The Latvian Gambit

[16thompsong]

I was having a brainstorm one day about opening, and I considered looking at openings against 1.e4. Being a Sicilian player, I didn't those lines. However, I don't like the French or Caro-Kann. So after (briefly) looking at the Alekhine, I turned to the open games. I didn't like the idea of playing 2... Nc6 because of somebody playing some random opening that I don't feel like looking at. So I finally found a great opening which, I bet you can guess it, is

THE LATVIAN GAMBIT!

Immediately I discounted such sources as the internet, because every single person seems to say it's bad. Even I thought it sucked a week ago. I also jumped over such opening books as SCO and WCO both by Eric Schiller. They irritate me a lot. So I turned to MCO. What I (surprisingly) found was that Black was better in every line except for 3. Nxe5. These lines are the lines given by MCO that I prefer. However, I am playing a side line against 3. Nxe5, as I believe Black's horrible in the main line.

8

7

6

5

4

3

2

1

abcdefgh

( ? ) vs ( ? )

Date:

Event:

Site:

Round: ?

Result:

1. e4 e5 2. Nf3 f5 3. Bc4 ( 3. exf5 e4 4. Ne5 Nf6 5. Be2 d6 6. Bh5+ Ke7 7. Nf7 Qe8 8. Nxh8 ( 8. Nc3 g6 9. fxg6 ( 9. Nxh8 gxh5 ) 9... hxg6 10. Nxh8 gxh5 ) 8... Qxh5 9. Qxh5 Nxh5 10. g4 Nf6 11. Nc3 ( 11. Rg1 Nc6 12. Rg3 Nd4 13. Kd1 g6 ) 11... Nxg4 12. Nd5+ ) ( 3. d4 fxe4 4. Nxe5 Nf6 5. Bg5 d6 6. Nc3 ( 6. Nc4 Be7 7. Ne3 c6 8. Qd2 d5 9. c4 O-O 10. Be2 Be6 ) 6... dxe5 7. dxe5 Qxd1+ 8. Rxd1 h6 9. Bxf6 gxf6 10. Nd5 Kd7 11. Nb6+ Kc6 12. Nxa8 fxe5 13. Bc4 Bd6 ) ( 3. Nxe5 Nc6 4. Nxc6 ( 4. Qh5+ g6 5. Nxg6 Nf6 6. Qh3 hxg6 7. Qxh8 Qe7 ) 4... dxc6 5. Nc3 Bc5 6. d3 Nf6 ) 3...  fxe4 4. Nxe5 d5 5. Qh5+ g6 6. Nxg6 hxg6 7. Qxg6+ ( 7. Qxh8 Kf7 8. Qd4 Be6 9. Be2 Nc6 ) 7...  Kd7 8. Bxd5 Nf6 9. Nc3 Qe7 10. d3 exd3+

1. e4 e5 2. Nf3 f5 3. Bc4   ( 3. exf5 e4 4. Ne5 Nf6 5. Be2 d6 6. Bh5+ Ke7 7. Nf7 Qe8 8. Nxh8   ( 8. Nc3 g6 9. fxg6   ( 9. Nxh8 gxh5  )  hxg6 10. Nxh8 gxh5  ) Qxh5 9. Qxh5 Nxh5 10. g4 Nf6 11. Nc3   ( 11. Rg1 Nc6 12. Rg3 Nd4 13. Kd1 g6  )  Nxg4 12. Nd5+  )   ( 3. d4 fxe4 4. Nxe5 Nf6 5. Bg5 d6 6. Nc3   ( 6. Nc4 Be7 7. Ne3 c6 8. Qd2 d5 9. c4 O-O 10. Be2 Be6  )  dxe5 7. dxe5 Qxd1+ 8. Rxd1 h6 9. Bxf6 gxf6 10. Nd5 Kd7 11. Nb6+ Kc6 12. Nxa8 fxe5 13. Bc4 Bd6  )   ( 3. Nxe5 Nc6 4. Nxc6   ( 4. Qh5+ g6 5. Nxg6 Nf6 6. Qh3 hxg6 7. Qxh8 Qe7  )  4... dxc6 5. Nc3 Bc5 6. d3 Nf6  ) 3... fxe4 4. Nxe5 d5 5. Qh5+ g6 6. Nxg6 hxg6 7. Qxg6+   ( 7. Qxh8 Kf7 8. Qd4 Be6 9. Be2 Nc6  )  Kd7 8. Bxd5 Nf6 9. Nc3 Qe7 10. d3 exd3+ 

I apologize for the 'un' user-friendly text that it's in, but that's generally how I think.

