Playing solo competitive as I always do. Started out at 2800, climbed to 3200, then dropped to 2700, climbed up to 3100. I’ve been in and out of diamond 8-9 times now haha.
This isn’t a complaint btw. I know playing solo that I have very little control and you are at the mercy of matchmaking. But what does make me smile is situations where I’ll go on a 10-12 loss streak… get down to 2700, the game then seems to go “oh sh*t, this isn’t right” and there will be a massive overcorrection and I’ll go on a 10+ win streak.
Over a few days I’ll drop 400-500 SR then gain it all back again.
A lot of players have reported having the same phenomenon. I documented the wins and losses over 100 games and every streak of wins or losses are almost perfectly mirrored. Based on reports and my own data it definitely seems like the matchmaker has some sort of influence.
It’s an imperfect system with too many people taking advantage of ways to manipulate it (smurfing, throwing, getting boosted, etc) so when you lose a game you’re statistically designed to win, you get games that are harder to win. This is because the algorithm is using the data it has (i.e. you losing a game you should have won) to decide that you belong at lower SR. Sometimes it over-corrects and you get lower than you should be, then it over-corrects again and you get higher than you should be.
Nah, see above. People need to accept that they belong in a range of SR. If he swings between 2800 and 3200 he’s comfortably a “good plat” or a “bad diamond”. That’s not that big of a swing when you consider it’s on a 5000 point scale. There are personal factors beyond people manipulating the system like how you sleep, hormones, etc that impact decision making and mechanical ability.
I guess it hasn’t occurred to ya’ll that if you hover in a few hundred SR range, it’s not some conspiracy to keep you there but rather it’s where you belong?
I have multiple accounts and they’re all within a few hundred of each other at most and most are within 50-100. That type of consistency doesn’t happen by accident. The algorithm literally measures your APM and gives you games designed to keep you according to that. If you step on the bathroom scale and it reads higher than you want it to, is that a conspiracy too? No, it’s measuring you the same way the algorithm does.
It would be interesting to know the signal-to-noise thresholds by rank. This requires knowing the pdf of SR by % players (which know this on paper and from statements) to the SR by % accounts (live parked actual).
You can absolutely belong in a different SR range if the signal/noise is too low. Terrible gold players floating around buffered by new entrants and smurfs going in every direction. Their signal is low the noise is high they float around too high for where they ought to be. Bottom500 is another great example because it’s an EXTREME case where the effects are maximized. You can be silver/gold and just sit in that rank. The bottom 0.1% is way more populated than it should be (alts) and the statistical noise works out to plat-like gameplay and gold-like stats expectations. Meaning you’ll just float there and never rank up unless you break out by teaming up with smurfs or using fresh mmr accnts. A proper ranking system would instantly recognize you’re playing like, e.g. an 800sr player and shoot you out of <500sr. I mean just what kind of afk bots should <500 sr even be? Sadly it’s filled with T500 icons.
Secondly, “thin” ranks like masters can become made “thick” by the same small% repeat placing accounts in that rank, shifting SR to where it doesn’t belong. It is NOT about your skill or your rank (mmr, alts) it IS about where SR is located (ruined by alts and no-resets). So the signal/noise again for Diamond say, can be washed out by someone with 9 accnts who just parks them on the edge of plat/dia to gatekeep.
To be clear, are you asserting that this statement is beyond the realm of assumption and rises to the level of fact? Or are you theorizing?
“Nah?” So Blizzard implements no matchmaker-based strategies to boost engagement? There was no versioning or split-testing analysis to finalize a version of the matchmaker that offered the best or most profitable metrics for Blizzard. Interesting. What confirms your thinking on this?
Why do people need to accept that they belong in a range of SR? What convinces you that this is the only viable theory?
Why would a person care about the size of the swing of the total population, if their performance has been consistent? This would be like ignoring the idiosyncratic tendencies of my professor when it comes to creating exams which effect my grades in the class, and being told to find comfort in the fact that my percentile relative to the total population hasn’t changed enough to be cause for concern. Why the hell shouldn’t someone be concerned about wasting their time and energy in what is arguably artificial grind to drive engagement?
And there are corporate factors like corporations being corporations. The reasonable person is open to both possibilities, are you?
Here’s another question: why do developers implement grind into their games?
Issues like smurfing are objectively worse in lower ranks. There are simply more people who are higher ranked and have the capacity to smurf there. GM smurfs in masters, masters smurf in diamond, etc, but bronze players have nowhere to smurf.
There are occassionally people who are on an account WAY below their main (T500 in bronze) but it takes a LOT of effort to stay that low. I’m only diamond/masters and I win over 100 SR in bronze games with PBSR. Most “smurfs” are within a few hundred of their main and it’s their grouping account, an account they play other heros whose kit they haven’t maximized but game sense alone keeps them not too far off, etc.
But let’s be real, people in bronze aren’t there because of smurfs. They are there because they are bad. They cannot aim, can’t make plays that have impact, etc. I’ve climbed out many times and it takes like 2 hours to go from bronze to gold for me and again I’m only diamond/low masters.
Just quoted a random sentence to make it easier for people to follow.
It’s their game so they can do whatever they want. We can opine all day on how strict W/L with no finger on the scale for performance but they paid money for the algorithm, they’re gonna use it.
