Puzzle without a phase designation?

DetroitMichael

Hello,

I just did this puzzle and when I finished I was surprised that it wasn't listed as either a middlegame or endgame. Is this something new? An oversight? I've done literally 23k+ puzzles and have tracked the last 5k or so in a spreadsheet. This is literally the first in all the time I've been recording them without a phase listed.

lichess.org/training/nlPWl

Just curious.

pepellou

From the lichess puzzle database (database.lichess.org/#puzzles):

Total puzzles: 2,649,613
Tagged as endgame: 1,115,809 (42.11%)
Tagged as middlegame: 1,220,333 (46.05%)

There are no puzzles with both "endgame" and "middlegame" themes (although funnily there are 2 "middlegame" puzzles with themes "rookEndgame" and "bishopEndgame", which we can ignore because it's just 2).

That means there are 313,473 puzzles (11.83%) without either of those 2 tags.

The first puzzle ids are:
000mr, 001w5, 00206, 002Tf, 003YT, 003o0, 0045Q, 004X6, 0055Y, 005jJ, 005wy, 008D5, 0092V, 009FS, 009XT, 00AVR, 00DkJ, 00Ea3, 00EbJ, 00Evs, ...

DetroitMichael

Thanks for the response, but I think you misread part of my statement. I'm sure of the 11.83% of the puzzles without an endgame or middlegame tag, they are likely from the opening. I've encountered plenty of opening puzzles. What I was getting at was that the particular puzzle listed had neither opening nor middlegame nor endgame tag at all. That is what I found weird.

DetroitMichael

btw, I remember after encountering that one, I encountered 2 more. I just checked my spreadsheet and they were puzzles nlPWl and oi4sD.

lichess.org/training/nlPWl
lichess.org/training/oi4sD

something I noticed about all three is that they were played less than 100 times, but I've played plenty of new puzzles and never had a missing phase designation. It's no big deal, but like I said, I found it odd. I also only notice because like I said, for ever puzzle I do, I enter the results into my spreadsheet so I'm acutely aware of the puzzle lengths, phase, goals, motifs, etc.

pepellou

DetroitMichael

thank you for the analysis. so here's the question, who assigns the phase of the game? clearly, phase is not an option for us to add as a theme. are these puzzles themeless because they could be construed as either the end of the middlegame or beginning of the endgame? are they lost in a weird intermezzo if you will? to me, this is creating a philosophical, metaphysical quandary. how can a chess game not be in one of three states, either opening, middlegame or endgame? If a chess game is not listed as being in the opening, middle or end, does it really exist? Is it a chess game? If a chess game falls in the woods and no one is there to hear it, does it make a sound? so many questions, so few answers...

pepellou

@DetroitMichael said in #6:
> are they lost in a weird intermezzo if you will?

That would be my guess.
Those tags were automated, so at some point, you had to set a rule of what qualifies as endgame, as middlegame, etc.
This is a tricky rule to set, I think.

With my students, for example, I use the rule "ignore pawns, count the rest of the material and if one side has 14 points or less then we'll call it an endgame" (I use that to consider R+Q an endgame, but it's not a perfect rule).

An analysis on that list would be interesting to determine what's common to those positions, but... I think my curiosity is satisfied enough at this point ^__^

DetroitMichael

that's an interesting theory. looking back at my spreadsheet, I have a few QRR+5 vs QRR+5 that were labeled as endgames, so who knows what their algorithm says qualifies. and I have examples with piece value counts above and below that on either side of the middlegame/endgame divide, so there's some more calculus to it somewhere, that's for sure. perhaps move count as well? oh well, like you, not that interested in proceeding further, especially knowing there are over 100k such unlabeled puzzles.

This topic has been archived and can no longer be replied to.