due to this extra step in the system to blame. As an addition to the above, I'm not sure on the maths on this, would role influence allow you to push faster rank changes as a result? Like, good tanks (for eg) can carry a game, and you are a good one, vs support having little effect and you are a good one; would the tank player in such a situation have an advantage for reaching their "true" elo? All things equal, they'd have to play less games surely? or do the mistakes of a good support in a losing match having less penalty keep things balanced?
I feel like I'm stuck in Elo hell, and nothing can convince me otherwise. Here's why: I have a friend in real life who's not very good at the game, yet he only plays ranked and expects me to join him. He's in Silver, whereas I'm in Plat, but when I play with him, I end up dropping to Gold because of his performance. This limits my ability to climb higher than Plat since I don't play enough solo games. However, I recently started playing on an alternate account and got stuck in Silver on it, while on my main account, I consistently fluctuate between Gold 1 and Plat 1 because of my friend dragging me down. My plan now is to try playing with him on my alt account to see if I can break through the Plat barrier.
I think this video is a cope too. Sorry I watched only till 19:00 but that's the part I want to discuss. Your is a very well rationalized projection is something you know partially (as you stated). The details on how the ELO system is implemented in OW2 are not public. What's public is the Blizzard Match Making Rating system patent. In few words, this system is built to assure a fast and efficient match. Now, fast is paramount, and efficient has many meanings. I don't want to discuss this because the info is publicly available, you can see what they see as efficient. The problem with this video is that you analyse the system as something very abstract. The player you describe is a type of player, the type that stuck in bronze or silver, maybe gold. Now, players can interact in many ways with the system. You failed to mention the groups. Groups (duos, trios and 5 stacks) exist and have a big influence on the game. Within the grouping, one can do many things. One can "boost", i.e. have good group mates, or "min-maxing", i.e. having group mates worse than you. These tricks are well documented and you can experience it very easly in a game session. Besides the tricks, there's the smurfing too. If the anti-smurf system catches you, it boosts you. This is why every content creator can make an "unranked to GM". I invite you to search on YT for a "bronze to GM". For what I know, there's only one guy who made it and it's a very interest document to demonstrate how bad the system is. You cannot get better in this system. Not because the system is that bad, but because players abuse the system flaws and the people who can interveen don't really know what to do (hence, this season reset and changes). Also, what about the floating players? (unranked players placed in ranked matches for placements) I have a lot to say but I stop there. I just don't understand why the community cannot come together on these issue. All other games can pint-point the flaws of their games, this one does not. I really can't see why. Very good video BTW, you put a lot of effort and knowledge in your videos and the results are great. Congrats
@@TheViveros I finished the video and I have nothing to retract or change. Just thanks for mentioning that climbing low ranks is systemically harder than climbing mid and high ranks. But honestly, the psychological analysis is likewise partial and subjective, but I'm not interested in arguing that, it's your view, very articulate, but partial. Anyway, I would have liked from you an answer on the points I raised, I think those are the most important.
Alright, fine, let's deconstruct it point by point. 1) This video is not a cope, because frankly, I have no idea what you think it would even be coping for. I don't think you're using the word right. 2) If I only partially know this after all my research, then you know even less given the lack of research demonstrated by the lack of understanding of the system even after having it explained, so acting like you have the capacity to act as an authority and yet I apparently don't have the capacity to opine on it in a way that could be correct is just acting in bad faith. 3) Every other player is also operating under the matchmaking system. You benefit from the system as much as you are hurt by it because it's random chance that evens out in the end. It's a wash. 4) Rank is inherently, fundamentally abstract. The whole concept of Elo is based on the principle understanding of skill as something nebulous that cannot be objectively defined, and the ranking system is built around trying to evaluate it under that pretense. This was mentioned in Part 1. 5) The matchmaking system deliberately tries to match groups with other groups; if the other team has a duo, your team also likely has a duo. If you're five-stacking, the other team is also usually going to have a four- or five-stack. Grouping up does have an effect, and that's why the matchmaking system explicitly tries to balance that out by matching groups with other groups. 6) Yes, boosting exists. Please refer back to the section of the video where I said "You can win for reasons outside of your control as much as you can lose for reasons outside of your control". The ranked system is still working as intended because all it can do is try to evaluate how a team with that rank will perform against another team. I'm terribly sorry, but there is no way to devise a ranking system that can peer into your soul and discern your exact skill. This is the best we've got, and at the end of the day, you're just as likely to be playing against a boosted player who's worse than you or playing with someone who's trying to boost their friend as you are to be playing against them. 7) You are describing the anti-smurfing system as if it's disproving the validity of the system when in reality it's demonstrating that it works. A smurf's rank gets dragged upwards if the system can identify that they're better than their rank currently indicates, which proves that the system *is* capable of recognizing when someone is clearly better than their rank. Beyond that, if you're 8) Yeah, Bronze to GM is rare because going from being among the bottom 5% of players to among the top 1% is extremely uncommon. Most people don't go from being objectively terrible to objectively amazing at a single game unless they didn't belong in the bottom 5% in the first place. If you earned that Bronze rank, then you probably will never be GM unless you put in a ton of work. It isn't a demonstration of the system failing to say "Players who are very good can start a fresh account and hit GM fast, but players who are very bad rarely - if ever - hit GM". That doesn't mean Bronze players can't climb, it just means that they would have to do a lot more work to climb to GM than someone who's starting out as pretty average. The fundamental difference in ability between a Bronze and GM player is so massive that a Bronze player will likely never become GM for the same reason a GM player will likely never become a Bronze player. That's not a flaw of the system. 9) You absolutely can get better at this game. My first ever competitive season, I placed at exactly 2000 SR back in early Overwatch 1, and a few seasons later I was in Masters. How? I played a ton, reviewed my vods, and figured out how to get better. Then, slowly, my rank fell. Why? Because I got busy with life and uni and wasn't playing as much, was falling behind the patch notes, and got worse at the game. Anyone can get better, and it's a cope to pretend otherwise. It's not the system's fault if people aren't good and refuse to try to get better. 10) Floating players aren't players that the game has no understanding of. There's a reason you have to play a bunch of QP before you can even try ranked, and it's partially so the game can get a feel for where it should start you out in your placements. It's not flying blind, it's just operating with less data and will correct sharply if it's wrong. 11) The community can, and has, come together on this. Just because some people are still clinging onto the cope for dear life doesn't make them correct; Elo hell is not real, it is a cope, and the sooner people accept that the sooner things can actually move on. Almost all of these arguments are points made in the video that were seemingly deliberately ignored. I have gotten probably a dozen essays like this in response to this video, and I cannot stress enough that this is *just* you coping.
@@TheViveros This is a very good answer and I thank you for that. I'm sorry I didn't catch these arguments in the video, but it seems to me they weren't as clear as other points you made. I have to say that your response about going from Bronze to GM while correct shows exactly what people refer to, incorrectly, by saying "Elo hell". That's not just a coping mechanism, but the right reaction to a big flaw in the system. Why the same person, the same player, has to win more than 200 games to get from bronze to GM and only 5 if the account is boosted? (Boosted from QP, as yourself admitted). As you said, it's the anti-smurf, it's how it works, but then why say that the last word on your progression is the rank and the rest is coping? The fact that the system works sometimes or even many times doesn't mean the system is always right and beyond questioning. I would say there are so many factors at play outside of the single player skillset that one cannot but persevere and hope, if one want to climb. I really don't like this emphasis on the individual and its performance. I don't like it in social sciences, I don't like it in a team based videogames. (I say “like”, I have arguments, but that’s off topic). There are ways to consider the single player skill in a team-based game, but if the system takes in account only wins and losses (or predominantly only win and losses), i.e. a volatile outcome, that’s an arbitrary set of data where the individual performance is very diluted. Also, all this argument on the "grind mind-set" vacillates a bit if you consider that the fastest way to climb if one's hard stuck is getting a fresh account. Or, as I said before, by grouping (example, "min-maxing"). Anyway, I don't want to force you to answer me on every point, I just wanted to argue that things are not as righteously monolithic, nor easily scientifically analysable (at least with the current set of information we dispose as public). All social and psychological factors implied in every human skill progression are at play in OW, as for every game. There are, on the other hand, transparent games (like chess or, in some measure, sports) and others more opaque. OW2 clearly falls in the 2nd category. That's what I meant by saying "this video is a cope": things are not as clear and determined as you say and your explanation is not all-encompassing. Sorry if it seemed otherwise.
When I came back to Overwatch 1, I got put at about 680 SR. Near the bottom of bronze. For awhile I believed that I was trapped in ELO hell at the time because I had been low silver in 2016. It was truly a cope and I realized I didn’t know shit about the game and couldn’t aim. So I poured hours and hours into improving. I remember being at about 950 SR playing Ana and it was a game on Temple of Anubis. I literally killed 4 every fight that match and our team’s Pharah came into voice telling us to thank us for the carry and my Rein jumped in and was like “it’s pretty clear our Ana is by far the best player in the lobby.” It was at that point that I knew I was improving and within a month or two I hit plat for the first time. At every new level there was a new set of skills, knowledge, or mechanic I needed to learn to improve and I would once again get that Anubis feeling of dominating. Eventually I hit diamond and then switched to Valorant, where I once again went from bronze to diamond. Now I’m playing Apex, and am on my way to diamond again. Improving is a process and your rank is the feedback as to whether you’re improving faster than the general player base. If ELO hell exists how am I now in my third climb in a third different competitive game?
I cant agree that its just cope. But the reality is that you do just have to get better. The reality is that to confidently get past the limits of "Elo Hell" it requires that you be the X-Factor. You cant afford to lose the duels. You cant afford to NEED peel from your teammates. You cant afford to be out of position too often. You cant afford to rely on your teammates to pick off the Mercy who has rezzed free behind a wall everytime. You HAVE to be the teams X-Factor if YOU want to climb.
Two things: 1. I've been looking forward to this video pretty much since it was first introduced in the polls, and it did not disappoint! 2. That Widow gameplay 😳
Honestly I'm shocked that it took this long, I really thought that this topic would win in its first poll. And you have no idea how happy I was while I was playing that Widow game knowing that I'd be able to actually include some good gameplay in this video lmao
I believe a concept similar to Elo Hell exists, though it's not just constant losses or anything. My original account, before OW2's multiple changes to the rank system, was hard-stuck Bronze (1200 SR) on Tank (Tank being my main role). Getting down that low was 100% my fault; I entered ranked as a new player, inexperienced and not very knowledgeable. But as I improved, I didn't see any improvement with my rank. Every game was frustrating, win or lose. I felt like I had to hard carry every single game. I joined a PUGs group of variable skill levels around that time, and suddenly that issue disappeared. My performance outside of what most would think of a bronze player in those PUGs made me think. I made a new account, leveled it up, and played my placement games. That new account placed Gold (2100 SR) on tank and until OW2 it fluctuated between high Silver (1900ish) and low Gold (2100ish) every couple of seasons. While I don't doubt that I may have been able to get to that rank on my original account eventually, the dedication and continued frustration that would have been required made it effectively Elo Hell.
Your experience with the ranked system is normal. It's common to have a peak rank and fall down and fall down a rank or two and struggle to carry the games on your own at times, you will get back to your average rank eventually, the only way to rank up is to be the best player in the lobby. Also, new accounts start around gold you just can't see the exact elo before placements (used to be plat in ow1 they changed it sometime early on in ow2). So that could have been what got you out of bronze without having to play games in it.
Yeah the problem isn't that people don't eventually reach their true rank, it's that "eventually" part. It's monkeys and typewriters. Sure EVENTUALLY they'll write shakespeare but I wouldn't blame someone who got bored after a thousand years of waiting and nothing remotely interesting happening. The same concept is true in overwatch. While yeah, given a thousand games the system has enough data to pretty comfortably say you're in the right rank, those 1000 games are incredibly frustrating to have to play through when 499 of them were losses you felt you had no hand in and yeah you technically won more than you lost but not enough to where you felt that 1000 could've been reduced down to 100. In short: the problem isn't that the system doesn't work. It's that it works slowly, and humans don't have the patience of a machine. Elo hell feels real even if it objectively isn't because just like Pi could have a one million long stretch of just 0s, so true could an elo system have a stretch of 1000 bad losses and/or frustratingly difficult and unsatisfying wins before you reach your "correct" rank.
i refused to give up my ow1 cosmetics and it took me over a year to get from the pits of bronze 5 to silver (which i finally hit 2 days ago). worth it tbh, it's been a very gratifying anime underdog story
it boils down to this: game lost: Could have done better. game won: Nice, I can do better. game lost because people are throwing: It is what it is. next game
Honestly, if you take this mentality into your games you will immediately perform better. I think it's an undervalued lesson that lots of people learn from youth sports, where win or lose the game is over when the final buzzer rings and you've gotta move on. Someone might've been an asshole or pulled some dirty moves on you during the game, but you can't throw hands in the parking lot over *everything* - you've gotta just shake their hand, say "Good game" and move onto the next one. If you sit and stew over every game that doesn't go your way, you'll become bitter, and when you're bitter you can't get better.
