1.g4 opening is losing?

[Ovyron]

Black's advantage is much higher than that, I think some 1.60 would be unstoppable.

Another interesting question is: if 1.g4 is lost by force, is 1...d5 the only winning move? Not that it's hard to find, but I find really weird that 1.g4 leads to a position where almost everything draws, and only one move wins, just like the opening position has all moves drawing, and only one losing. That's so weird that I think 1.g4 being able to be saved (if we had 32men TBs) had to be more likely. Or perhaps 1...Nc6 beats it as well.

[mmt]

Black's advantage is much higher than that, I think some 1.60 would be unstoppable.

According to? Not according to LC0 with a large net after 10+ hours of searching with 2080Ti and TitanV. That seems to be the longest search that's been posted so far. Based on that, it's doubtful that 1. g4 is losing.

If you'd like to, I can play it against your preferred lines to see if this is incorrect.

[Ovyron]

Not according to LC0 with a large net after 10+ hours of searching with 2080Ti and TitanV.

I spent 3 weeks analyzing a single position of one of the variations of the 1.g4 d5 2.g5 mainline. How much is that, 504 hours? They weren't 24/7 but they were interactive so they may be equivalent to 1.5 weeks, but still...

If you'd like to, I can play it against your preferred lines to see if this is incorrect.

Challenge accepted! I'll create our game soon...

[mmt]

I spent 3 weeks analyzing a single position of one of the variations of the 1.g4 d5 2.g5 mainline. How much is that, 504 hours? They weren't 24/7 but they were interactive so they may be equivalent to 1.5 weeks, but still...

LC0 doesn't play 2. g5 though. I can have it look at this position. From a quick 15M nodes look, it seems to prefer 2. h3 and 2. c4 to 2. g5 also.

Challenge accepted! I'll create our game soon...

Cool!

[BrendanJNorman]

When I played 30 games between SF10 and SFdev 16 games were 0-1 (here 10 first). Other games were draws. Does this indicate, that 1.g4 is losing move?

Easy question.

Since chess hasn't been solved, no first move is losing.

This really is common sense, and no engine games are really needed to stumble upon this conclusion.

That being said, 1.g4 sucks, and that's why we never see it played by anyone serious.

It violates basic opening principles and basically says "Hey black...wanna play white? I'm happy to pass the ball back because there's a trick I'm hoping for in the 1.g4 d5 line! I'm a bit of a gambler ya know?"

Back in 2000-2001 I knew a young guy (I was young too at the time) who was like 1300 Elo and he stubbornly played 1.g4 and whenever I'd suggest he develop a real opening repertoire, he'd tell me some story about how he "beat someone with his Grob at so and so club, and they had no idea what to do" or whatever...

Some excuses...

He's still 1300 Elo.

I actually think Grob players are a cult.

[Uri Blass]

When I played 30 games between SF10 and SFdev 16 games were 0-1 (here 10 first). Other games were draws. Does this indicate, that 1.g4 is losing move?

Easy question.

Since chess hasn't been solved, no first move is losing.

This really is common sense, and no engine games are really needed to stumble upon this conclusion.

That being said, 1.g4 sucks, and that's why we never see it played by anyone serious.

It violates basic opening principles and basically says "Hey black...wanna play white? I'm happy to pass the ball back because there's a trick I'm hoping for in the 1.g4 d5 line! I'm a bit of a gambler ya know?"

Back in 2000-2001 I knew a young guy (I was young too at the time) who was like 1300 Elo and he stubbornly played 1.g4 and whenever I'd suggest he develop a real opening repertoire, he'd tell me some story about how he "beat someone with his Grob at so and so club, and they had no idea what to do" or whatever...

Some excuses...

He's still 1300 Elo.

I actually think Grob players are a cult.

He is 1300 mainly not because of the first move.
I am sure strong players can easily get a lot more than 1300 and always play 1.g4 with white.

playing g4 is going to cause worse results relative to playing a normal opening but
I believe the main improvement this 1300 player can get is not by learning opening but by being better in tactics.

[Ovyron]

Challenge accepted! I'll create our game soon...

Cool!

Here's our game thread, good luck!

[mmt]

[d]1rq1r1k1/p3bpp1/2n2n1p/2Np3P/3P3R/BP1Q4/P2NPP2/R3K3 w Q - 5 24
LC0 gets itself in trouble here. It plays 24. Rd1 because it doesn't find the response 24... Ng4! in a reasonable time, while Stockfish 11 sees it instantly.

Leads to a nice sacrifice
[d]1rq1r1k1/p3bpp1/2n4p/2Np3P/3P2n1/BP1Q4/P2NPP2/3RK2R b - - 8 25
25.... Rb4!

This is a position resulting from playing out 1.g4 d5 2.Bg2

[BrendanJNorman]

He is 1300 mainly not because of the first move.
I am sure strong players can easily get a lot more than 1300 and always play 1.g4 with white.

He is 1300 because of his bull-headedness, laziness and perhaps lack of talent.

Bull-headed because he ignored a player 600 Elo higher when I told him to play 1.e4 and study the classics from Morphy, Capablanca and Fischer if he wants to improve his chess understanding.

Instead, he studied 1.g4 with Fritz 6 and used it to cheat online to "test his lines".

If a player 600 Elo higher than me (so a top ten player in the world) offered to give me advice, I'd be leaping for my notepad and pen.

But I guess this "learning attitude" is why I gained hundreds of Elo since then and he is still 1300 on chess.com.

I am sure strong players can easily get a lot more than 1300

What does this mean? By DEFINITION "strong" players are already higher than 1300.

playing g4 is going to cause worse results relative to playing a normal opening but
I believe the main improvement this 1300 player can get is not by learning opening but by being better in tactics.

A 1300 a player should absolutely IGNORE openings study. Yes, they should study tactics daily, but also the study of classics (as mentioned above) will build over time an overall understanding of openings, middlegame strategy, attacking formations, tactical motifs AND endgames.

Doing this took me from unrated to 1550 in my first year of playing chess (and 8/11 in my first tournament), then >1603 then >1730 and so on to 1980 , once I started playing training games against ChessMaster 3000 (got it from a garage sale )

BTW...I'm saying this as a coach who has taken a 1300 guy (who WAS listening and following all instructions) from 1300 to National Master (around 2100 Elo) in 18 months. I know what works.

[Alayan]

Looking at the tree in the Bg2 line, one line that ChessDBCN liked (and matching sergio's net choices until 12. ... Bd6) goes :

Code: Select all

1. g4 d5 2. Bg2 Bxg4 3. c4 c6 4. Qb3 e6
5. Qxb7 Nd7 6. Nc3 Ne7 7. cxd5 exd5 8. d4 Rb8
9. Qa6 Rb6 10. Qd3 Ng6 11. h3 Be6 12. Nf3 Bd6
13. h4 h5 14. Na4 Rb4 15. b3 O-O 16. Bg5 Qc8
17. Rc1 Rb8 18. O-O Re8 19. Qc3 Ba3 20. Qxc6 Bxc1
21. Rxc1 Qxc6 22. Rxc6

FEN : 1r2r1k1/p2n1pp1/2R1b1n1/3p2Bp/N2P3P/1P3N2/P3PPB1/6K1 b - - 0 22

But this end position is lost. Each line from there leads to ever-increasing black scores.

So, white needs to deviate from this line somewhere. ChessDBCN hasn't found something convincing yet, its score after 7. ... Ne7 has moved to -1.30 after I fed it the refutations to its preferred line.