Forums

reti gambit

Sort:
Cali_boy613
I recently found this relatively unknown but very intriguing gambit against the French defense.

It goes:
1.e4, e6 2. b3, d5 3. Bb2, dxe4


As a player who loves to attack (my repertoire includes the Dutch, kings gambit, Grand Prix, and I'm learning the dragon) I hate the French. It seems every variation is similar and relatively slow and boring. But this gambit is totally different and exciting. I've seen very few videos, articles or forum posts about it. I'd never even heard about it till I saw it on a top ten aggressive openings list (thanks to kingcrusher, the best chess YouTuber by far.) does anyone know if this is sound, playable or good (in your opinion)? I figured it must be unsound due to the lack of publicity, but the lines look pretty good for white. However, as a 1470 player, I'd like to get another opinion
Cali_boy613
Also, check out kingcrusher's videos out on YouTube. He's got over 5000 famous and current games and gives great commentary
Cali_boy613
Anyone?
strikingtan

White temporarily gives up a pawn in exchange for rapid development with Bb2, Nc3, Qd2, and O-O-O. White generally wins back the pawn.

JSLigon

You say King Crusher, I say King Scrusher. Anyway I've never played this but it's certainly a way to get a French Defense opponent out of book. The main line, if there is such a thing with a variation this rare, would be 1. e4 e6 2. b3 d5 3. Bb2 dxe4 4. Nc3 Nf6 5. Qe2 (5. g4 looks more adventurous). When I ran this line through Stockfish it also liked 3... a6 for black, which hasn't been played much at all but it was the choice of 2700 GM Vallejo Pons, so that might be worth a look. After 3... a6 Ultra Super GM Fish, Stock recommends 4. g3 going for the double fianchetto.

strikingtan
JSLigon wrote:

You say King Crusher, I say King Scrusher. Anyway I've never played this but it's certainly a way to get a French Defense opponent out of book. The main line, if there is such a thing with a variation this rare, would be 1. e4 e6 2. b3 d5 3. Bb2 dxe4 4. Nc3 Nf6 5. Qe2 (5. g4 looks more adventurous). When I ran this line through Stockfish it also liked 3... a6 for black, which hasn't been played much at all but it was the choice of 2700 GM Vallejo Pons, so that might be worth a look. After 3... a6 Ultra Super GM Fish, Stock recommends 4. g3 going for the double fianchetto.

Actually, 3... dxe4? is not correct in light of 3... d4, followed by 4... e5 (or 4... c5).

JSLigon

3... dxe4 is black's main move after 1. e4 e6 2. b3 d5 3. Bb2, not the only move available but I don't see how you can give it a question mark. I'm not qualified to argue for its correctness but it's the most common choice in that position.

strikingtan
[COMMENT DELETED]
strikingtan
JSLigon wrote:

3... dxe4 is black's main move after 1. e4 e6 2. b3 d5 3. Bb2, not the only move available but I don't see how you can give it a question mark. I'm not qualified to argue for its correctness but it's the most common choice in that position.


You can't play a move if you don't know why it's played.

 

Also, 3... d4 is played to close the position and inactivate white's fianchettoed bishop. 3...dxe4 does not work to accomplish this.

JSLigon

3... d4 doesn't work to accomplish that either. 4. Nf3 c5 5. c3 and the bishop isn't going to stay blocked.

strikingtan
JSLigon wrote:

3... d4 doesn't work to accomplish that either. 4. Nf3 c5 5. c3 and the bishop isn't going to stay blocked.

5... Nc6 6. Bb5 Bd7 7. Bxc6 Bxc6 8. cxd4 cxd4 and black gets the bishop pair with the initiative, e.g. 9. Nxd4 Bxe4 threatening ...Bxg2 or ...Bd3; or 9. Bxd4 Bxe4 threatening either ...Nxf3 followed by Qxd4 or ...Bd3.

chesster3145

7. O-O! Nf6 8. cxd4! Nxd4 9. Bxd7+ Qxd7 10. e5 Nd5 11. Bxd4 and White wins a pawn for insufficient compensation.

strikingtan
chesster3145 wrote:

7. O-O! Nf6 8. cxd4! Nxd4 9. Bxd7+ Qxd7 10. e5 Nd5 11. Bxd4 and White wins a pawn for insufficient compensation.

You are correct.

JSLigon

I've been running these lines through Stockfish, interested in maybe playing this as white. 3... d4 is at least a reasonable try for black. After 7. O-O Nge7 (instead of Nf6) it's saying white has no advantage.