I started overwatch when overwatch 2 launched, got myself in bronze 5 dps (genji one trick mostly) I really believed in elo hell because I went from the whole of season 1 and 2 with a net total of 5 blade kills. Season 3 came and I was having 10 blade kill games and I’m beginning to top my lobby in kills and I’m somehow only bronze 4. I hated the system because I’d queue with friends and they’d go up from silver 4 to silver 1 and I’d go up from bronze 5 to 4 which reinforced my belief in elo hell. This video helped explain why I felt like bronze felt like a never ending pit. Thank you.
I explicitly do not play Overwatch, and I have fuck all knowledge about how rankings work in general, but this mf has somehow roped me into several of his video essays about the damn thing That is to say, you’re doing an amazing job, and I hope you keep making content because it's genuinely amazing and well thought out
Wow, great video. I truly started to enjoy overwatch so much more after accepting that I'm just not as good as I thought. Everything you've said I've already accepted and knew years ago, but you just put it perfectly. Also, you keep mentioning bronze players but just from my own experience, diamond/masters/gm players need to watch/understand this video just as much if not more, since they have a general understanding of the game and are decent enough to have a massive ego but also not top 1% -> 0.1% like they believe they are. Diamond/masters is rotten with this "elo hell" ideal. I've also always been a masters player barely peaking into GM for literally the last 6 years... so i have little experience at the lower ranks but this 100% applies to higher ranked players just as much as bronze (if not more).
Given the new information on the matchmaking algorithm I wonder how it affects this approach to elo. During ml7’s dev interview the dev mentioned something about a machine learning algorithm and boosting mercy players with lower healing since high elo mercy’s have lower healing since they are boosting more but the model picked up on just the lower healing and might’ve thought it made a better mercy. Anyways, nice vid!
Really I’d just love to know what data and formats the model is fed, like I didn’t realize that it does or at one point could see some form of healing stat.
The only thing I would like to add is that elo like most evaluation works better with more data to pull on. I mean the rank of people who play more regularly would probably be closer to their "correct" rank since their placement is less influenced by outliers. Naturally this works both ways if I play 2 matches a week or something games I won or lost simply because someone left the match are more influencial than these types of games for people that play 5 games a day simply because of the sample size the ranking system can pull from. I hope this makes sense
One question I have. If the system is using all the math to determine which team is likely to win, what are they basing that likelihood on? If it is based on performances from previous games and your current modifier status (positive or negative), how does the system track what is going on in the game? If you are on an upward trajectory then you are probably maining a character and farming certain stats and plays. What happens when 10 people constantly flick between heroes to counter others despite not being particularly good with those characters? Surely if you want to climb then you must stick with a hero you are good at and play that character consistently to avoid swings in your calculated team value which could negatively impact your next game. It makes sense that the system is more stable with consistent information.
perdón por el comentario en otro idioma pero una vez literalmente le dije al otro support "deja de llorar y empeza a curar" porque se estaba quejando del tanque pensando que era el problema, cuando el problema es que tenia menos de la mitad de curación que el resto de suports en la partida, cuestión escucho y ganamos tan fuerte que para el final de la partida solo quedaba un support en el equipo enemigo (le dimos la jugada de la partida)
The problem with Elo in pvp team based systems in games isn't that it takes i to account everyone's score to score you. It is that it doesn't correctly evaluate individual impact. I have lost and won many games. And a lot of wins and losses aren't too off depending on if you actually work together with your team. But I have 100% been in games where I far overperformed and still lost because my teammates sinply didn't get on the point and I lost rank because of it. Many people deal with this issue. The system needs to not only evaluate a number over the number but also how well everyone performs individually as in who they took a pick on (how important their kills were) l, when they were on the objective (how important was it to be on the objective if they were on it alone), if they tried to keep someone alive as a support, how much they supported a teammate being attacked, and if their deaths were the fault of them not being with their team or their team all dying and leaving them alone. Many factors play into winning and losing, not just numbers, but actual gamesense. And a system based on numbers could never accurately rank someone's individual improvement in that. The only way to get out of elo hell is to find friends who undertsand how each other play to play with and work together. It is a team based game not a solo based game. This is why making "stacking" not be allowed in high rank is an ignorant and foolish way of working with the game. If the best players are also friends and play together then they are the best because they are good at doing the one thing the game is about, TEAMWORK. Without trying to assist your team in a team based game you will never win. If we could see our number and the game would follow actual individual gameplay properly and tell you what you can improve on then there would be no reason to "cope" and people who get annoyed would actually have no reason for it. But the game doesn't care about how well you actually do. It just cares about the number and the math. It is just a poorly made system for multiple player systems. Literally just be nice to people and try to find groups you are decent with and play together in a group with them specifically. Everyone plays bad, everyone plays good. I have not found a single person who is 100% consistent. And the math cares TOOO much about cinsistency. A hunt never goes as planned, and a pvp match is just a hunt with more complex problems to deal with.
the problem is even if im the constant variable, i still need a copious amount of games to average out, and who REALLY has time for 100+ overwatch games a season? most people dont, or even have the will
Since beginning of season 9 every 3-4 games i encounter a person playing music in voice chat and then leaving after a minute or two. Outside factors, like trolls and people deranking their accounts for whatever reason bring up lots of questions about the validity of "no such thing as elo hell".
@@bismuth7730 And realistically you remember more the people who played badly than the people who played well due to a negative bias. I garantee you that it's less than every 4 games for a long time and even if it is it's every 4 games, just play better and if you lose because of these people every 4 games you will rank up and get out of the rank range in which people are always bad.
One factor that I didn’t catch being mentioned was leavers possibly being more prevalent at lower elos. That’s one reason why I favor the idea of a surrender option to just cut the bad matches short: You shouldn’t have to play out an entire defense rd 3v5 after 2 players leave on your team. Rolls are so frequent they have become normalized. Much of the OW toxicity comes from being trapped in these frustrating, humiliating, helpless feeling situations. If 3 players on a team elect to surrender, the match ends and the enemy team gets their full pts while the enemy team gets half downgraded what a full loss would have cost [say 15% calibrated down rather than 30%]. Damage mitigated; frustration ended swiftly; easily forgotten. Bad games ended quicker, good games sooner and longer. It’s much easier not to tilt and adopt a mindset of: ‘Whatever, GG, Go Next’ If you can actually go next. If a player leaves the match, that player is blocked from queuing for 24 hours, takes a loss and automatic derank one division penalty. The abandoned players receive a draw. Three strikes and your acct is suspended till next szn from comp; your internet cannot handle it if that really is the reason. This part may just be my anger talking: Season-Long Suspended leavers must also start next szn in a new small shame rank below bronze: iron. Iron cannot be deranked to, you can only be placed there by leaving suspension. You must gain 300 sr and have an endorsement rating of 3 or better to reach Bronze 5. At the very least, offer forfeit as an option the enemy can extend and the badly losing opponents accept. Maybe it won’t;t get used often, but the times it will people will be grateful for. Maybe even incentivize it with a 10% sr boost to the winners for offering. Many would take a sure win with a lil extra boost over spawn camping for 5-10 minutes.
I agree that if you’re just blaming your teammates and not yourself for being at your rank, that is wrong but I feel like we can all agree that there is that couple of games where it entirely was your teammates, fault hands you were the only person on your team that was positive and order contributing to the overall goal
From this video’s script: “It’s also important to remember that the pendulum swings both ways here. You can absolutely lose an Overwatch game through no fault of your own - but that also inherently means that you can win an Overwatch game through no success of your own. It’s a team game, and the results - barring the rare instances where one person absolutely single-handedly carries through a dominant performance - will always reflect the overall performance of a team, not the individuals. You will lose games you performed well in, and you will win games you performed poorly in. Complaints about Elo hell exclusively focus on the feeling that it’s only ever affecting you and your team and only ever doing so negatively, but statistically it helps as much as it hurts your team and the other team is dealing with the same situation.” This specific point was addressed - multiple times, in fact - in this video. Maybe watch the whole thing before commenting.
@@TheViveros thank you I completely agree and like you said you can completely lose a game due to no fault of your own but if you’re consistently trying to climb and you can’t, that’s where you belong
I think one problem is something you said in a previous video about why Overwatch is a failure as a competitive game. Because of the randomness of hero switching, unselectable match types, and meta changes every quarter season that can occasionally change how a person's main (or strongest set of heroes), no matter how accurate the method of measure is, it's measuring something that isnt a straight line that the players are walking. It's more like taking a leap of faith every match and sometimes you land flat on your face because there was nothing to land on. It doesn't justify it, but hopefully the communication implemented in season 9 makes this easier. I stopped playing competitive because i realized the pressure to do well was causing me to fail. I do excellent in casual because I can concentrate on just getting better, not dragging others down with me.
also I'm soooo hyped for the new study! the original study was what drew me in to your channel. I loved how much data you had and I really enjoyed seeing it presented, so I look forward to the new study results
not trying to be rude, but i think you should try to talk a little bit slower. It's hard to compreheend it sometimes. Keep up the good work, this is one of the few channels that actualy put interesting content
I think one part of this you didn't mention on the why we cope section but is very important is the amount of effort and mental stress is necessary to stop blaming and keep on trying your best when you feel like one of your teammates is not performing. If you have say an 0-14 Genji, you could put a lot of effort to try and compensate his poor plays but this will just make the game last longer while the odds are still not in your favor. This results in a game where you felt like you played well and tried your best but still lost, very tilting and you will keep that resentment on the next games. This the problem the devs tried to solve when releasing Overwatch 2 and the card system by the way. They wanted players to not focus on the outcome of each match but instead look at the results of a collective of games. They failed spectacularly of course, but I suspect a lot of tilting will come back after getting hit with "Expected" or "Reversal" following a match with somebody you felt was throwing.
How I would describe what "Elo hell" is to me in OW, is that at lower ranks, some skills that are very useful in high ranks just do not translate. I feel that in higher ranks enabling team mates, by creating distractions for the enemy team, your teammates are able to capitalize on that, this is especially obvious when playing support. I love playing support in diamond, but hate it in low plat/gold. Because what you do affects the outcome less often, you consequently lose motivation and focus and start to perform less and less. Because OW is a multiplayer game, and not single player, the "victory requirements" are also determined by the players you play with, thus the "climate" of a certain rank. For instance in Diamond I can count on my teammates doing certain things and following up on things I do, this then allows me to count on them and giving me more options. Whilst in gold/plat they would just keep doing their own thing. Every rank has it's own set of rules. So I am not saying that the gods of the ELO algorithm are conspiring against me, just that the skills i have and how I LIKE to play (and thus tend to do) are not as valuable in every rank.
elo hell may well be a cope, but i can safely say i don’t remember ever having a comp match in bronze when i didn’t have gold objective time, even when i was healing, meaning i never had anyone on my team who was on the objective more than i was. when staying on the objective IS the thing that makes you win or lose, it’s frustrating to know my team never cares as much about it as i do. it just doesn’t feel fair. i’m very willing to admit that i’m not a gold player, but i just hate consistently being stuck on teams that don’t seem to actually try to win
I never really played much rank in OW, but I played a lot of Ranked in league and DOTA, and the idea of ELO hell is a lot less prevalent in DOTA. My theory, DOTA shows the number. I had a huge jump in the last DOTA MMR reset, indicating I was probably under-ranked before they changed the system and reset my MMR. But I never felt like I was stuck under my ‘true’ rating. Because my true rating was right there in my profile. Not as visible as the fancy badge, but easy enough to find.
i know quickplay is not part of the discussion, but whenever people tilt in quickplay, i always tell them to quit the lobby if they hate playing with me so much. In ranked, i feel less inclined, as i usually try to do my job, and switch whenever i have to but just yesterday, when a rando went "dva never play the game again" to a friend of mine, i first of all told them to shutup and quit if they didn't want to play with us. Then, since my friend WAS affected by that, i made sure to let him know that i didn't care about his skills (he *isn't* good at the game, but that's why i play Quickplay with him) I'm okay with people silently complaining to me, and I'm also okay when people complain directly to me, as i am 100% aware that my gameplay is not perfect, and i usually take their comment as a way to see my gameplay under a different lens, or, in the case of toxicity, i take it as a hint to push myself a little harder, think a little harder, and see how that goes. i won a game with 2 toxic people in the team in qp, and as much as at the start i was doing relatively shit, and my team *was* carrying me, i eventually managed to feel like i was pulling my own weight, and one of them did recognize that, as much as it was salty. the other just went "got carried lol" and i ignored them completely, since they didn't realize just how much the team as a whole did. but yeah.
