Sharp, yet reasonably balanced opening lines requested.

[Dann Corbit]

Sample rows in the file look like this:

Code: Select all

r1bqkb1r/1p3ppp/p1nppn2/8/3NP3/2N1BP2/PPPQ2PP/R3KB1R b KQkq - acd 40; bm h5; ce -69; c4 "Abdalla, Luiz Guilherme Aurelli; pm=Be7; 0-1;"; c5 "Be7"; c6 "5784176"; c8 "So.0164393.008b.01937"; pm Be7 {591} Qc7 {77} Nxd4 {38} Bd7 {24} h5 {14} d5 {10} Ne5 {5} g6 {1} Nd7 {1} Rb8 {1}; pv h5; white_wins 3292; black_wins 1824; draws 1484; Opening B80h: Sicilian: Scheveningen, English, 7...Nc6 8.Qd2; id "ddots.0005";

Breakdown:
r1bqkb1r/1p3ppp/p1nppn2/8/3NP3/2N1BP2/PPPQ2PP/R3KB1R b KQkq - (the position)
acd 40 (depth of analysis was 40 plies)
bm h5 (computer analysis line started with h5)
ce -69 (score was -69 centipawns for the side to move)
c4 "Abdalla, Luiz Guilherme Aurelli; pm=Be7; 0-1;" (this is one player, who chose the move Be7, and won the game)
c5 "Be7" (Be7 was chosen in some study I was examining)
c6 "5784176" (this particular opening was also studied by Frank, who gave it this ID)
c8 "So.0164393.008b.01937" (Wesley So played this move. The annotation tells me exactly what move in exactly what game)
pm Be7 {591} Qc7 {77} Nxd4 {38} Bd7 {24} h5 {14} d5 {10} Ne5 {5} g6 {1} Nd7 {1} Rb8 {1} (these are moves commonly played for this position)
pv h5 (the pv emitted by computer analysis was h5 {probably a hash table hit})
white_wins 3292 (I have 3292 analyzed positions where white won from this position [*])
black_wins 1824 (I have 1824 analyzed positions where black won from this position [*])
draws 1484 (I have 1484 analyzed positions where there was a draw from this position [*])
Opening B80h: Sicilian: Scheveningen, English, 7...Nc6 8.Qd2 (this is the name of the opening)
id "ddots.0005" (this is the current ID of the opening. It is based on the name of DOTS which is created by a friend of mine, but I generated these).

[Dann Corbit]

[*] The white wins, black wins and draws are probably not what you think.
I have a unique index on a table that contains the position and the engine/player. So there will be a single row from Robert J. Fischer with e4 in it, instead of hundreds of them. Each time a different player is added who has played the position, a new row will get inserted. But subsequent identical positions with the same player name will be rejected. Even at that, there are about a billion rows in that table. So you can see why I need the index.

[Graham Banks]

Is it possible to get a PGN?

[Dann Corbit]

Here is a link to a PGN file with those positions:

[Graham Banks]

Here is a link to a PGN file with those positions:

Thanks, Dann. Much appreciated.

[abgursu]

I have another two, one you're maybe familiar with, one you're probably not.
This interesting line from the good old romantic theory. I say theory, because it has been played a lot in those times.
[pgn]1.e4 e5 2.f4 exf4 3.Nf3 g5 4.h4 g4 5.Ne5 Nf6 6.Nxg4 Nxe4 7.d3 Ng3 8.Bxf4 Nxh1[/pgn]
This great line with the following 9.Qe2+ Qe7 10.Nf6+ Kd8 11.Bxc7+ Kxc7 12.Nd5+ Kd8 13.Nxe7 Bxe7 line is really interesting, but drawish for engines. I remember Hikaru analyzing this Morphy - Anderssen game and got disappointed by how engines say 0.00 to this interesting position.

My other one is kinda rare, but in my deep analysis I find this crazy line quite playable, just like the one I published before:
[pgn]1.d4 f5 2.Bg5 h6 3.Bh4 g5 4.Bg3 Bg7 5.e3 c5 6.dxc5 Bxb2 7.Nd2 Bxa1 8. Qxa1 Nf6 9.Bd3[/pgn]
This line seems a quiet exchange sacrifice but things get really tricky. Dutch openings are really rich in gameplay.

Anyways, I have analysed this line and the other one I previously commented deeply in my lichess study, you're free to check it out if you're interested.
https://lichess.org/study/ffvUGEwT

[Graham Banks]

I have another two, one you're maybe familiar with, one you're probably not.

Thanks.

[Graham Banks]

Still keen to get more lines that can lead to interesting games.

[Rebel]

Still keen to get more lines that can lead to interesting games.

Hi Graham, if you are serious I can ask Jeroen Noomen.

[Graham Banks]

Still keen to get more lines that can lead to interesting games.

Hi Graham, if you are serious I can ask Jeroen Noomen.

Of course, but I think he's more invested in helping out TCEC.