Lichess Time Control Statistics

0.75+17, haha, that's original & unique :D

I suspected 5+0 would be the most played, it offers a quick game but still gives enough time to think through a tough spot here and there.

@Blind_VarLur that might have been me. I was bored, and discovered that you can type a custom increment by going to "Inspect element" (in Chrome) hovering over the increment slider and typing in a number. Useless, but I found it funny in a way.... EDIT: I think you can o that in minuotes per side as well, so you could have a 17+17 game

Thanks Thibault and Vhsdream for the info!

@vhsdreams

Thanks much for collecting that information!

I've never really done much playing or observing of slower time control games, so the placement of 10 0 surprised me a bit.

The top 6 all being some of the quick pairing time controls and the more popular tournament time controls makes a lot of sense.

I am still somewhat surprised that 1 0 isn't higher up; to the point @tpr made, I would have assumed that it had a good chance of being the most played by raw game count, since you can play so many more games in a particular amount of time.

If you make some sort of adjustment like @tpr mentioned, looking at something like time spent playing a particular time control, then there's a decent chance the 10 0 would be the most popular.

It certainly is in terms of the most time committed to playing (i.e., if you treat every game played as lasting the maximum duration).

Since there are clock times, I guess it's technically possible to get the actual duration of each game and sum it for each time control to get actual time spent playing, but that may be overkill :)

On a separate note, @vhsdreams, what tool did you use to generate this?

I've never dealt with pgn files of this size before, and have always used pgn-extract as my command line tool of choice, but it seems a bit clunky for this task.

Thanks again!

@a_pleasant_illusion

Just a simple script (read the file line by line -> count occurences of '[TimeControl "some time control"]').

Yes, it's possible. I think the simplest way would be to calculate a real duration of average game of every (or better just top 50 or something) time control, multiply it by number of games and compare total time spent playing. I might do it later as I'm curious myself.

I wonder how the results would be affected by berserking in tournaments.

@a_pleasant_illusion

I'm not so sure about my results (didn't check everything too carefully and I'm not a programmer), but they look believable at first sight, so here they are:

pastebin.com/GsCPia6X (time control / hours spent playing)

or with more data (you can copy & paste it to MS Excel):

pastebin.com/0EU6uxrf

Notes:
- only games with at least 2 moves (I mean one by White and one by Black)
- both time increment and berserking were taken into consideration
- there's an option where you can add 15 seconds to your opponent's clock and I didn't take it into acccount (too complicated to implement and in some games it's undetectable anyway)

PS: If the results are correct you were right about 10+0 ;)

To really know the popularity of a time control from the database, we must search quantity of players playing that time control. I mean there are two variables, players and quantity of games per time control. I mean popularity refers to how many players prefer an specific time control, and how many games they play in that format. For example some player prefer super fast TC so they play a lot of games but maybe more players prefer a bit slower TC and produce less games for the database. It also must be taken into account the fact that a player can prefer slower TC and also appear in the database with almost the same quantity of games played in fast TC because they are easier to accumulate.

@ezequielmeza

I think my approach covers everything you said (I measure time control popularity by the total time players spent playing it - look at the second link in my last post).