Just for fun, here's my pet theory for where the elo hell cope comes from The meta shifts depending on the rank you're at. However "good" or "correct" play is often defined relative to play at the high ranks. It doesn't matter if you're playing "correctly" if the way you're playing doesn't suite the actual lobby you're in. People cope be cause they're playing "correctly", but get stuck because they never stop to reflect on the actual problem
For an example: Dead by Daylight is a 4v1 asymmetric horror game. 4 survivors try to escape 1 killer. When they introduced skill based matchmaking, they made it so that if a survivor escaped, it counted as a win for them, and if they died, it was a loss. At high ranks, altruism is optimal. It increases the whole team's chances of survival considerably. However, SBMM only cared about your individual survival. If people play selfishly, the person playing altruistically almost certainly dies. So there's a sort of prisoner's dilemma situation going on. As a result, a number of really good streamers got kinda stuck in the mid ranks because they kept trying to play altruistically, and got punished for it
An easier way to rank up is to duo with someone who's good at a role that isn't your main role, so if you're a pretty good tank player you should duo with someone who's good at support or dps.
My cope list Step1: blame myself Step 2: look at scoreboard to see who to blame Step 3: explain how the neg dps was causing the loss justfor them to yell back Step 4: yell and scream Step 5: cry and cope
i never believed i was in ELO hell (or the related "losers queue" concept) but i was entrapped by the opposite idea. A "winners queue" type ELO hell. This would be "i suck at the game, i always get rolled. I should be bronze but im gold cuz my team carries me" Tho not with overwatch cuz i dont play competitive there. i dont know, maybe it is kinda coping on my part too cuz of how one-sided matches tends to feel, tho maybe thats just blinding myself by purposely forgetting the genuinely good times.
Every other person in the game has the same chance to be throwing. means if you never throw, your team will always have an advantage statistically, you are guaranteed to have a above 50% wr given enough time
I am guilty of not roasting my duo more. Except when he plays support. Oh boy, this man CANNOT lamp a Tracer bomb to save his life while playing Bap. I told him twice in one game he died to pulse, and I watched it happen both times. Both landed on the floor. My buddy made it to plat before I did on support
I feel like mercy is an exception to this rule. Even GM players tend to struggle trying to climb ranks with her because she is so heavily dependent on your other team members. But then again, this may just be another form of justifying why i've been hardstuck in plat lmao
I mean yeah but some characters just can’t carry and if you need to carry sometimes mercy isn’t the pick and a more damage oriented support is. Having the ability to reasonably well play a carry character in your pocket is important.
I'll say that I trust Elo, sorta (there are other systems), but I don't trust Blizzard's ability to code without bugs at this point. They have admitted problems in the rankings, and some of those took more than one try to fix. Without open code, I can't fully trust them at this point. I think another point not often raised is that Blizzard have admitted that ranks aren't equally distributed. There are larger masses in different ranks. This can mean that moving in one of those ranks actually takes many more games than someone at another rank. They've made changes to make this more visible, and I hope they do more on this front.
I see what you mean but I think some people are in elo hell and some people are using that as cope there are a lot of factors that play into a win and 1 person will not change the tide of the whole game sometimes not all the time
Once again, I am begging people to actually listen to what I’m saying before commenting. A quote from the script for this video: “It’s also important to remember that the pendulum swings both ways here. You can absolutely lose an Overwatch game through no fault of your own - but that also inherently means that you can win an Overwatch game through no success of your own. It’s a team game, and the results - barring the rare instances where one person absolutely single-handedly carries through a dominant performance - will always reflect the overall performance of a team, not the individuals. You will lose games you performed well in, and you will win games you performed poorly in. Complaints about Elo hell exclusively focus on the feeling that it’s only ever affecting you and your team and only ever doing so negatively, but statistically it helps as much as it hurts your team and the other team is dealing with the same situation.” Your point was mentioned. Elo hell is not real, and pretending that these factors only affect you - and only affect you negatively - is a cope.
elo hell is an actual thing and im currently stuck in it, i was low masters in overwatch 1 and havent played since 2020 so i decayed to silver 1 in ow2, when im going 35/6 in a match where my tank is constantly feeding and my other dps is going 3/12 genuinely how the fuck is it somehow my fault that we lost that match
Had a similar experience with coming back to OW2 on an old OW1 account, i played close to a hundred of comp matches and my rank barely moved in gold. Made an alt account, placed diamond 2, made another alt, diamond 3, few more matches, rank jumped up. wouldn't be an issue if not for cosmetics on my account i guess. Feels like your MMR moves less the more matches you've played so it gets harder to climb as you improve. Viveros just likes the word cope
This doesn't factor in specifically that Aaron Keller hates me in particular and hath decreed that I shall not climb. For real though I've been on teams where one person has admitted to being a smurf for a much higher skill level and then other teammates STILL insist that they are being treated unfairly by the game because they lost.
Elo hell is 100% real. Last season i struggled to get out of plat 5, after the reset, higher and lower elos etc im now in diamond 3 and FLYING through it.
I just hope people get a good laugh out of the idea that you don’t just think Elo hell exists, but that you think it exists as high up the rankings as Plat.
Yeah bro it’s called the game underwent a significant change overnight and it’s affecting people’s performances. A whole bunch of people can suddenly hit headshots now that they couldn’t before. You were hardstuck bc you were at the correct rank for the state of the game back then. Your rank has changed because the state of the game has changed to something more favourable. This is not complicated.
@@TheViveros Look man its def not a one and done if im still performing well and holding my own in higher ranks. I get ur point but my experience SHOWED its real.
IF its any consolation for people I used to play comp a lot in Overwatch 1 back in 2017/20119 era. I was gold/plat. When I was a baby brand new to overwatch I was silver my very first placement and rose pretty quickly. Bronze at that time were people I assumed did not know how to play video games in general. Like not just lower rank but like genuinely perhaps children or those with sub 10 hours of FPS game time. I booted up OW2 when it came out and it threw my ass in bronze for the first time ever. These bozo's in bronze look like gold back in 2018 no freaking lie. I don't play a lot of comp and being placed in bronze def made me not want to play it so it wasn't till several seasons later that I put any effort into getting out of bronze. But as a whole if you have been playing this game for a long time and have gotten better over the years but your rank is the same it isn't that you haven't gotten better. It is that the quality of each rank has risen with the years. At least on the lower end as someone who played a long time ago and has recently jumped back into comp this last year. Its like inflation in the real world. "I used to make 50k a year and support a family of 5!" "Yeah that was in 2011. That doesn't cut it anymore." Elo hell isn't real. So if you need a new cope thats one for you. *Edit. As a part time league player I actually think an Iron rank would be great. If bronze is the largest rank in terms of broad SR it only makes sense to break it up. The number of actual sub 500 sr comp players has to be pretty small so 1000 and below can be Iron and 1000-1500 can be Bronze. I don't think that would be an issue. If someone is very low ranking giving them something to show for their progress is a motivator to improve. If someone just sees Bronze 5 every time they rank up no matter how good they do for say two, three or even four rank ups in a row it will just make them think the game is rigged against them. And I couldn't blame them personally. Not because it is but because OW2 not showing SR has been terrible. Just show us the number. Hiding the number does nothing and just complicates everything. Its like a big secret that the players shouldnt' and can't have access too. Poor design and no one likes it.
So in overwatch1 2.9k was considered elo hell. Even t500 players doing bronze-gm knew most there time would be spent in 2.9k. Why? Aaa a long list of reasons. It was a crossover rank but the people in it were either in all aim no brain or all brain no aim. In Overwatch2 this 2.9k rank doesn't exist anymore cuz the ranking system brought these sim players up and brain players a lil down. My opinion is that you shouldn't focus on the rank too hard. Just try to play the game
I don't personally believe in an Elo Hell formed by the actual game algorithms, mainly for the reasons that you've outlined in this video, but I do believe that some games have something fairly akin to it in their lowest echelons, primarily the lowest rungs of games like league of legends. Those places are absolutely rife with boosters, smurfs, trolls, and all manner of other situations that completely destroy games at a completely random rate that cannot be accounted for. In theory they would happen to everyone equally, and you being one of 5 on a team *decreases* the odds of one of those players being on your team, but in reality it often only takes one on either side to affect the outcome in a way that messes with whatever system they use.
For context I've been at above average ranks in every season of LoL that I've decided to play it, so this is not me coping at all. I completely believe I deserve to be exactly where I am. I just think that the gaming experience at the lowest ranks is completely awful for your mental, which will sometimes cause a downward trend in winrate even if your skill is high.
I don't believe elo hell is real, but people make it real. What I mean is, Elo hell is as real as a psychosomatic injury, or a phantom pain. There's enough non-merit factors that may determine whether or not an individual game or games may have results altered (a pc crash, bad ping, a teammate not playing their best that day), but it's not reflective of the entire season for the person claiming elo hell. Regardless, those determinant factors may be enough to convince the individual they should under-perform, and by bias, flake up their own lack of skill to it being a "already lost game", etc.
While the word cope is definitely triggering here lol… while i do disagree with the fact that “elo hell” as a concept is just nonexistent (def exists to an extent).. a video like this is refreshing tho instead of the very tired “JUST PLAY BETTER!” Coming from a streamer who only queues up with their fellow top 500 friends and plays for 15 hrs a day. I think there are definitely some rank ranges where stuff gets really weird (your bronze 5 case for example). And sure the teams should be “even” but a game that depends so heavily on dps/tank success (im willing to hear if this is not the case to someone else) i can easily see a support especially being kind of stuck somewhere at some point. Great video though! 👍🏾
8:15 I can’t get over how y’all didn’t contest there, tank was back and Kiri prolly had suzu so you would’ve lived. Great vid though honestly just gotta guh guh
what i dont understand is how im hardstuck low diamond on my main account but im my usual masters 3 on a side account. i load onto my main and i get back to back one sided stomps while i go to my side and the matches are fair.
Volatility is the reason. It's a "certainty factor" in the ranking system. When it's uncertain of your rank you'll move up or down more, which can boost newer accounts to a higher rank than intended for awhile (your Smurf) if you go on a lucky win streak. It happens, and has resulted in my Smurf also being higher rank than my main (quite often). The more you play on it the more likely it is you'll move towards your intended rank.
I think elo hell exists but doesn’t affect many. I’ve personally never experienced it but I have friends who I play with who are low ranked but pop off or play very well in my lobbies (higher lobbies) but in solo q they can’t carry. Idk what the problem is for him
I dont believe elo hell is real. Im pretty good at calling myself out on playing like a smoothbrain. But the whole tilting is a skill issue thing kinda opened my eyes. I mean i know i shouldnt tilt but hearing somebody say it like you did has helped me.
Only issue is, they use an MMR system, inspired by Rocket League. Rl is maximum 3v3, though, so it works better considering you can individually carry 1 or 2 people much easier than 4, also considering the roles in OW, where not everyone does everything, but you virtually have to cooperate for anything to work. Thats why it doesnt really work well in OW. Not saying this to cope, though, just criticizing OW's ranked system because its hardly good enough, even with the update.
While I agree that elo hell is just cope, there’s some serious fuckery that happens at the lowest ranks. Around 95% of my games when I still played was roll or get rolled and it made individual skill expression near impossible. I’m pretty sure this is due to the fact that accurate skill ratings start to fall apart a bit at the extremes. At low bronze you can have a person that doesn’t know strafing is a thing and someone who just came to overwatch from something like apex be only 100 sr apart, it’s really hard for the match maker to make remotely fair games
I mean, yeah? Blizzard’s openly acknowledged that this is the ranking system they use. They haven’t kept this a secret, I didn’t realize people would be disputing the actual ranking system given how well-established that fact is. Overwatch is even specifically named on the Elo ranking system Wikipedia page as a game that uses it lol Overwatch uses Elo, by their own admission, and this is how Elo works, so I don’t know what evidence you’re even asking for here.
I will just take your word for it then. I wasn't trying to throw any accusations at you, nor was I doubting your integrity. I am simply unaware of what blizzard has revealed about this. Just from playing the game I don't think you would get this information, but I could be wrong there as well. Just from playing the game the ranking system just seemed unclear and confusing to me. Although it seems a bit better now in season 9.@@TheViveros
I wanna say that elo hell is real because it does effect everyone i mean we are more than halfway through this season and t500 is still masters people are lower than deserved imo
elo clearly stands for electric light orchestra duh
So Elo Hell is the light section at Lowe's
i used to work at lowes and i can confirm lowes is hell
Mr Blue Sky!
Thank you 🤌
"I have become cope, I have become Elo hell."
The lesser known Oppenheimer quote.
Copenheimer
Ok, so as an (intermediate) competitive chess player and casual math nerd, I have a few comments.