Regarding belonging in a range, what else could it be? You’re telling me that every single day you play at the same level? You have no external life factors (stressors/anxiety/sleep/nutrition/substances) that impact your decision making and your aim? I dunno about you but I definitely have days where my aim is ON POINT and I’m getting kills I wouldn’t always get. These “good days” impact your MMR, bringing it up. Then when you have a “bad day”, it drags your MMR down. That’s why people go on SR swings.
There’s also randomness in your opponents. You might be a Doom one trick and you just happen to get a bunch of Sombra mains or a Pharah one trick and there just happen to be a bunch of hitscan mains that day shutting you down. It’s impossible to nail your skill down to one exact SR, that’s why people fluctuate.
Agreed I’ve been educating people on this topic for years. You don’t need direct math (data-based evidence) just a potential algebra.
It is like 18 minutes for a bronze dps queue 11 mins prio. Then you start at lowest possible bronze with anchor stats that need to be overcome. You are magically silver/gold skill level (watched a Jayne vid). But you solo queue, so the game finds you throwers/smurfs/stacks for mmr-rigged matches that adapt to your gains every step of the way, and you get hosted on some brazillian 2cp server every 5th match. Not to mention cancels and leavers. If you sweat, you only ever move up 20sr. You can go 45-2 on genji and gain ??? three times in a row in the bottom500 because they don’t even track your pbsr there.
How are you going to gain 1000sr in 2 hours under these conditions?
Evidence for that statement on their psychology? I’m not doubting it I’m simply saying we don’t need to make any claims about the motivations of smurfs. They’re all alts to me and my math, ruining the curve of SR by % players into an SR by % accounts.
Alt effects (on rank integrity) can become much worse the higher up you go. A single masters player filling high plat / low dia border with alts gives them a lot of gatekeeping power and disposability. Passing through diamond should be thin (beat a few % of players). But if there are too many accounts there then there ought to be, the progression is scammed away. They could practically wintrade friends into higher rank with that kind of SR where it doesn’t belong.
Every alt ruins what ranks and elos mean.
Making climbing effectively moot. Fake accounts in fake ranks in fake rigged lobbies.
On the topic of inconsistencies, there’s also a great deal of randomness among teammates that people don’t seem aware of. Using the context of striving for “random matches among your peers” as optimal, it’s not going to be the matchmaker’s fault that you got two lucio OTP’s, or someone who decides to force rein on dorado, or someone who refuses to swap ball against sombra/mei and doesn’t actually know how to play into the countercomp, or that you loaded into your literal worst map against someone’s literal best map. Or perhaps a smurf who was chilling under the radar at 75% attention and decided to “turn up” for the match and carry you because someone on the other team tac crouched.
Those are cherrypicked examples sure but I don’t think it matters which cherry I pick up when we’re in an orchard
You can easily get hit with streaks of these consecutively, both up and down the ladder. Happens often enough that people will come to the forums and their first post would be something absurdly not matchmaker specific, followed by, “fix your matchmaker!” You also see it a lot in their social media feeds.
the “snr” is a good helper analogy for why people range above and below the rank they actually “deserve”. I’m using the term deserve with the lightest of implications, lol.
Throwing aside any speculative or nefariously patented yet not proven to be implemented mechanics that may or may not just be part of a “corporate IP arms race”, the system’s accuracy could almost certainly be improved by getting rid of superfluous inconsistencies that are baked in by game design/admin.
It’s already low given the competitive environment of this game is so… unfocused? I’d compare it to trying to network without having any standardization for the insulating you’re requiring on your wires. There’s no standardization for what’s throwing, what’s trying your best, what’s an acceptable use of an alt account, very little done to attract like minded individuals together (casual + tryhard cause no real tryhard mode besides scrims and casual modes = joke). They’ve basically got a ranked matchmaker who’s goal is to appeal to literally everyone at the same time, except arcade players.
Then there are the inconsistencies the game itself adds (or encourages) both in its audience and its design. The obvious one, tank synergies. You’re playing at your proper elo with little room to carry and you’ve got hog dva against double shield. GG NEXT MOST LIKELY.
When you add dishonest ratings and dishonest intentions into the mix, even if you used the classic “play until you lose two” method to help iron out your own biological inconsistencies from your playtime, I would not be surprised to see long term ranging of +/- 500sr given no other improvements to the player.
It’s rather insulting that you think I’m just that stupid. Maybe it hasn’t occurred to you to read what I actually posted or maybe not comments at all if you’re not going to read it.
That’s y I asked blizzard to just implement a deranking button years ago. Nothing worse that being in basically enforced loss or win streak. Just Skip the BS streak and return to a median rank or something.
Tell me about it. I’ve been between 2900-3100 on support for the most part since role queue came in. Some seasons admittedly I didn’t play a lot, I’d place and maybe not play more, but some seasons I would play a decent amount.
This season started really badly for me and I’ve barely won a game since. I dropped from 2986 all the way down to about 2575 in the space of about a week.
Some of those games I know I didn’t play so well, but the games just seemed unwinnable most of the time. Once I got that low, I immediately won a few games in a row and the games started to feel really easy again. Now I’m around 2675ish I think, and I gotta climb back to where I usually am.
This has already been happening to me a lot on tank for the last couple of years. I’ll be around low plat then drop all the way to low gold and then climb up again. Have done that drop and climb about 5 times now.
