<Comment deleted by user>
I would like to know how many times some games are replayed identical. What game holds the record as the most repeated game, things like that.
You can use the opening explorer for that lichess.org/analysis#explorer
Are the very short one movers in the database? If they are they also would impact the percentages of the WDL. Image playing ...
1. e4 and not continuing to play. Having many games resigning early surely impacts the results displayed in the WDL analysis.
Maybe a filter while viewing the explorer analysis would be useful, like setting a range of moves and/or a range by numbers of exchanges in a game. If you have 30 exchanges than it's a drawn game. If you want to remove the short games you would use a range slide or type a min-max length of moves.
Example:
View games from move 6 to 50. I would assume the WDL percentages would be different.
Or
Material Exchanges by groups of quantities: Early exchanges 1 to 10; Mid exchanges 11 to 20; End games from 21 to 30 exchanges.
1. e4 and not continuing to play. Having many games resigning early surely impacts the results displayed in the WDL analysis.
Maybe a filter while viewing the explorer analysis would be useful, like setting a range of moves and/or a range by numbers of exchanges in a game. If you have 30 exchanges than it's a drawn game. If you want to remove the short games you would use a range slide or type a min-max length of moves.
Example:
View games from move 6 to 50. I would assume the WDL percentages would be different.
Or
Material Exchanges by groups of quantities: Early exchanges 1 to 10; Mid exchanges 11 to 20; End games from 21 to 30 exchanges.
@thibault said in #23:
> You can use the opening explorer for that lichess.org/analysis#explorer
Not really, isn't the book maker eliminating duplicates at some maxdepth. or is that old news?
I always wondered at what depth would there be no more actually played duplicate prefixes of games.
I might have gotten some thing wrong about that.. But the book, like the opening explorer is about branching relative popularity. 2 identical games prior to maxdepth would be counting as 1. Yet, we can follow many games to their end..
This has always been a mystery to me. (I gave up figuring bymyself soeme time ago, digging through issues on github). But here someone else has the same curiosity (also some people elsewhere have been asking about the probability that in some pool per some time period, 2 games be played the same. It does appear to me that the question depends on the depth of common prefix sequences.
And my past issue digging on GitHub (closed issues are kind of full of interesting things), did mention duplication rejection upon some parameter called something like maxdepth. So I would be curious about how did that evolve, update my understanding (also help others understand the difference between a game database, and a book digestion of a game database, such as what the opening explorer give us).
There could be a huge maxdepth.. In the end being able to have the game database statistics before any digestion might put into perspective sequence popularity per depth (not book branching popularity, which would be relative to the maxdepth zobrification induced pruning of the full database).
or how does lichess choose maxdepth. I don't think many of us can have the whole database at home.
Repeating, a bit: what happens to the games that were excluded from the book digestion? do we loose access to them. How many of the games, that we can still follow in the explorer to their outcome, beyond maxdepth?
> You can use the opening explorer for that lichess.org/analysis#explorer
Not really, isn't the book maker eliminating duplicates at some maxdepth. or is that old news?
I always wondered at what depth would there be no more actually played duplicate prefixes of games.
I might have gotten some thing wrong about that.. But the book, like the opening explorer is about branching relative popularity. 2 identical games prior to maxdepth would be counting as 1. Yet, we can follow many games to their end..
This has always been a mystery to me. (I gave up figuring bymyself soeme time ago, digging through issues on github). But here someone else has the same curiosity (also some people elsewhere have been asking about the probability that in some pool per some time period, 2 games be played the same. It does appear to me that the question depends on the depth of common prefix sequences.
And my past issue digging on GitHub (closed issues are kind of full of interesting things), did mention duplication rejection upon some parameter called something like maxdepth. So I would be curious about how did that evolve, update my understanding (also help others understand the difference between a game database, and a book digestion of a game database, such as what the opening explorer give us).
There could be a huge maxdepth.. In the end being able to have the game database statistics before any digestion might put into perspective sequence popularity per depth (not book branching popularity, which would be relative to the maxdepth zobrification induced pruning of the full database).
or how does lichess choose maxdepth. I don't think many of us can have the whole database at home.
Repeating, a bit: what happens to the games that were excluded from the book digestion? do we loose access to them. How many of the games, that we can still follow in the explorer to their outcome, beyond maxdepth?
4,996,127,933 (standard chess) + 24,898,428 (antichess) + 19,442,472 (atomic) + 17,547,492 (chess960) + 20,962,446 (crazyhouse) + 4,839,286 (horde) + 5,617,956 (king of the hill) + 4,350,326 (racing kings) + 7,172,612 (three-check) = 5,100,958,951 (all games)
My games account for 0.000000268% of games. I want compensation for the 0.000000268% that I contributed.
@bufferunderrun said in #11:
> A bit more than the 8 minutes. Anyways, 10 billion by the end of next year?
10 billion by the end of the year !!
> A bit more than the 8 minutes. Anyways, 10 billion by the end of next year?
10 billion by the end of the year !!
An option to remove flagged games might give a different perspective to the results.
Can we do something like this please?
youtu.be/WaVCzMfKf8g?t=52
youtu.be/WaVCzMfKf8g?t=52
This topic has been archived and can no longer be replied to.