1. Awesome video! Despite what I'm about to say, I completely agree with all of your conclusions. I've been following you for a while as well and I just really like how thorough your analysis always is. Keep up the good work :)
2. So /technically/ the Elo system is kinda terrible, and was quickly replaced by the Glicko and then Glicko-2 systems. Long story short, if Elo assumes that a player's performance is a bell curve (which is the nerd way of expressing that d20 metaphor), Glicko and Glicko-2 basically give variables to how wide the bell curve is. This allows the system to account for some players being more volatile than others, and just makes for a better system overall. Based of your explanation of the Overwatch rating system (I've never played Overwatch in my life, I don't even own the game!), I would be shocked if they weren't using some version of the Glicko-2 system.
3. Even more um-actually-ly, the Glicko rating system is known to be slightly deflationary, benefiting lower rated players ever so slightly. It's actually been a problem for the International Chess Federation, and they are currently considering increasing the skill floor needed to get a rating and just adding a bunch of elo to people rating below/around that range (which is insane, but also super interesting). In theory, this means that it's quite likely that pure implementations of the Glicko system would benefit lower rated players slightly, which is another interesting thing to note about the whole "Elo Hell" thing. However...
4. ...Most games don't implement a pure version of the Glicko system, which can lead to unexpected behavior. A good example of this is Splatoon 3's implementation (X Power), which guarantees a certain number of rating points if you can win 5 games out of a set of 7. This leads to effectively unlimited ratings if you meet a certain skill requirement and are willing to grind, and as a result the whole system has to be reset every month. This makes it a little difficult to make the argument of "Elo Hell doesn't exist because the system doesn't allow it!", because we don't really /know/ if the system allows it. Unless I'm missing something, Overwatch doesn't show people's Elo's, nor does it make the equations public. For all we know, the Overwatch team could've implemented some tweaks to make the game "less frustrating" that result in Elo Hell being a totally plausible thing. Now, do I think that is the case? No. However, does it have precedent in other games and is technically possible? Yes.
5. Last one I promise. So, from my experience, most people who get deep into intellectualizing "Elo Hell" eventually reach a singular argument as their endpoint: There statistically almost certainly exists at least one person stuck in "Elo Hell" to some degree. What do I mean by that. Well, let's say we're playing a game where we flip coins. On heads, we gain $10, and on tails, we lose $9. Statistically, we would expect to gain money playing this game, and most likely that's what'll happen! However, if we get a couple million people to play this game, it's highly likely that at least a couple people are going to be deep in the hole, and some very deep. Theoretically, it's entirely possible that, in a game with random attributes, somebody is getting murdered by lady luck. Elo Hell is a good cope because you can almost guarantee that somebody exists there, and so you can just say you're that person. Now, the chances of that being the case go down pretty quick as you play more games and as the actual skill gap increases, but they're never quite zero. Of course, there are a number of things you can do to get yourself out of Elo Hell even if that statistical version of it does exist, but explaining that to a person with a defeatist mental attitude is quite difficult I've found, and I'm fairly certain that there's a strong correlation between people with a defeatist mental attitude and people who believe in Elo Hell. (Looks like you basically said this last sentence yourself in the vid, good to see I'm not hallucinating!)
Thanks for reading my horribly rambly essay, and have an amazing day!
Well said. I had a conversation on an Overwatch stream today about how Elo hell isn't really a place, but a timeframe. If you are truly below your actual rank, if you play enough games you'll get out of Elo hell. But it's definitely possible to get an incredibly bad string of games and get stuck below your rank. There was a season where I dropped from diamond 1 to plat 2 and got stuck there because of an incredibly unlucky bout of leavers and throwers, but the very next season I was able to peak masters 5 after the matchmaker started being nice again. And while I cannot say I'm a perfect player and didn't lose any games that I could have won if I had performed better, it's incredibly demotivating to keep playing when you get an unwinnable game after an unwinnable game in a row.
I read every word! Thanks for your brilliant insight 👍
I like the concept of following someone because of their personality/interesting insights even if you don't play the game. It's the same for me with WarOwl, even though I've not played Counter-Strike in more than a decade, save for some nostalgic days here and there.
29:58
Now I've become ELO hell, the destroyer of copes.
While I agree that using elo hell as an excuse for a person not ranking up over time is an absolute cope, I actually think that people in lower ranks DO experience a form of elo hell. Its just that elo hell doesn't drag your rank down, it instead takes away some of your agency in affecting the outcome of matches. What do I mean by that? Well I’d like to compare my experience in two competitive seasons, one where I went from low silver to mid gold, and another where I went from low/mid gold to low diamond.
For both these seasons I started at a mid gold skill level. I know this because I had been in mid gold for like the past 7 seasons of competitive, and I hadn't really been trying to improve until that climb to diamond.
In my climb back to gold, it felt like many of my matches were decided before they even started. Despite my gold level play, I lost a ton of matches due to teammates that seemed to lack a basic understanding of the game. I also of course won matches that were steam rolls against those types of players. But generally, win or lose, it felt like I had less control over the outcome of my matches, my level of play didn't really matter if it was gonna be a steamroll one way or the other dependent on who got those 5/20 players (to use your dice roll analogy). Matches that were incredibly close and both teams felt mostly even, ones where I felt my impact was expressed in the outcome, felt few and far between.
My climb to diamond on the other hand, was a much more fun experience. While actually more difficult personally, as I actually had to improve my play to earn the higher ranks instead of just climbing back to where I belonged, I felt like I had more impact in every match. Even when I had teammates that were like a 5/20, in platinum they still had a baseline competence that meant it was easier to play around their shortcomings. Those close matches, where the teams were largely even in skill, felt like they were happening far more often.
In my experience, elo hell does exist in the sense that the variance in skill level at lower ranks has a much larger impact on the outcome of matches, with many resulting in one sided experiences, whereas in higher ranks the baseline competence of players ensures most matches are at least somewhat close.
You will still drift towards your correct rank no matter where you are on the ladder, but I think that the removal of agency in lower ranks results in a more frustrating climb, and makes it easier to fall back on the elo hell cope. It feels like your rank is being dragged around at the whim of the matchmaking system rather than your actual level of play.
All of this is of course subjective experience, hence why I overwhelmingly use "felt" to describe stuff, so I accept that is could be a bunch of baloney, but that's my two cents.
I can understand that. Sometimes the disparity of both teams seems so big, that one or another just get rolled. Some wins and loses just feel free.
I've won so many games being absolutely bad, and also lost a little doing the most as I could.
Now that I'm on plat 3, this seems to be more rare to happen.
something I've noticed about my friends in gold and silver a bunch is that ones who dont play comp sessions over a hourish in length or play in anyway consistently, believe they are stuck in elo hell alot more than the ones who play regularly. ive even vod reviewed their games, and i cant deny the game session they do have when they hop on are very streaky. it leads to nights where they stop playing after 4 matches as they didnt win a single one and a player a rank higher would not have made a difference. and other nights they play and they win 80% of thier matches mostly due to the other team just being substanically worse.
Definitely agree on some things. i think bronze is the hardest rank to get out of and i WILL die on this hill. i’m high plat/low dia now but i was hardstuck bronze for like 2-3 seasons in ow1 and man it was so tough to climb. nobody knows what theyre supposed to do down there
@@llll13lllllI have the exact same experience in top 500
I have been hard stuck silver for many seasons now and it feels like every time I’m 1-2 wins away from gold, I go on a losing streak and end up right back where I started. It’s frustrating because I see things like tanks repeatedly running in to 1v5 matches and dying the complaining they didn’t get healed to me as a support having higher dps than one of my dps while still maintaining heals on my teammates yet I somehow I am solely responsible for my rank being what it is.
I think actually elo hell does exist, just not in a way that you think. You think elo helll is just someone complaining about them being hard stuck in a rank that they deserve, but think that they don't. When in reality, elo hell is the 51/50. I like to call it where if you win 51% of your games. You will climb, but you will climb so slowly that it will feel like you're stuck in the rank because you will win 5 game, Lose 4 games, win 5 games, lose 4 games. And that is what elo really is. Now, if someone doesn't play enough games to see their rank eventually slowly climb and go up, it will Feel like they're Stuck at a certain rank. So instead of it taking 10 games to get to their True rank of maybe plat 5. It takes 500 games to go from silver to Plat 5 because of all the winning and losing, even though technically they deserve plat 5, but because of the 51/50, they climb so slow. It feels like their hard stuck in silver/gold. and that is real elo hell, and it does exist. imagine someone who doesn't have the tenacity to see their games through, tilts and gets angry ,upset. Then what happens? Is they end up throwing games, which now even lowers thier rank even more, and so even though they could be placed right into a plat game and hold their own. And in fact, even win a lot of games. But because of them Not having the tenacity to see their games through because of how frustrating it is to win 5 loose 4. They just don't ever see their true rank. And that is real Elo hell
Yes elo hell is when someone is focused on their rank instead of actually trying to get better. That's not elo hell though, it's the hell of not being able to hold oneself accountable and that hell is almost never isolated to a game.
This is a deeper personality problem that people write thesis about and it will forever fascinate me
elo hell was truly truly real the first 8 seasons. I spent a few hundred hours to figure this out (doing multiple unranked to GMs, bronze to GM, climbing on my main, etc.) and it's because it was nearly impossible to actually rank up. It took me like 50 straight wins to get out of bronze which is frankly ridiculous. I'm just hoping season 9 fixes this issue now that there's more direct feedback.
Your point about players at the bottom being incentivized to tilt is what I would define as elo hell. Not saying its the games fault that they’re at the bottom but the feeling of hopelessness when you win more games but don’t see any visual progress due to the huge gap in bronze, makes it feel like hell. So I believe in the feeling of elo hell but not the theory of elo hell because I remember what it was like to start at Bronze 5 and see no indication of climbing after hours of grinding. Took a while but I forced my way out lol
I was kind of on the fence about this at the beginning of the video. I don't play comp but from an outsider's perspective, the idea of elo hell seemed plausible. Your explanation is excellent. You're right, it just took a little bit of patience to humor it for long enough for you to change my mind.
Good video, man. Good video.
ermmm akshully elo hell is real because lifeweaver isnt my boyfriend irl
21:39 i'm exactly the type of person you're talking about here. i was DEEP in the bronze 5 SR hole, maybe because of my performance in ow1 or something, hard to know. it took me almost a year to get out of bronze 5 but once i got out of bronze 5 i was able to hit silver 5 in a matter of just a month or two. that massive SR pit at the bottom of bronze 5 is the closest thing to elo hell that i believe in, only because when the SR stat is hidden it can feel like you're making zero progress no matter how much you're actually improving.
i think you hit the nail on the head about this topic in the "what do" section. it's really hard to accept you're bad at something you like doing and want to be able to be proud of. the self esteem hit from this realization has definitely, at times, made me feel almost as much lack of control over my ability as the elo hell mindset brought in the first place. but it's possible to overcome that and being able to see the light at the end of the long *long* bronze tunnel definitely helped
My biggest problem with my own personal "Elo hell" was just my pure, unadulterated, HORRIBLE luck. More than 70% of my games had literal trolls (Throwing themselves off cliffs, running around being unhelpful hiding in corners (Note, i'm not talking about flankers or sombra here. Just literally people squirreled off doing jack all.), or verbally abusing everyone on the team. (Often more than one of these.) After a while, I gave up even bothering TRYING to rank up. Then I ended up quitting shortly after 2 came out. I probably was somewhat close to my actual rank, but I definitely knew I could play better than I did when playing with randoms in queue, because I did play with higher ranked players in various tournaments and I invariably performed much better with teammates of a higher rank who understood strategy and comms. It's a team game. If the team can't team, then the game don't game. Like I said, I gave up. It wasn't worth the frustration anymore.
This will always exist as an outlier type deal, statistical improbabilities that disprove the existence of certain phenomena doesn't mean it's not possible for that phenomena to manifest in certain cases for certain periods. I also have some pretty abysmal luck, such that it does actively play a part in slowing my rank climbing in a noticeable manner. However, after a certain point it still does boil down to a skill issue. That's why I'll generally believe someone that says they're, say, competent enough to compete well in low diamond but are stuck in midplat because of throwers and lack of coordination, but they're almost certainly doing something wrong if they claim to be masters or something. Over time, it is almost certain that you will skew towards your real rank, and it just isn't feasible to say that your skill isn't holding you back. Either you aren't good enough to play consistently and don't deserve that rank, or you're good enough but unlucky so you're slightly below where you should be, thus making it not that big of an issue.
With that said, there is one nuance to this in that you could be in a situation of the likes of being GM of one specific hero but don't perform nearly as well in any of the other heroes, so you end up stagnating as a result. Or, more commonly, not being able to play hitscan well or something. That is most certainly still a skill issue, but a more understandable one since nobody wants to actively play unfun heroes just to gain rank. Still though, in terms of ranking for the role, that's definitely a massive skill issue.
