Who are the Elo gatekeepers?

Recently I've seen a number of references to the 'elo gatekeepers'. What is meant by this? Do they exist or is this a metaphor? If they do exist - what do they do, where, with whom and why? 

There is some dude JankogajdaskoLEM who thinks chess.com have a conspiracy that keeps people at low elo. These accounts that "keep them at low elo" are called gatekeepers, and they don't exist. Or, they do in a way, but they are just like other people who want to win. They are not stronger than their rating, and the reason Janko is bieng "gatekept" is that he's not improving, while the community of ~600 rated people is not getting worse.

Aha @Superplayer7472 thanks for the info. I have read theories that ratings are fixed algorithmically, for example if a new member wins their first games they will be encouraged to stay on the site, which benefits chess.com, and suggesting the possibility that there are bots disguised as humans to achieve this. But this sounds a bit far-fetched to me.

Yes, that's ... just a conspiracy theory.

Some people prefer to point blame at others for their own lack of chess progress (at a conspiratorial "system" intent on deflating their ratings, for example) ... In most cases though, the difficulties actually come from the player's own choices.

I'd say the biggest barriers to progress are one's study habits (or lack thereof), and the desire to rush into new games, instead of taking the time to learn from the games recently played ...

@EloGateKeeper what evidence is there?

I have not been oppressed or 'gatekept'. I am in my early 40's and am improving. I just hit 1500. Months ago, 1500's were too strong for me, but now they are tough games I have the ability to win sometimes. 
My 7 yr old(soon to be 8) son has gone from 100 to 1000 since February.

If he keeps practising and studying he'll get even better, if not he'll stagnate.

it's a nonsensical conspiracy theory that chess.com has bots that gatekeep how high peoples ELO can get or something like that. Like they'll claim if you get too high of a winstreak you'll get force paired with one of these higher performing bots to regulate your ELO.

Not sure if im getting it 100% right, but its something similar to that, and its all idiocy. Chess.com wouldnt waste their time with that. some people cant accept that they can lose a game, and in fact lose more than one game in a row fair and square.

I've read basketstorm's post https://www.chess.com/forum/view/general/checking-if-elo-system-is-oppressive-with-proofs but it doesn't really make sense to me. Admittedly, I don't know much about coding and algorithms. But basically I have to ask myself "why would chess.com do this? How would they benefit from restricting people's progress?"

"Joined 17 hours ago"

Beware, by asking such questions, you basically invite trolls who would try to convince you that everything is ok, that it's just the skill issue of novice players. Do your own research. I can explain though, there are three versions:

1) Chess.com belongs to the evil Chess Matrix org. That org forces them to do the actual Elo gatekeeping. From arranging rigged games to more serious things like time manipulations. It is the most adequate version but it does not answer your "why would they benefit" question, so read next:

2) They've deliberately crafted this unfair system to encourage users to pay for subscription that offers lessons and puzzles. Your rating drops - you think you're not good and need teachings and advanced reviews. But those things aren't free here. That's how they would benefit. That's a possibility, I'm not claiming that's true. But just think about it.

And version #3) They just don't care about accuracy. This version has the most proofs. Because the original Elo rating system (that was designed for FIDE by Arpad Elo) has certain guidelines and warnings that are clearly ignored here and even on lichess.

Some other proofs are:

- mathematical simulation that displays impossibility to obtain fair ratings just by using increments especially when a hard rating floor is present.

- Arpad Elo used recalculations and only allowed increments in non-localized pools. Before merging localized pools he advised special tournaments with subsequent recalculations. FIDE recently applied recalculation to all ratings up to 2000. On chess.com severe pool localisation is ignored, recalculations never happen and increments are the only tool to adjust ratings. This makes possible for really strong players to climb steadily, but unfairly puts the rest in a random chaotic wandering in low-Elo zone.

- Rating of engines is precise, measured against GMs and corresponds to FIDE ratings at GM level. Can be scaled down (also precisely), because it's all just math. It doesn't lie. And engines don't have bad days. So we can use engines as reference. And the thing is, most novice players here (100-400 chess.com-rated) beat 1000 or even 1300 bots at least 50% of the time while beating 250-rated bots 90-99% of the time. That means that their actual Elo level is 1000+. Because Elo scales up and down. And because just by taking an established rating of someone and your winrate against him you can estimate your rating. That's the idea behind Elo. Only few people understand this, sadly. Here we don't have a proper chain from GM level down to various rating levels. They are all distorted. We don't know how actually strong 1000-chess.com-rated player is. But obviously much more stronger than 1000-FIDE-rated. So this issue touches not only very-low-Elo players.

And the most obvious proofs are here in these videos, just watch every vid till the end to learn the truth

' The only thing holding me back- is everyone else'

No need for me to try and get better- as the whole world is against me.

^ That type of thinking is just sad.

People seem to forget it took everyone of us months, just to learn how to tie our own shoelaces. why expect chess genius level in just a few months?

It took me two years to get to a peak of 1700 elo on this site. Apparently it also took me two years to turn into a cyborg created solely to keep other peoples elo low by being better at chess than them.

Dont waste peoples time with conspiracy theories until you have some actual evidence. This hogwash is not necessary - for example everyone who has some chess experience understand that the bots on the site are not very accurately rated. Theyre not meant to be, the rating is approx adequate but the purpose of these bots is to simulate certain styles rather than ratings.

You are wrong. Bots on this site is just Komodo engine with various settings. It's true that it has style setting but it can also play at various strength levels for each style. I've used this engine in a different program. Komodo wasn't even made by chess.com originally, chess.com reportedly started purchase process in 2018 and finalized it last year. Komodo rating is precise and it is calibrated against pool of all engines and grandmasters also competed in that pool, so it is in line with proper true FIDE ratings. It's the players' ratings that are not accurate, not bots' ratings.

One of the original authors of Komodo was GM Larry Kaufman. And by the way he pointed at FIDE versus chess.com rating differences and provided some proofs and calculations. I don't know if those calculations work now, it was some time ago. But he knew. He also worked on other famous engines like Rybka, so he's not some random dude, he knows a couple of things about chess. In 2023 he joined chess.com Komodo chess team as a consultant. So stop implying that bots here are something unprofessional or fake or badly configured. Great minds are behind it.

It's certainly a better theory than the idea that I lose games because I play like a chimpanzee.

Yes, they are Gate keeping big time. your rating will always be +- 200 points once your established and non provisional player. Ive been researching this for years and I know for a fact chess.com on ALL levels of play 800-2200 keeps you around your current chess.com rating. They do this so you keep playing and trying to get better etc. I went on lichess back in the day and got all my ratings over 2000 for bullet rapid and blitz and then QUIT lichess and moved on to chess.com. My friend did the same exact thing and said lichess is a joke. Chess.com on the other hand makes sure you win and then you lose. I watch 2100 rated players make the same blunders as the 1800s and play exactly the same openings and same tactics all the time. Theres 1600s on chess.com could beat a 2000+ player no problem!

To be Frank, everytime im about to cross like 1900 I get paired with someone with 156k games and their chess.com INSIGHTS shows that they win 90% of their games against my opening i usually play as black or white! And this happens over and over again untill guess what? Im back to my normal rating.... its so frustrating and makes me do more puzzles and pay chess.com premium so i can review my games LMAO>

I am the gatekeeper

Lichess is heavily inflated... 2000 lichess is like 1400 chesscom