[narag]

This seems like a wild claim. On chess.com I've gone from 500 to ~1050 in a handful of months without any real study,

Is rating on chess.com similar to rating in lichess? The article mentions 1%, but you start with 1500 in lichess??? If you lose often, you quickly go down. But if you keep an even record, you stay in 1500.

I don't understand how people learn from slamming blitz/bullet games.

As in the article, speed training helps recognizing patterns and dealing with them. At first it seems impossible to play significant moves with so little time, but you can try to just doing it. You'll adapt very soon.

I recommend starting with blitz 3+2 and later bullet 2+1. There are tournaments of 1+0... or even less time. 30 seconds is really crazy.

[VonGallifrey]

> If you lose often, you quickly go down. But if you keep an even record, you stay in 1500.

Is that actually true? You start at a 1500 in lichess, but that is a provisional rating and the outcome of your first few games are going to cause BIG swings in your rating. After your rating is no longer a provisional rating then that is going to slow down.

Losing a few games at the start might take a lot more wins to "even out" in rating. Just keeping an even record might not be enough to stay 1500.

[narag]

After your rating is no longer a provisional raing and the outcome of your first few games are going to cause BIG swings in your rating.

The rating algorithm is suppossed to be predictive. It's only natural that the first data points cause big swings, but the amplitude will narrow after a few games.

Then, providing you win the same number of games that you lose, you stabilize. Will you keep around 1500? I think so, because the algorithm selects oponents with similar rating and with my own mediocre ratings over 1500, lichess tells me that I'm better than 60% of blitz players and 65% of bullet players.

There's a caveat though. Ratings in Lichess are not "lineal". There are annoying discontinuities. I get stuck at some level but, if I manage to go up 50 points, I get easier (but better rated) oponents and raise another 100 points. Conversely, if I get down a certain level, I can expect to sink even more.

[getoffmycase]

I am currently in a run playing rapid 5|5 because I found myself consistently running out of time in longer time controls. To me, improving at rapid is making significant progress in my overall chess abilities. I developed a better understanding of end games and openings , so that I can spend more time being careful in the middle game. Recently I noticed that I don’t run out of time in Rapid, and that’s progress to me, regardless of my ELO

[smatija]

Rating on chesscom is wildly different than on lichess in lower rating ranges.

300 cc should be similiar to around 800 lichess. Around 2000-2200 ratings even out on both sites and on higher ratings lichess ratings tend to be lower than on cc.

For improving longer time controls are better. The old adage, that if you want to play blitz better you should play rapid and that if you want to improve rapid you should play classical is still valid today. Blitz is beneficial only at higher rating ranges (2000+) to allow practising openings, and even that only in moderation. Players who don't play classical time controls (90min+30sec/move at least) tend to plateau around 2000-2200 online.

This ofc doesn't mean that blitz isn't fun - vast majority of my games are in blitz time controls.

[narag]

300 cc should be similiar to around 800 lichess.

800 is very low in Lichess. If you can think just two moves ahead without losing pieces, you get 1000 easily.

For improving longer time controls are better.

Yes and no. There are different aspects to the game. Every variety is good for something.

After playing a lot of bullet, I win easily in blitz just because the time advantage. But the reverse is also true. After playing more slower games, my fast games get better.

And there are some parts of the game that you just need to sit down with a book and a couple of boards.