@@DragonEdge10 yup you're spot on with this one. Given enough time and a decent amount of games, everyone is at their intended rank within a relatively small margin of error. The outliers can certainly be a few divisions off, but by no means are there masters players stuck in plat for long.
What's interesting is that it wasn't anywhere near 70% but you have slowly increased that number over time in your head
Fighting every fiber of my being to not spam this in my friends DMs
Ever since OW2, Blizzard has added an insidiuous element to ranked: grading on a curve. They're not hiding your sr, they've genuinely stopped tracking it. They only measure your rank relative to the position of other people. This means that if you go 5 wins in a row, depending on how long it took you to do so (say you play 1 game a day) you could actually *drop* in rank. It's why I quit.
Edit: I forgot to mention why I thought this was relevant: believing in elo hell is a sign of tilt, and either way you're screwed here: you either doomqueue and end up losing a bunch of games due to losing focus/tilt, or you try and take it slow and limit your games per day and get kneecapped by the system. I got lucky and caught it red-handed, but the diminished returns from not playing often enough could mean the difference between going positive or negative in a 5-3 cycle, for example. The player would never know they're being screwed over, only knowing they dropped in rank and probably blaming it on elo hell.
Like, I genuinely cannot see how my own biases can in any way influence the fact I went 5 wins 0 losses and went from top 28% to top 35%. That's just a drop. Either my theory is true, that I'm being punished for not playing often enough, which honestly; what the fuck Blizzard, or I'm being made to play matches that have only decreased my rank despite me winning, and then why would I play?
Dude, thank you so much for this video. I’ve genuinely thought ive been stuck in elo hell, but this video kinda made me take an objective look at things, and ive noticed that ive been coping hard.
That’s something I’m so thankful for: the homies will just tell each other they messed up and need to lock in
nah frr like i’ll tell my hgs when they needa lock in n vise versa 😭 like we not gon sit here n pretend we’re good
Honestly, the best counter-argument to the concept of Elo Hell I've heard comes from Flats: the enemy team is just as likely to have a bad player as your team. If you can guarantee for yourself that you are a good player, your team will have 4 potential bad players. The enemy will have 5. Even with 'elo hell', the odds are in your favor if you really are a competent player.
Edit: oh you literally just mentioned that at 12 minutes haha
I don't play anymore, and I accept now that the average player's skill increases at a rate faster than my own, so I accept whatever rank is given to me.
I used to cope hard in OW1. I could peak plat, but my average match would be silver, and I was left holding the bag. And, unfortunately, the first 2-3 seasons of OW2 were the devs saying "oh we oopsie'd the ranked system on our end", which really made it easier to cope. I now believe that any...frustrations, or beliefs, that the system is rigged against you might still exist, with varying levels of merit, but it is not ELO specifically causing the problem. Could be a bad patch, could be time of day, region (look at how OCE/AUS/NZ players talk about their ranked experiences, for example); could be a myriad number of things. And, I feel like for a lot of people, especially lately, there's more underlying, negative factors, that come together consistently enough to give off the impression it's an ELO thing; which is just not the case
as someone who personally took 100 wins at a 65% wr to get from b5 to b4 it feels like your going in circles for a long time im gold now but that frustration i wont forget and i agree with how dumb the lack of transparency is
Yea the problem with b5 is that it is the size of an entire full rank
I would like to propose an anecdote and ask for an explanation:
My friends and I started the game around gold. Left and came back years later and placed immediately into diamond and stayed there. Following season we got bored and threw back to gold (no we don’t throw anymore). Following seasons couldn’t climb out of gold at all. Recently came back after they reset the mmr and am back to high diamond/masters, but literally couldn’t climb out of gold before the reset.
nah brother, i didnt believe in elo hell til i got a game where both my dps had literally 0 kills. Not a single one. tank is trying to create space and supports were on par. Get those shitters out my game bruh
What's crazy is that on average, you're just as bad as they are over time. If you can't understand that, you don't deserve to climb
Sure, but is every game like that? Probably not
I started playing Overwatch again recently and that led me to your channel and I fell in LOVE. Your content is great! I love the way you meld Overwatch and more deep topics together, keep up the good work!!!!!!
I've liked a saying Eskay said on one of her Lucio guides. about 30% of games you can't win, 30% you can't lose, and 40% your individual skill can decide the difference
The problem with the ELO system is that it isnt well designed for a hero-shooter like OW but for games like Halo where everyone has the same kit. Overwatch clearly has hard counters; the bronze moira will shutdown the Gold Genji with ease but still can be less skilled overall. An amazing mercy isnt going to win against a winston diving her. The widow with a mercy pocket and Shield tank will always come out on top against the better widow if they dont have the same mirror composition. That right there makes ranks inconsistent and not reflective of anyones skill.
In ranked they expect you to win games not taking into account which heroes your teammates are picking and as a DPS player it is near impossible to carry bad hero picks especially when it's your tanks picking wrong and also positiioning poorly. I woulnd't mind the grind if they made the game less team-work dependedent and give more agency to individual skill. The game is so counter-pick meta that how can any system guage skill by just looking at wins? I've won games without breaking a sweat just because my HERO countered theirs, not because I out-shot/out-smarted them....I simply outpicked them. Going Sombra against a widow isn't exactly a brilliant play just the right one. On the other hand I've busted my ass on 76 or Ashe and dropped 60 elims just to lose because Tank decided to chase and die while on onvertime and they rushed us.
Another whole issue with ranked is that noone is actually their rank, people have smurf accounts, this makes each game inconsitent among skill levels. SO many smurf accounts I know can barely climb out out plat on one account but hit GM with ease on another....that right there is all you need to realize the system is quite all over the place and not a reliable source of determining skill.
Just got out of a Splatoon tournament and I’m feeling the whole high to low energy drain so I’m so excited to watch this. The Dunning Krueger video was why I subscribed so it’s nice to see this topic again
splatoon on top
Your scripts are so good. Like seriously they always strike the heart and mind perfectly. It's crazy, keep it up.
Thank you so much! Kind of a fun fact about this script, but I actually wrote this video and the upcoming one out of order - the members’ choice poll for this month was a dead tie right up until the end so I didn’t start writing this script until the 11th, and while I was waiting for the members to finish deciding I ended up writing a whole other video that’ll be coming out on March 1st. I had to burn the candle at both ends a little bit to get this one out on time, so I’m super glad that it still worked 🥰
10/10 no notes. A great deconstruction of "Elo Hell" and what you can do to change that mindset. I make this comment longer for the algorithm because I have nothing more to add to the conversation than what you have already said.
my "elo hell" is just my wifi randomly disconnecting
So if elo hell doesn’t truly exist, does that mean boosted ranks also doesn’t exist as well?
boosted deff exists
lol. i don't think that's how it works. if you play with someone who has a much better skill rating than you, you will likely start to win more games. it's once you stop playing with that duo, it becomes clear to everyone in the rank: you are boosted and you don't deserve that rank.
elo hell, to me at least, is relying on the idea of "these events are out of my control," which is not even true. and being boosted is having your duo, imo, equalize your mistakes or making more positives plays (has a better understanding of how to control things ig?), and possibly having a better understanding of the mistakes they are making (in terms of, positioning, etc).
as an example of elo hell, i lost some of my placements yesterday, i could whine about how it's my team mistakes/elo hell, but i also could've joined vc and called out when i was going to ult or when i saw a flanker. and an example of being boosted, is when i duo with someone who has a higher skill rating (or smurfing) than me, this would be extremely unfair to the enemy team as i already have an advantage (assuming they aren't being boosted).
basically, these are two different things. being boosted does exist lol
@@kitenlovar699 people also miss that the higher ranked player in the duo is also boosted because they are always playing in lobbies at ranks below their ability level.
Those 2 things are not inverses of each other so your question doesn't really make sense
In groups? Yeah, of course. Solo q but just lucky? Over time it will get to their actual rank. I don't see the problem here
Elo Hell exists, that is to say in the mid ranks some of your games will be effected by other players fluctuating in their rank up and down and that will mean you will lose some games irrespective of your performance.. that happens in every rank but it’s more likely in a mid-rank like Plat.
You also don’t factor in smurf accounts and grouped up players who are synergising certain heroes and utilising one of them being lower ranked to skew the matchmaker in their favor especially against solo queue players that cannot go in with a team strategy and coordination baked-in. then you have players that have been boosted, sometimes because they grouped-up and won games based off those advantages that they now no longer have, all these factors contribute to the sluggish feeling of what most people would call Elo Hell, it’s not fair to just say there is no validity in these issues that really do effect the game and the ranking system as just people trying to cope
You could make the argument that this is all more of an Elo Limbo, you will overall break through the effect with enough games played, if you are a masters-grandmasters you’ll probably completely blast through it unless you are using a highly team reliant hero like lifeweaver, who has great healing yes, but can’t dictate the outcome of a game in the same way, say. an ana can, and yes some heroes are better at climbing than others given the skill tier and that heroes specific balance state.
However, there is a more stark version of Elo Hell in Mid-ranks on console, why? because there are a lot of smurf/alt-account kbm xim players on console that will be in your games, they thusly have an artificial advantage in mechanical aim that drastically effects their team and your team in terms of this particular “skill” metric, and may also come attached with a pocket mercy to amplify the advantage of this cheating behaviour, but they often times won’t have comparable gamesense and other skills for the given rank or they would have risen from the rank they are in, so they can stomp some games off the back of this advantage but equally lose games when their advantage is minimised/countered and their pocket mercy isn’t there to bail them out or amplify this factor.. this creates an environment in which games are sometimes won and lost at the matchmaking level because the system wasn’t designed to factor in these ximming wild cards, it’s most noticeable in High Plat/Low Diamond on console, and varies with the region you play in, but effects console across the board across all ranks, by time you get to GM almost everyone is using Xim and it’s really not fair to say they are more skilled than people using controller when they are cheating.
No because all of this is a constant everyone faces therefore it isn't effecting your rank at all really
Absolute amazing video definitely needed to watch it! Made me realise a lot. Thank you!
Love the vid Viv. Theoreticals ahead:
Early in OW1, you had 1 competetive score. Shared between all 4 types of role, but there was no role lock. So it was perfectly possible to have games where you could preform as a support and just not have the space to fit yourself into the team, causing you to (on average for mono role) lose more games. Being a plat support, but not being able to hit shots from hitscan dps or confident in positional tanking would lead you to being in a gold/silver position due to circumstances barely within your control. I don't think this is called ELO hell but it certainly is frustrating, even during the learning process, as it can often feel like the quickest (not best!!) way to improve is to find players who allow you to focus on strengths rather than shore up weaknesses. I do wonder if ELO hell is more of a cope in games without role lock due to this extra step in the system to blame.
As an addition to the above, I'm not sure on the maths on this, would role influence allow you to push faster rank changes as a result? Like, good tanks (for eg) can carry a game, and you are a good one, vs support having little effect and you are a good one; would the tank player in such a situation have an advantage for reaching their "true" elo? All things equal, they'd have to play less games surely? or do the mistakes of a good support in a losing match having less penalty keep things balanced?
I feel like I'm stuck in Elo hell, and nothing can convince me otherwise. Here's why: I have a friend in real life who's not very good at the game, yet he only plays ranked and expects me to join him. He's in Silver, whereas I'm in Plat, but when I play with him, I end up dropping to Gold because of his performance. This limits my ability to climb higher than Plat since I don't play enough solo games. However, I recently started playing on an alternate account and got stuck in Silver on it, while on my main account, I consistently fluctuate between Gold 1 and Plat 1 because of my friend dragging me down. My plan now is to try playing with him on my alt account to see if I can break through the Plat barrier.
I think this video is a cope too.
Sorry I watched only till 19:00 but that's the part I want to discuss.
Your is a very well rationalized projection is something you know partially (as you stated). The details on how the ELO system is implemented in OW2 are not public.
What's public is the Blizzard Match Making Rating system patent. In few words, this system is built to assure a fast and efficient match. Now, fast is paramount, and efficient has many meanings. I don't want to discuss this because the info is publicly available, you can see what they see as efficient.
The problem with this video is that you analyse the system as something very abstract. The player you describe is a type of player, the type that stuck in bronze or silver, maybe gold. Now, players can interact in many ways with the system. You failed to mention the groups. Groups (duos, trios and 5 stacks) exist and have a big influence on the game.
Within the grouping, one can do many things. One can "boost", i.e. have good group mates, or "min-maxing", i.e. having group mates worse than you.
These tricks are well documented and you can experience it very easly in a game session.
Besides the tricks, there's the smurfing too. If the anti-smurf system catches you, it boosts you. This is why every content creator can make an "unranked to GM".
