Puzzle Rush

As long as the puzzles get some quality control. I've come across quite a few tactic puzzles here on lichess that get voted down to zero by the players because they are considered pretty bad. With the current way lichess picks its puzzles automatically from user games, some bad ones are bound to get through.

Puzzle Rush! Puzzle Rush! Puzzle Rush! Puzzle Rush! Puzzle Rush! Puzzle Rush!

I have seen a few requests for puzzle rush. While i am not against the idea i would much prefer a tactics battle. This would involve some number of players joining puzzle tourney then they all get the same puzzle set and race to see who can solve the most in a fixed time.
For me this would be more interesting.

@glambit +1

And generally speaking I like the idea of trying to innovate and create something new and even better, over simply copying their idea.

> "From what I've seen, lichess prefers to innovate rather than copy."

So let's innovate! ;-)

The problem at the moment with Chess.com's Puzzle Rush on the free account is that the daily limit is *one* Puzzle Rush per day. It's interesting that the limit is set so low - how are newish users supposed to get hooked when each rush lasts a maximum of 5 minutes?

As for the current puzzles on Lichess - my main "issue" is the wild (often downward) swing in rating that's possible if one tries to blitz them or do them without absolute full attention (especially when difficulty varies randomly from puzzle to puzzle).

A series of puzzles of *gradually* increasing difficulty that doesn't affect the Training points would resolve some of this.

I also wonder if a time limit of 7 minutes would be nicer than 5, with a tweaked difficulty-increase curve (55 points for the current all-time best seems like a low number that doesn't provide a lot of differentiation).

Is it too early for name suggestions? Speed Tactics? Fast Puzzles? Maybe not Speed Solve though cause that's what the Rubik's Cubing community uses...

Whether or not it affects rating points seems a little immaterial. Puzzle Rush does seem to be a reasonable way to help people work on tactic speed, but it's inaccessible without paying for it. Providing a resource to people who cannot afford it (for whatever reason) seems to me to be a form of innovation.

The puzzle tournament system proposed by @glambit seems interesting, but I would likely not participate in those tournaments. I see tactics puzzles as a way to improve my own play -- and for that, I'd like to have the opportunity to play without feeling like I need to compare my play to other peoples'.

There are tunables we could tweak: how many you have to get wrong before the spree ends, or perhaps you need to get N in a row incorrect, or something like this. Also, we might be able to do something interesting with spaced repetition, where you have to re-do puzzles you get incorrect after some period of time. This is an incremental change, but indeed an innovation upon the idea.

Recognizing that patrons don't get anything special, I just decided to become a monthly patron in support of this idea. Lichess has already given me so much, so I feel like I already owe for that. At the same time I hope that tactic speed training can be another thing that Lichess can help me out with (especially since most of my time is spent in bullet and blitz games). I'd rather play here and support this community.

Ruzzle Plush please.

Sounds like a good idea, just think the quality of the puzzles would need upping before doing this. The Chess.com puzzles make sense to me and don't seem to penalise me for breaking order or insisting on some nonsensical rubbish that a human would never play. The puzzles on lichess make me rage quit more often than not.

Please downvote puzzles that do this, and upvote puzzles that you like (or maybe even those that you don't hate), even if you don't solve them. It'd be possible to require a threshold for a puzzle to appear in rush, say anything with fewer than 10 upvotes. (I don't know what percentage of puzzles fall into this category, but I see a good few.)

Admittedly, I need to get better at doing this myself, but it's a little weird with the puzzle interface on mobile being semi-broken.

Here is another idea : instead of losing after three strikes, why not remove something like 20 seconds for a failed puzzle ?

As for the puzzles quality, it is true that sometimes one can find a variation that gives say +4 and lose because stockfish had a variation with +4.3 in mind : this can be indeed very frustrating. Maybe puzzles with a lot of downvotes should simply be deleted (is it already the case ?).