 

1. e4 e6 2. b3 d5 3. Bb2 d4 4. Nf3 c5 5. c3 Nc6 6. Bb5 Bd7 7. O‑O Nge7

8. Bc4 dxc3 9. dxc3 Ng6 10. Na3 Be7 11. Be2 O-O 12. g3 a6 13. Nd2 b5 14. f4 Qb6 15. Qe1 Bc8 16. Qf2 Rd8 17. Rfd1 Ra7 18. Nf3 Rad7 19. Rxd7 Rxd7 20. Nc2 Bb7 21. Rd1 Rxd1+ 22. Bxd1 Qd8 23. Qd2 Qb6 24. Qf2
= (0.00) Depth=42/68 0:134:17 40349 MN

8. Na3 Ng6 9. Nc4 dxc3 10. Bxc3 f6 11. e5 Ncxe5 12. Bxd7+ Nxd7 13. Qe2 e5 14. d4 cxd4 15. Nxd4 Nf4 16. Qf3 Nc5 17. Rae1 Ncd3 18. Nb5 Qd7 19. a4 Be7 20. Bxe5 fxe5 21. Nxe5 Nxe5 22. Qxf4 Ng6 23. Nd6+ Kd8 24. Nf7+ Ke8 25. Nd6+ Kd8
= (0.00) Depth=42/68 0:134:17 40349 MN

8. cxd4 cxd4 9. Na3 Ng6 10. Nc4 Rc8 11. Rc1 Bc5 12. b4 Bxb4 13. Qb3 Be7 14. Bxc6 Bxc6 15. Nxd4 Bxe4 16. Rfe1 Bd5 17. Qa4+ Kf8 18. Ne3 Rxc1 19. Rxc1 f6 20. Qxa7 Kf7 21. g3 e5 22. Ndf5 Be4 23. Qa4 Qd3 24. Rc3 Qb1+ 25. Rc1 Qd3
= (0.00) Depth=42/68 0:134:17 40349 MN

 

7. Qe2 looks more promising however.

Cherub_Enjel

I wouldn't recommend using an Engine to analyze the Reti Gambit. Not one of those openings. 

JSLigon

I know that's a common point of view, and it's held by plenty of strong players, but I don't understand it. Every opening position is also a chess position, and engines play chess way better than people do. And in this case we're dealing with a rare enough opening that there isn't much existing theory to fall back on. I'd trust engine analysis over anything else.

Cherub_Enjel

The idea is twofold, mainly:

 

(1) Engines don't care about the playability of the position. More precisely, this means that engines don't have an evaluation factor that consider how many playable moves there are in the tree of variations, but only consider the best possible continuation for one side. This is the same reason why no one recommends the Najdorf to 1600s. 

 

(2) Humans can play better moves than engines in certain positions, especially ones dealing with clarification of pawn structure. In computer-assisted correspondence chess, a beginner spitting out engine moves will lose against a master using an engine, which proves that human ability is not fully encapsulated by engine ability. 

 

The first reason one alone should be more than enough to convince anyone who's tried to play a computer-generated line in a real game. 

 

There are more reasons, but these are the main ones. You don't expect many inexperienced players to come up with either of them on their own, though. 

Cherub_Enjel

On the other hand, I would trust engine analysis in very sharp openings - in fact, engine analysis is absolutely necessary in those lines. 

But someone playing the Closed Ruy wouldn't really care if his/her opponents had engine preparation against it (lol). 

JSLigon

The first reason I get as a practical objection to playing engine lines, although it's not an argument against the validity of the engine's analysis. But I might end up with something that's way above my ability to play it, whereas my opponent's side will be much easier for them to play. Meaning that regardless of its objective merit, at my level the engine line is simply no good. And where engine lines are concerned the line might not even be practical for human grandmasters to attempt. It's like trying to fly an alien spacecraft. It might be technologically superior to any of our ships, but if I try to fly it I'm probably going to get myself killed. Ok, that's something to watch out for with engine lines. Or with super grandmaster lines for average club players. I don't think it would be the best idea for me to try to play the Najdorf either.

As for the second one, I'm aware of positions that engines famously don't play correctly, but I tend to think these positions are anomolies. And also that as engines improve and computers continue to get faster, the already small proportion of such positions is going to continue to shrink. One of the main things I've noticed from having Stockfish analyze common positions in opening theory, is that initially it may want to make an unusual move that theory considers dubious. But very often it will switch over to what is considered the main move if I let it run for long enough. Or if I put in the first (considered dubious) move along with the opponent's most common response, and have Stockfish analyze that, its evaluation will start to show that the dubious move isn't so good after all.

JSLigon

And also I'm not just interested in engine analysis for my own openings. It fits in with a project I'm working on.