I invite you to search on YT for a "bronze to GM". For what I know, there's only one guy who made it and it's a very interest document to demonstrate how bad the system is.
You cannot get better in this system.
Not because the system is that bad, but because players abuse the system flaws and the people who can interveen don't really know what to do (hence, this season reset and changes).
Also, what about the floating players? (unranked players placed in ranked matches for placements)
I have a lot to say but I stop there. I just don't understand why the community cannot come together on these issue. All other games can pint-point the flaws of their games, this one does not.
I really can't see why.
Very good video BTW, you put a lot of effort and knowledge in your videos and the results are great. Congrats
I am *begging* you guys to actually finish watching the video before popping off in the comments, jfc.
@@TheViveros I finished the video and I have nothing to retract or change. Just thanks for mentioning that climbing low ranks is systemically harder than climbing mid and high ranks. But honestly, the psychological analysis is likewise partial and subjective, but I'm not interested in arguing that, it's your view, very articulate, but partial.
Anyway, I would have liked from you an answer on the points I raised, I think those are the most important.
Alright, fine, let's deconstruct it point by point.
1) This video is not a cope, because frankly, I have no idea what you think it would even be coping for. I don't think you're using the word right.
2) If I only partially know this after all my research, then you know even less given the lack of research demonstrated by the lack of understanding of the system even after having it explained, so acting like you have the capacity to act as an authority and yet I apparently don't have the capacity to opine on it in a way that could be correct is just acting in bad faith.
3) Every other player is also operating under the matchmaking system. You benefit from the system as much as you are hurt by it because it's random chance that evens out in the end. It's a wash.
4) Rank is inherently, fundamentally abstract. The whole concept of Elo is based on the principle understanding of skill as something nebulous that cannot be objectively defined, and the ranking system is built around trying to evaluate it under that pretense. This was mentioned in Part 1.
5) The matchmaking system deliberately tries to match groups with other groups; if the other team has a duo, your team also likely has a duo. If you're five-stacking, the other team is also usually going to have a four- or five-stack. Grouping up does have an effect, and that's why the matchmaking system explicitly tries to balance that out by matching groups with other groups.
6) Yes, boosting exists. Please refer back to the section of the video where I said "You can win for reasons outside of your control as much as you can lose for reasons outside of your control". The ranked system is still working as intended because all it can do is try to evaluate how a team with that rank will perform against another team. I'm terribly sorry, but there is no way to devise a ranking system that can peer into your soul and discern your exact skill. This is the best we've got, and at the end of the day, you're just as likely to be playing against a boosted player who's worse than you or playing with someone who's trying to boost their friend as you are to be playing against them.
7) You are describing the anti-smurfing system as if it's disproving the validity of the system when in reality it's demonstrating that it works. A smurf's rank gets dragged upwards if the system can identify that they're better than their rank currently indicates, which proves that the system *is* capable of recognizing when someone is clearly better than their rank. Beyond that, if you're
8) Yeah, Bronze to GM is rare because going from being among the bottom 5% of players to among the top 1% is extremely uncommon. Most people don't go from being objectively terrible to objectively amazing at a single game unless they didn't belong in the bottom 5% in the first place. If you earned that Bronze rank, then you probably will never be GM unless you put in a ton of work. It isn't a demonstration of the system failing to say "Players who are very good can start a fresh account and hit GM fast, but players who are very bad rarely - if ever - hit GM". That doesn't mean Bronze players can't climb, it just means that they would have to do a lot more work to climb to GM than someone who's starting out as pretty average. The fundamental difference in ability between a Bronze and GM player is so massive that a Bronze player will likely never become GM for the same reason a GM player will likely never become a Bronze player. That's not a flaw of the system.
9) You absolutely can get better at this game. My first ever competitive season, I placed at exactly 2000 SR back in early Overwatch 1, and a few seasons later I was in Masters. How? I played a ton, reviewed my vods, and figured out how to get better. Then, slowly, my rank fell. Why? Because I got busy with life and uni and wasn't playing as much, was falling behind the patch notes, and got worse at the game. Anyone can get better, and it's a cope to pretend otherwise. It's not the system's fault if people aren't good and refuse to try to get better.
10) Floating players aren't players that the game has no understanding of. There's a reason you have to play a bunch of QP before you can even try ranked, and it's partially so the game can get a feel for where it should start you out in your placements. It's not flying blind, it's just operating with less data and will correct sharply if it's wrong.
11) The community can, and has, come together on this. Just because some people are still clinging onto the cope for dear life doesn't make them correct; Elo hell is not real, it is a cope, and the sooner people accept that the sooner things can actually move on.
Almost all of these arguments are points made in the video that were seemingly deliberately ignored. I have gotten probably a dozen essays like this in response to this video, and I cannot stress enough that this is *just* you coping.
@@TheViveros This is a very good answer and I thank you for that.
I'm sorry I didn't catch these arguments in the video, but it seems to me they weren't as clear as other points you made.
I have to say that your response about going from Bronze to GM while correct shows exactly what people refer to, incorrectly, by saying "Elo hell". That's not just a coping mechanism, but the right reaction to a big flaw in the system. Why the same person, the same player, has to win more than 200 games to get from bronze to GM and only 5 if the account is boosted? (Boosted from QP, as yourself admitted). As you said, it's the anti-smurf, it's how it works, but then why say that the last word on your progression is the rank and the rest is coping?
The fact that the system works sometimes or even many times doesn't mean the system is always right and beyond questioning.
I would say there are so many factors at play outside of the single player skillset that one cannot but persevere and hope, if one want to climb. I really don't like this emphasis on the individual and its performance. I don't like it in social sciences, I don't like it in a team based videogames. (I say “like”, I have arguments, but that’s off topic).
There are ways to consider the single player skill in a team-based game, but if the system takes in account only wins and losses (or predominantly only win and losses), i.e. a volatile outcome, that’s an arbitrary set of data where the individual performance is very diluted.
Also, all this argument on the "grind mind-set" vacillates a bit if you consider that the fastest way to climb if one's hard stuck is getting a fresh account.
Or, as I said before, by grouping (example, "min-maxing").
Anyway, I don't want to force you to answer me on every point, I just wanted to argue that things are not as righteously monolithic, nor easily scientifically analysable (at least with the current set of information we dispose as public).
All social and psychological factors implied in every human skill progression are at play in OW, as for every game. There are, on the other hand, transparent games (like chess or, in some measure, sports) and others more opaque. OW2 clearly falls in the 2nd category.
That's what I meant by saying "this video is a cope": things are not as clear and determined as you say and your explanation is not all-encompassing. Sorry if it seemed otherwise.
When I came back to Overwatch 1, I got put at about 680 SR. Near the bottom of bronze. For awhile I believed that I was trapped in ELO hell at the time because I had been low silver in 2016.
It was truly a cope and I realized I didn’t know shit about the game and couldn’t aim. So I poured hours and hours into improving. I remember being at about 950 SR playing Ana and it was a game on Temple of Anubis. I literally killed 4 every fight that match and our team’s Pharah came into voice telling us to thank us for the carry and my Rein jumped in and was like “it’s pretty clear our Ana is by far the best player in the lobby.”
It was at that point that I knew I was improving and within a month or two I hit plat for the first time. At every new level there was a new set of skills, knowledge, or mechanic I needed to learn to improve and I would once again get that Anubis feeling of dominating.
Eventually I hit diamond and then switched to Valorant, where I once again went from bronze to diamond. Now I’m playing Apex, and am on my way to diamond again.
Improving is a process and your rank is the feedback as to whether you’re improving faster than the general player base. If ELO hell exists how am I now in my third climb in a third different competitive game?
Just as I was wondering the same question LMAOO good video as always!
I cant agree that its just cope.
But the reality is that you do just have to get better.
The reality is that to confidently get past the limits of "Elo Hell" it requires that you be the X-Factor.
You cant afford to lose the duels.
You cant afford to NEED peel from your teammates.
You cant afford to be out of position too often.
You cant afford to rely on your teammates to pick off the Mercy who has rezzed free behind a wall everytime.
You HAVE to be the teams X-Factor if YOU want to climb.
Isn't this just repeating what he said??
Two things:
1. I've been looking forward to this video pretty much since it was first introduced in the polls, and it did not disappoint!
2. That Widow gameplay 😳
Honestly I'm shocked that it took this long, I really thought that this topic would win in its first poll. And you have no idea how happy I was while I was playing that Widow game knowing that I'd be able to actually include some good gameplay in this video lmao
I believe a concept similar to Elo Hell exists, though it's not just constant losses or anything. My original account, before OW2's multiple changes to the rank system, was hard-stuck Bronze (1200 SR) on Tank (Tank being my main role). Getting down that low was 100% my fault; I entered ranked as a new player, inexperienced and not very knowledgeable. But as I improved, I didn't see any improvement with my rank. Every game was frustrating, win or lose. I felt like I had to hard carry every single game.
I joined a PUGs group of variable skill levels around that time, and suddenly that issue disappeared. My performance outside of what most would think of a bronze player in those PUGs made me think. I made a new account, leveled it up, and played my placement games. That new account placed Gold (2100 SR) on tank and until OW2 it fluctuated between high Silver (1900ish) and low Gold (2100ish) every couple of seasons. While I don't doubt that I may have been able to get to that rank on my original account eventually, the dedication and continued frustration that would have been required made it effectively Elo Hell.
I sat down and did tons of games to verify that what you're describing exists. It took me going like 50-4 to get out of bronze as a GM player.
Your experience with the ranked system is normal. It's common to have a peak rank and fall down and fall down a rank or two and struggle to carry the games on your own at times, you will get back to your average rank eventually, the only way to rank up is to be the best player in the lobby.
Also, new accounts start around gold you just can't see the exact elo before placements (used to be plat in ow1 they changed it sometime early on in ow2). So that could have been what got you out of bronze without having to play games in it.
Yeah the problem isn't that people don't eventually reach their true rank, it's that "eventually" part. It's monkeys and typewriters. Sure EVENTUALLY they'll write shakespeare but I wouldn't blame someone who got bored after a thousand years of waiting and nothing remotely interesting happening. The same concept is true in overwatch. While yeah, given a thousand games the system has enough data to pretty comfortably say you're in the right rank, those 1000 games are incredibly frustrating to have to play through when 499 of them were losses you felt you had no hand in and yeah you technically won more than you lost but not enough to where you felt that 1000 could've been reduced down to 100.
In short: the problem isn't that the system doesn't work. It's that it works slowly, and humans don't have the patience of a machine. Elo hell feels real even if it objectively isn't because just like Pi could have a one million long stretch of just 0s, so true could an elo system have a stretch of 1000 bad losses and/or frustratingly difficult and unsatisfying wins before you reach your "correct" rank.
i refused to give up my ow1 cosmetics and it took me over a year to get from the pits of bronze 5 to silver (which i finally hit 2 days ago). worth it tbh, it's been a very gratifying anime underdog story
8:56 has me in tears oh my god bro
yoooo the slide lmao
I didn’t even have time to consider trying to life grip him, that man just disappeared straight into the river.
@@TheViveros you replied agh 😭 I love your videos they’re amazing
SAME OMG
it boils down to this:
game lost: Could have done better.
game won: Nice, I can do better.
game lost because people are throwing: It is what it is. next game
Honestly, if you take this mentality into your games you will immediately perform better. I think it's an undervalued lesson that lots of people learn from youth sports, where win or lose the game is over when the final buzzer rings and you've gotta move on. Someone might've been an asshole or pulled some dirty moves on you during the game, but you can't throw hands in the parking lot over *everything* - you've gotta just shake their hand, say "Good game" and move onto the next one. If you sit and stew over every game that doesn't go your way, you'll become bitter, and when you're bitter you can't get better.
I started overwatch when overwatch 2 launched, got myself in bronze 5 dps (genji one trick mostly) I really believed in elo hell because I went from the whole of season 1 and 2 with a net total of 5 blade kills. Season 3 came and I was having 10 blade kill games and I’m beginning to top my lobby in kills and I’m somehow only bronze 4. I hated the system because I’d queue with friends and they’d go up from silver 4 to silver 1 and I’d go up from bronze 5 to 4 which reinforced my belief in elo hell. This video helped explain why I felt like bronze felt like a never ending pit. Thank you.
I explicitly do not play Overwatch, and I have fuck all knowledge about how rankings work in general, but this mf has somehow roped me into several of his video essays about the damn thing
That is to say, you’re doing an amazing job, and I hope you keep making content because it's genuinely amazing and well thought out
YEEEE I love filling out surveys for a good cause 😊
Wow, great video. I truly started to enjoy overwatch so much more after accepting that I'm just not as good as I thought. Everything you've said I've already accepted and knew years ago, but you just put it perfectly.
