Reaching 2200 Elo in bullet chess

I wrote a post last year with some reflections on reaching a 2000 Elo rating in bullet chess. I reached this rating on Chess.com as I mostly play chess there with friends and colleagues. I even went as far as to say that I liked certain game mechanics better on Chess.com than Lichess (in particular how the premoves work). However, since then I have not played any bullet chess on Chess.com.

Recently, I have played bullet chess on Lichess. Since March, I have played 1,187 games with a total time spent playing of 1 day and 10 hours. My rating is 2203 with a rating deviation of 45.02 (pretty sure the decimal places are of no use here). The highest rating I have reached is 2288. Accordingly, I believe it is fair to say that my bullet rating is at 2200. In this post I will outline a few observations on reaching 2200 Elo in bullet chess on Lichess.

As in my previous post, it should go without saying that I am not 2200 Elo great at chess in general. I still believe these numbers primarily reflect the ability to recognise patterns and play the clock at pace (even if I only rely on my mousepad). And I now have data to show it. I have been playing bullet chess for at least 15 years, on and off, and I am sure I am good at it, but it still feels like a "fake" rating (or inflated at best).

On Lichess I can explore how exactly the games ended for games where I won and lost. In the figure below we can see the results for my 563 victories and 549 defeats (75 games ended in stalemate). There is a clear discrepancy between the types of games I won and lost. I won almost two out of three games by running down my opponent's clock, and only around one in eight games by checkmate. When I lost, I almost lost an equal number of games via checkmate, clock flag, and resignation.

It is fun to point out that, in my previous post, I said never ever resign when you play bullet chess, but when I look at my own games, I do that in ~33% of the games I lose. I guess this is to not waste my time (or the time of my opponent). If we both have, say, 45 seconds left and I am in a losing position, there is no need to finish the game to checkmate - even if I could potentially reach a stalemate in a few of those games. Life's too short.

The fact that I win a majority of my games due to time is another reason why I do not enjoy playing bullet chess anymore. I am not getting better at chess as I am stuck in a local maximum. That is, it is not a good sign that I am good at playing the clock and only win one out of eight games with a checkmate. As I wrote in my previous post, the irony is that I play bullet chess in order not to think, and if I want to improve, I should play less bullet chess.

In addition to playing bullet chess I also tried a few games of ultrabullet chess. Here you have only 15 seconds on the clock with zero increment, and here you rely primarily on premoves and most games, if not all of them, are down to who can make the best premoves (and in particular to make it difficult for the opponent to make good premoves). This is way too stressful and not something I would like to be better at. It is a good skill to have in certain endgames, but not really useful or fun with 32 pieces on the board.

In sum, since March I have played more than 10 hours of bullet chess on Lichess. That is one day and ten hours I will never get back. I believe in the importance of wasting time (and I am obviously very good at it), but there has to be a better quality to wasting time. Accordingly, I have decided to close my account on Lichess. No more bullet chess for now.