#Weird bug with eval graph an API

4 messages · Page 1 of 1 (latest)

shut hinge
#

OK, so this is a weird one that I have no clue how it happened. So there's this PGN, right...

[Site "https://lichess.org/vi2zOlBO"]
[Date "2022.12.03"]
[White "rasulovvugar"]
[Black "penguingim1"]
[Result "1-0"]
[UTCDate "2022.12.03"]
[UTCTime "21:54:32"]
[WhiteElo "2388"]
[BlackElo "2363"]
[WhiteRatingDiff "+5"]
[BlackRatingDiff "-7"]
[WhiteTitle "GM"]
[BlackTitle "GM"]
[Variant "Chess960"]
[TimeControl "180+2"]
[ECO "?"]
[Termination "Normal"]
[FEN "rbbqnknr/pppppppp/8/8/8/8/PPPPPPPP/RBBQNKNR w KQkq - 0 1"]
[SetUp "1"]

1. d4 { [%eval 0.25] [%clk 0:01:30] } 1... d5 { [%eval 0.13] [%clk 0:01:30] } 2. Ngf3 { [%eval 0.24] [%clk 0:01:30] } 2... c6 { [%eval 0.04] [%clk 0:01:30] } 3. c3 { [%eval 0.62] [%clk 0:01:29] } 3... g6 { [%eval 0.75] [%clk 0:01:28] } 4. e4 { [%eval 0.92] [%clk 0:01:28] } 4... dxe4 { [%eval 0.55] [%clk 0:01:28] } 5. Bxe4 { [%eval 1.77] [%clk 0:01:28] } 5... Nd6 { [%eval 1.43] [%clk 0:01:27] } 6. Bc2 { [%eval 1.6] [%clk 0:01:27] } 6... Bf5 { [%eval 1.59] [%clk 0:01:27] } 7. Nd3 { [%eval 1.81] [%clk 0:01:26] } 7... h5 { [%eval 1.86] [%clk 0:01:25] } 8. O-O { [%eval 1.96] [%clk 0:01:25] } 8... Nf6 { [%eval 1.79] [%clk 0:01:25] } 9. Re1 { [%eval 2.29] [%clk 0:01:23] } 9... h4 { [%eval 1.99] [%clk 0:01:21] } 10. h3 { [%eval 3.11] [%clk 0:01:22] } 10... Nc4 { [%eval 2.14] [%clk 0:01:19] } 11. b3 { [%eval 3.36] [%clk 0:01:19] } ......```

Well, all looks good until you go to the Lichess link (u can get it from above PGN headers), and see this eval graph:
#

As you can see, in the API export, 1...d5 shows 0.13, and in the eval graph, 1...d5 shows 0.3. So, the problem appears to be that for move(i), the eval graph is one ahead of the API export data (what the api is showing as 1...d5, the eval graph is showing as 2.Ngf3.) I haven't looked into which is correct, but one of these is just giving completely useless data.

marble tree
#

Do you have any "normal" (non-variant) game like this? Maybe that could help.

shut hinge