Also, you keep mentioning bronze players but just from my own experience, diamond/masters/gm players need to watch/understand this video just as much if not more, since they have a general understanding of the game and are decent enough to have a massive ego but also not top 1% -> 0.1% like they believe they are. Diamond/masters is rotten with this "elo hell" ideal. I've also always been a masters player barely peaking into GM for literally the last 6 years... so i have little experience at the lower ranks but this 100% applies to higher ranked players just as much as bronze (if not more).
23:07 ayo what he doin?
Shocked that this is the first comment I’ve gotten about it tbh
Given the new information on the matchmaking algorithm I wonder how it affects this approach to elo. During ml7’s dev interview the dev mentioned something about a machine learning algorithm and boosting mercy players with lower healing since high elo mercy’s have lower healing since they are boosting more but the model picked up on just the lower healing and might’ve thought it made a better mercy. Anyways, nice vid!
Really I’d just love to know what data and formats the model is fed, like I didn’t realize that it does or at one point could see some form of healing stat.
The only thing I would like to add is that elo like most evaluation works better with more data to pull on. I mean the rank of people who play more regularly would probably be closer to their "correct" rank since their placement is less influenced by outliers. Naturally this works both ways if I play 2 matches a week or something games I won or lost simply because someone left the match are more influencial than these types of games for people that play 5 games a day simply because of the sample size the ranking system can pull from.
I hope this makes sense
One question I have. If the system is using all the math to determine which team is likely to win, what are they basing that likelihood on? If it is based on performances from previous games and your current modifier status (positive or negative), how does the system track what is going on in the game? If you are on an upward trajectory then you are probably maining a character and farming certain stats and plays. What happens when 10 people constantly flick between heroes to counter others despite not being particularly good with those characters? Surely if you want to climb then you must stick with a hero you are good at and play that character consistently to avoid swings in your calculated team value which could negatively impact your next game. It makes sense that the system is more stable with consistent information.
Babe wake up viveros dropped a new video
perdón por el comentario en otro idioma pero una vez literalmente le dije al otro support "deja de llorar y empeza a curar" porque se estaba quejando del tanque pensando que era el problema, cuando el problema es que tenia menos de la mitad de curación que el resto de suports en la partida, cuestión escucho y ganamos tan fuerte que para el final de la partida solo quedaba un support en el equipo enemigo (le dimos la jugada de la partida)
The problem with Elo in pvp team based systems in games isn't that it takes i to account everyone's score to score you. It is that it doesn't correctly evaluate individual impact. I have lost and won many games. And a lot of wins and losses aren't too off depending on if you actually work together with your team. But I have 100% been in games where I far overperformed and still lost because my teammates sinply didn't get on the point and I lost rank because of it. Many people deal with this issue. The system needs to not only evaluate a number over the number but also how well everyone performs individually as in who they took a pick on (how important their kills were) l, when they were on the objective (how important was it to be on the objective if they were on it alone), if they tried to keep someone alive as a support, how much they supported a teammate being attacked, and if their deaths were the fault of them not being with their team or their team all dying and leaving them alone.
Many factors play into winning and losing, not just numbers, but actual gamesense. And a system based on numbers could never accurately rank someone's individual improvement in that. The only way to get out of elo hell is to find friends who undertsand how each other play to play with and work together. It is a team based game not a solo based game. This is why making "stacking" not be allowed in high rank is an ignorant and foolish way of working with the game. If the best players are also friends and play together then they are the best because they are good at doing the one thing the game is about, TEAMWORK. Without trying to assist your team in a team based game you will never win. If we could see our number and the game would follow actual individual gameplay properly and tell you what you can improve on then there would be no reason to "cope" and people who get annoyed would actually have no reason for it. But the game doesn't care about how well you actually do. It just cares about the number and the math. It is just a poorly made system for multiple player systems. Literally just be nice to people and try to find groups you are decent with and play together in a group with them specifically. Everyone plays bad, everyone plays good. I have not found a single person who is 100% consistent. And the math cares TOOO much about cinsistency. A hunt never goes as planned, and a pvp match is just a hunt with more complex problems to deal with.
the problem is even if im the constant variable, i still need a copious amount of games to average out, and who REALLY has time for 100+ overwatch games a season? most people dont, or even have the will
Since beginning of season 9 every 3-4 games i encounter a person playing music in voice chat and then leaving after a minute or two. Outside factors, like trolls and people deranking their accounts for whatever reason bring up lots of questions about the validity of "no such thing as elo hell".
Here is what im talking about, i just played this game 30 minutes ago TR7NKN
The other team gets those people too, just as often as you do.
@@TheViveros Realistically someone is always going to have more of those games.
@@bismuth7730 And realistically you remember more the people who played badly than the people who played well due to a negative bias. I garantee you that it's less than every 4 games for a long time and even if it is it's every 4 games, just play better and if you lose because of these people every 4 games you will rank up and get out of the rank range in which people are always bad.
what is the opposite of elo hell where it feels like you're being matched against people that feel instrumentally more powerful or skilled than you?
Boosted
One factor that I didn’t catch being mentioned was leavers possibly being more prevalent at lower elos. That’s one reason why I favor the idea of a surrender option to just cut the bad matches short:
You shouldn’t have to play out an entire defense rd 3v5 after 2 players leave on your team. Rolls are so frequent they have become normalized. Much of the OW toxicity comes from being trapped in these frustrating, humiliating, helpless feeling situations.
If 3 players on a team elect to surrender, the match ends and the enemy team gets their full pts while the enemy team gets half downgraded what a full loss would have cost [say 15% calibrated down rather than 30%]. Damage mitigated; frustration ended swiftly; easily forgotten.
Bad games ended quicker, good games sooner and longer.
It’s much easier not to tilt and adopt a mindset of: ‘Whatever, GG, Go Next’
If you can actually go next.
If a player leaves the match, that player is blocked from queuing for 24 hours, takes a loss and automatic derank one division penalty.
The abandoned players receive a draw.
Three strikes and your acct is suspended till next szn from comp; your internet cannot handle it if that really is the reason.
This part may just be my anger talking:
Season-Long Suspended leavers must also start next szn in a new small shame rank below bronze: iron. Iron cannot be deranked to, you can only be placed there by leaving suspension. You must gain 300 sr and have an endorsement rating of 3 or better to reach Bronze 5.
At the very least, offer forfeit as an option the enemy can extend and the badly losing opponents accept. Maybe it won’t;t get used often, but the times it will people will be grateful for. Maybe even incentivize it with a 10% sr boost to the winners for offering. Many would take a sure win with a lil extra boost over spawn camping for 5-10 minutes.
Yes people can always make new accounts, but with stiffer penalties how many people would bother?
I agree that if you’re just blaming your teammates and not yourself for being at your rank, that is wrong but I feel like we can all agree that there is that couple of games where it entirely was your teammates, fault hands you were the only person on your team that was positive and order contributing to the overall goal
From this video’s script: “It’s also important to remember that the pendulum swings both ways here. You can absolutely lose an Overwatch game through no fault of your own - but that also inherently means that you can win an Overwatch game through no success of your own. It’s a team game, and the results - barring the rare instances where one person absolutely single-handedly carries through a dominant performance - will always reflect the overall performance of a team, not the individuals. You will lose games you performed well in, and you will win games you performed poorly in. Complaints about Elo hell exclusively focus on the feeling that it’s only ever affecting you and your team and only ever doing so negatively, but statistically it helps as much as it hurts your team and the other team is dealing with the same situation.”
This specific point was addressed - multiple times, in fact - in this video. Maybe watch the whole thing before commenting.
@@TheViveros thank you I completely agree and like you said you can completely lose a game due to no fault of your own but if you’re consistently trying to climb and you can’t, that’s where you belong
@@TheViveros you had nothing but absolutely great points
This video has taught me I am so much worse at video games than I ever could've imagined, lmao!
I think one problem is something you said in a previous video about why Overwatch is a failure as a competitive game. Because of the randomness of hero switching, unselectable match types, and meta changes every quarter season that can occasionally change how a person's main (or strongest set of heroes), no matter how accurate the method of measure is, it's measuring something that isnt a straight line that the players are walking. It's more like taking a leap of faith every match and sometimes you land flat on your face because there was nothing to land on. It doesn't justify it, but hopefully the communication implemented in season 9 makes this easier. I stopped playing competitive because i realized the pressure to do well was causing me to fail. I do excellent in casual because I can concentrate on just getting better, not dragging others down with me.
oh how this video has AGED!
ITS HERE!
omg watching you play widow is so impressive!
also I'm soooo hyped for the new study! the original study was what drew me in to your channel. I loved how much data you had and I really enjoyed seeing it presented, so I look forward to the new study results
not trying to be rude, but i think you should try to talk a little bit slower. It's hard to compreheend it sometimes. Keep up the good work, this is one of the few channels that actualy put interesting content
I think one part of this you didn't mention on the why we cope section but is very important is the amount of effort and mental stress is necessary to stop blaming and keep on trying your best when you feel like one of your teammates is not performing. If you have say an 0-14 Genji, you could put a lot of effort to try and compensate his poor plays but this will just make the game last longer while the odds are still not in your favor. This results in a game where you felt like you played well and tried your best but still lost, very tilting and you will keep that resentment on the next games.
This the problem the devs tried to solve when releasing Overwatch 2 and the card system by the way. They wanted players to not focus on the outcome of each match but instead look at the results of a collective of games. They failed spectacularly of course, but I suspect a lot of tilting will come back after getting hit with "Expected" or "Reversal" following a match with somebody you felt was throwing.
How I would describe what "Elo hell" is to me in OW, is that at lower ranks, some skills that are very useful in high ranks just do not translate. I feel that in higher ranks enabling team mates, by creating distractions for the enemy team, your teammates are able to capitalize on that, this is especially obvious when playing support. I love playing support in diamond, but hate it in low plat/gold. Because what you do affects the outcome less often, you consequently lose motivation and focus and start to perform less and less.
Because OW is a multiplayer game, and not single player, the "victory requirements" are also determined by the players you play with, thus the "climate" of a certain rank. For instance in Diamond I can count on my teammates doing certain things and following up on things I do, this then allows me to count on them and giving me more options. Whilst in gold/plat they would just keep doing their own thing. Every rank has it's own set of rules.
So I am not saying that the gods of the ELO algorithm are conspiring against me, just that the skills i have and how I LIKE to play (and thus tend to do) are not as valuable in every rank.
Fun fact, CSGO had an actual elo hell in NA. Average rank was like silver 5 or something
Elo hell is nothing but
Copium we smoking that pack.
elo hell may well be a cope, but i can safely say i don’t remember ever having a comp match in bronze when i didn’t have gold objective time, even when i was healing, meaning i never had anyone on my team who was on the objective more than i was. when staying on the objective IS the thing that makes you win or lose, it’s frustrating to know my team never cares as much about it as i do. it just doesn’t feel fair. i’m very willing to admit that i’m not a gold player, but i just hate consistently being stuck on teams that don’t seem to actually try to win
I never really played much rank in OW, but I played a lot of Ranked in league and DOTA, and the idea of ELO hell is a lot less prevalent in DOTA. My theory, DOTA shows the number. I had a huge jump in the last DOTA MMR reset, indicating I was probably under-ranked before they changed the system and reset my MMR. But I never felt like I was stuck under my ‘true’ rating. Because my true rating was right there in my profile. Not as visible as the fancy badge, but easy enough to find.
Mostly off topic but I just want to thank you for talking at a decent speed lol
i know quickplay is not part of the discussion, but whenever people tilt in quickplay, i always tell them to quit the lobby if they hate playing with me so much.
In ranked, i feel less inclined, as i usually try to do my job, and switch whenever i have to
but just yesterday, when a rando went "dva never play the game again" to a friend of mine, i first of all told them to shutup and quit if they didn't want to play with us. Then, since my friend WAS affected by that, i made sure to let him know that i didn't care about his skills (he *isn't* good at the game, but that's why i play Quickplay with him)
I'm okay with people silently complaining to me, and I'm also okay when people complain directly to me, as i am 100% aware that my gameplay is not perfect, and i usually take their comment as a way to see my gameplay under a different lens, or, in the case of toxicity, i take it as a hint to push myself a little harder, think a little harder, and see how that goes.
i won a game with 2 toxic people in the team in qp, and as much as at the start i was doing relatively shit, and my team *was* carrying me, i eventually managed to feel like i was pulling my own weight, and one of them did recognize that, as much as it was salty. the other just went "got carried lol" and i ignored them completely, since they didn't realize just how much the team as a whole did. but yeah.
Just for fun, here's my pet theory for where the elo hell cope comes from
The meta shifts depending on the rank you're at. However "good" or "correct" play is often defined relative to play at the high ranks. It doesn't matter if you're playing "correctly" if the way you're playing doesn't suite the actual lobby you're in. People cope be cause they're playing "correctly", but get stuck because they never stop to reflect on the actual problem
For an example: Dead by Daylight is a 4v1 asymmetric horror game. 4 survivors try to escape 1 killer. When they introduced skill based matchmaking, they made it so that if a survivor escaped, it counted as a win for them, and if they died, it was a loss.
At high ranks, altruism is optimal. It increases the whole team's chances of survival considerably. However, SBMM only cared about your individual survival. If people play selfishly, the person playing altruistically almost certainly dies. So there's a sort of prisoner's dilemma situation going on. As a result, a number of really good streamers got kinda stuck in the mid ranks because they kept trying to play altruistically, and got punished for it
Yeah it’s a lot of blindly applying technically correct concepts without understanding them past a surface level I guess
An easier way to rank up is to duo with someone who's good at a role that isn't your main role, so if you're a pretty good tank player you should duo with someone who's good at support or dps.
My cope list
Step1: blame myself
Step 2: look at scoreboard to see who to blame
Step 3: explain how the neg dps was causing the loss justfor them to yell back
Step 4: yell and scream
Step 5: cry and cope
The thing I hate most about comp is solo players going up against duos or stacks
i never believed i was in ELO hell (or the related "losers queue" concept) but i was entrapped by the opposite idea.
A "winners queue" type ELO hell.
This would be "i suck at the game, i always get rolled. I should be bronze but im gold cuz my team carries me"
Tho not with overwatch cuz i dont play competitive there.
i dont know, maybe it is kinda coping on my part too cuz of how one-sided matches tends to feel, tho maybe thats just blinding myself by purposely forgetting the genuinely good times.
Every other person in the game has the same chance to be throwing. means if you never throw, your team will always have an advantage statistically, you are guaranteed to have a above 50% wr given enough time
I am guilty of not roasting my duo more. Except when he plays support. Oh boy, this man CANNOT lamp a Tracer bomb to save his life while playing Bap. I told him twice in one game he died to pulse, and I watched it happen both times. Both landed on the floor. My buddy made it to plat before I did on support
If you consistently out perform your teammates but lose because of them you are in elo hell. This is a fact.
I feel like mercy is an exception to this rule. Even GM players tend to struggle trying to climb ranks with her because she is so heavily dependent on your other team members. But then again, this may just be another form of justifying why i've been hardstuck in plat lmao
I mean yeah but some characters just can’t carry and if you need to carry sometimes mercy isn’t the pick and a more damage oriented support is. Having the ability to reasonably well play a carry character in your pocket is important.
generally with friends I acknowledge when my friends are the issue and acknowledge whem im the issue (I say I'm the issue a majority of the time)
I'll say that I trust Elo, sorta (there are other systems), but I don't trust Blizzard's ability to code without bugs at this point. They have admitted problems in the rankings, and some of those took more than one try to fix. Without open code, I can't fully trust them at this point.
I think another point not often raised is that Blizzard have admitted that ranks aren't equally distributed. There are larger masses in different ranks. This can mean that moving in one of those ranks actually takes many more games than someone at another rank. They've made changes to make this more visible, and I hope they do more on this front.
I see what you mean but I think some people are in elo hell and some people are using that as cope there are a lot of factors that play into a win and 1 person will not change the tide of the whole game sometimes not all the time
Once again, I am begging people to actually listen to what I’m saying before commenting. A quote from the script for this video: “It’s also important to remember that the pendulum swings both ways here. You can absolutely lose an Overwatch game through no fault of your own - but that also inherently means that you can win an Overwatch game through no success of your own. It’s a team game, and the results - barring the rare instances where one person absolutely single-handedly carries through a dominant performance - will always reflect the overall performance of a team, not the individuals. You will lose games you performed well in, and you will win games you performed poorly in. Complaints about Elo hell exclusively focus on the feeling that it’s only ever affecting you and your team and only ever doing so negatively, but statistically it helps as much as it hurts your team and the other team is dealing with the same situation.”
Your point was mentioned. Elo hell is not real, and pretending that these factors only affect you - and only affect you negatively - is a cope.
elo hell is an actual thing and im currently stuck in it, i was low masters in overwatch 1 and havent played since 2020 so i decayed to silver 1 in ow2, when im going 35/6 in a match where my tank is constantly feeding and my other dps is going 3/12 genuinely how the fuck is it somehow my fault that we lost that match
Had a similar experience with coming back to OW2 on an old OW1 account, i played close to a hundred of comp matches and my rank barely moved in gold. Made an alt account, placed diamond 2, made another alt, diamond 3, few more matches, rank jumped up. wouldn't be an issue if not for cosmetics on my account i guess.
Feels like your MMR moves less the more matches you've played so it gets harder to climb as you improve. Viveros just likes the word cope
This doesn't factor in specifically that Aaron Keller hates me in particular and hath decreed that I shall not climb.
For real though I've been on teams where one person has admitted to being a smurf for a much higher skill level and then other teammates STILL insist that they are being treated unfairly by the game because they lost.
Elo hell is 100% real. Last season i struggled to get out of plat 5, after the reset, higher and lower elos etc im now in diamond 3 and FLYING through it.
I just hope people get a good laugh out of the idea that you don’t just think Elo hell exists, but that you think it exists as high up the rankings as Plat.
@@TheViveros so ur saying i was just hardstuck until i wasnt. Crazy
Yeah bro it’s called the game underwent a significant change overnight and it’s affecting people’s performances. A whole bunch of people can suddenly hit headshots now that they couldn’t before. You were hardstuck bc you were at the correct rank for the state of the game back then. Your rank has changed because the state of the game has changed to something more favourable. This is not complicated.
@@arlechinosleftball you're still in time to delete this bruh 😭
@@TheViveros Look man its def not a one and done if im still performing well and holding my own in higher ranks. I get ur point but my experience SHOWED its real.
IF its any consolation for people I used to play comp a lot in Overwatch 1 back in 2017/20119 era. I was gold/plat. When I was a baby brand new to overwatch I was silver my very first placement and rose pretty quickly. Bronze at that time were people I assumed did not know how to play video games in general. Like not just lower rank but like genuinely perhaps children or those with sub 10 hours of FPS game time. I booted up OW2 when it came out and it threw my ass in bronze for the first time ever. These bozo's in bronze look like gold back in 2018 no freaking lie. I don't play a lot of comp and being placed in bronze def made me not want to play it so it wasn't till several seasons later that I put any effort into getting out of bronze. But as a whole if you have been playing this game for a long time and have gotten better over the years but your rank is the same it isn't that you haven't gotten better. It is that the quality of each rank has risen with the years. At least on the lower end as someone who played a long time ago and has recently jumped back into comp this last year. Its like inflation in the real world. "I used to make 50k a year and support a family of 5!" "Yeah that was in 2011. That doesn't cut it anymore."
Elo hell isn't real. So if you need a new cope thats one for you.
*Edit. As a part time league player I actually think an Iron rank would be great. If bronze is the largest rank in terms of broad SR it only makes sense to break it up. The number of actual sub 500 sr comp players has to be pretty small so 1000 and below can be Iron and 1000-1500 can be Bronze. I don't think that would be an issue. If someone is very low ranking giving them something to show for their progress is a motivator to improve. If someone just sees Bronze 5 every time they rank up no matter how good they do for say two, three or even four rank ups in a row it will just make them think the game is rigged against them. And I couldn't blame them personally. Not because it is but because OW2 not showing SR has been terrible. Just show us the number. Hiding the number does nothing and just complicates everything. Its like a big secret that the players shouldnt' and can't have access too. Poor design and no one likes it.
So in overwatch1 2.9k was considered elo hell. Even t500 players doing bronze-gm knew most there time would be spent in 2.9k. Why? Aaa a long list of reasons. It was a crossover rank but the people in it were either in all aim no brain or all brain no aim. In Overwatch2 this 2.9k rank doesn't exist anymore cuz the ranking system brought these sim players up and brain players a lil down. My opinion is that you shouldn't focus on the rank too hard. Just try to play the game
I don't personally believe in an Elo Hell formed by the actual game algorithms, mainly for the reasons that you've outlined in this video, but I do believe that some games have something fairly akin to it in their lowest echelons, primarily the lowest rungs of games like league of legends. Those places are absolutely rife with boosters, smurfs, trolls, and all manner of other situations that completely destroy games at a completely random rate that cannot be accounted for. In theory they would happen to everyone equally, and you being one of 5 on a team *decreases* the odds of one of those players being on your team, but in reality it often only takes one on either side to affect the outcome in a way that messes with whatever system they use.
For context I've been at above average ranks in every season of LoL that I've decided to play it, so this is not me coping at all. I completely believe I deserve to be exactly where I am. I just think that the gaming experience at the lowest ranks is completely awful for your mental, which will sometimes cause a downward trend in winrate even if your skill is high.
I don't believe elo hell is real, but people make it real. What I mean is, Elo hell is as real as a psychosomatic injury, or a phantom pain. There's enough non-merit factors that may determine whether or not an individual game or games may have results altered (a pc crash, bad ping, a teammate not playing their best that day), but it's not reflective of the entire season for the person claiming elo hell. Regardless, those determinant factors may be enough to convince the individual they should under-perform, and by bias, flake up their own lack of skill to it being a "already lost game", etc.
While the word cope is definitely triggering here lol… while i do disagree with the fact that “elo hell” as a concept is just nonexistent (def exists to an extent).. a video like this is refreshing tho instead of the very tired “JUST PLAY BETTER!” Coming from a streamer who only queues up with their fellow top 500 friends and plays for 15 hrs a day. I think there are definitely some rank ranges where stuff gets really weird (your bronze 5 case for example). And sure the teams should be “even” but a game that depends so heavily on dps/tank success (im willing to hear if this is not the case to someone else) i can easily see a support especially being kind of stuck somewhere at some point. Great video though! 👍🏾
Self improvement in game should be prioritized over rank. Once you get better ranking up will come with it.
8:15 I can’t get over how y’all didn’t contest there, tank was back and Kiri prolly had suzu so you would’ve lived. Great vid though honestly just gotta guh guh
what i dont understand is how im hardstuck low diamond on my main account but im my usual masters 3 on a side account. i load onto my main and i get back to back one sided stomps while i go to my side and the matches are fair.
Volatility is the reason. It's a "certainty factor" in the ranking system. When it's uncertain of your rank you'll move up or down more, which can boost newer accounts to a higher rank than intended for awhile (your Smurf) if you go on a lucky win streak. It happens, and has resulted in my Smurf also being higher rank than my main (quite often). The more you play on it the more likely it is you'll move towards your intended rank.
Maybe play your games as if you are playing in master
I think elo hell exists but doesn’t affect many. I’ve personally never experienced it but I have friends who I play with who are low ranked but pop off or play very well in my lobbies (higher lobbies) but in solo q they can’t carry. Idk what the problem is for him
I dont believe elo hell is real. Im pretty good at calling myself out on playing like a smoothbrain. But the whole tilting is a skill issue thing kinda opened my eyes. I mean i know i shouldnt tilt but hearing somebody say it like you did has helped me.
Only issue is, they use an MMR system, inspired by Rocket League. Rl is maximum 3v3, though, so it works better considering you can individually carry 1 or 2 people much easier than 4, also considering the roles in OW, where not everyone does everything, but you virtually have to cooperate for anything to work. Thats why it doesnt really work well in OW. Not saying this to cope, though, just criticizing OW's ranked system because its hardly good enough, even with the update.
While I agree that elo hell is just cope, there’s some serious fuckery that happens at the lowest ranks. Around 95% of my games when I still played was roll or get rolled and it made individual skill expression near impossible. I’m pretty sure this is due to the fact that accurate skill ratings start to fall apart a bit at the extremes. At low bronze you can have a person that doesn’t know strafing is a thing and someone who just came to overwatch from something like apex be only 100 sr apart, it’s really hard for the match maker to make remotely fair games
Do you have any evidence or proof that this is indeed the way the Overwatch 2 ranked system and matchmaking work?
I mean, yeah? Blizzard’s openly acknowledged that this is the ranking system they use. They haven’t kept this a secret, I didn’t realize people would be disputing the actual ranking system given how well-established that fact is. Overwatch is even specifically named on the Elo ranking system Wikipedia page as a game that uses it lol
Overwatch uses Elo, by their own admission, and this is how Elo works, so I don’t know what evidence you’re even asking for here.
I will just take your word for it then. I wasn't trying to throw any accusations at you, nor was I doubting your integrity. I am simply unaware of what blizzard has revealed about this. Just from playing the game I don't think you would get this information, but I could be wrong there as well. Just from playing the game the ranking system just seemed unclear and confusing to me. Although it seems a bit better now in season 9.@@TheViveros
u getting t baged in the beginnign was funny
I wanna say that elo hell is real because it does effect everyone i mean we are more than halfway through this season and t500 is still masters people are lower than deserved imo