[rombelstielz]

by Eric Schiller. They irritate me a lot

Oh yeah!
The Sdtielz has seen half a dozen books by Schiller and they irritate him too.

[munich]

A pity you spent already some time for analyzing the latvian gambit. You wont get happy with it after a while (=on the long run), because it isnt so sound. Then, some day, you will need to look out for s.th. else again.

I see this trouble for you coming:

8

7

6

5

4

3

2

1

abcdefgh

( ? ) vs ( ? )

Date:

Event:

Site:

Round: ?

Result:

1. e4 e5 2. Nf3 f5 3. Nxe5 this is by far the most common response according to the chesstempo database. It might not be the case in Blitz, but OTB I believe a lot of people will come to the conclusion to take this pawn, even if there intentional thoughts might be wrong, it will often lead to make the white player play the move Nxe5 nevertheless. 3...  Nc6 your move 4. Nxc6 dxc6 black is a pawn down, and I can not see any weakness in white's position. Sure, black is a tempo up, because he does not need to clear the way for a bishop, whereas white needs to spend one pawn move to free his Bc1. But I dont think being one tempo ahead with no weaknesses of the opponent is enough compensation for the pawn. 5. Nc3 I am not sure why white would want to move this knight, as the black c6 pawn guards the squares b5 and d5. So a knight on c3 does not make so much sense here. Rather I would have the idea to play at some point Nb1-d2-f3, and I wont be the only one who will see that a knight on c3 has not future with a black pawn on c6. ( Instead: 5. d4 seems simple and good, but probably black will have not much of a game after 5.d3, too. Remember: black is a pawn down here! But before you find a good response to 5.d3, solve the trouble with the natural move 5.d4 first, as I believe this will be moved often here, especially if white has no theory knowledge about the latvian gambit. )

1.e4 e5 2.Nf3 f5 3.Nxe5 {this is by far the most common response according to the chesstempo database. It might not be the case in Blitz, but OTB I believe a lot of people will come to the conclusion to take this pawn, even if there intentional thoughts might be wrong, it will often lead to make the white player play the move Nxe5 nevertheless.} 3...Nc6 {your move} 4.Nxc6 dxc6 {black is a pawn down, and I can not see any weakness in white's position. Sure, black is a tempo up, because he does not need to clear the way for a bishop, whereas white needs to spend one pawn move to free his Bc1. But I dont think being one tempo ahead with no weaknesses of the opponent is enough compensation for the pawn.} 5.Nc3 {I am not sure why white would want to move this knight, as the black c6 pawn guards the squares b5 and d5. So a knight on c3 does not make so much sense here. Rather I would have the idea to play at some point Nb1-d2-f3, and I wont be the only one who will see that a knight on c3 has not future with a black pawn on c6. } ({Instead:} 5.d4 {seems simple and good, but probably black will have not much of a game after 5.d3, too. Remember: black is a pawn down here! But before you find a good response to 5.d3, solve the trouble with the natural move 5.d4 first, as I believe this will be moved often here, especially if white has no theory knowledge about the latvian gambit.})

You dont like 2...Nc6:

I didn't like the idea of playing 2... Nc6

So how about 1...Nc6 as an opening move. That is what I came up with.
You can look at the thread I opened for this move, but here I give you this so called "Nimzowitsch Defense" in a nutshell:

8

7

6

5

4

3

2

1

abcdefgh

( ? ) vs ( ? )

Date:

Event:

Site:

Round: ?

Result:

1. e4 Nc6 2. d4 d5 3. exd5 Qxd5 the white pawn is now attacked twice 4. Nf3 most common move 4...  Bg4 5. Be2 Nf6 6. Nc3 Qh5 7. h3 O-O-O 8. O-O ( 8. hxg4? Qxh1 the h-pawn is pinned! ) 8...  e5 9. d5 ( 9. hxg4 Nxg4 and now the next task of black is to distract the Nf3, so black can play Qh2#. I won many game like this, but there are some variations, which are too long and many to give in a nutshell (see the main Nimzo-thread for this). At the end it can be assessed that white should not accept the sac with 9.hxg4 ) 9...  e4 the d-pawn is pinned, too! White is now in some troubles, the initiative is clearly in black's hands, and it didnt cost you a pawn. Actually, white defended all the time as soon as with move 4.Nf3 which defends already the d4-pawn

1.e4 Nc6 2.d4 d5 3.exd5 Qxd5 {the white pawn is now attacked twice} 4.Nf3 {most common move} 4...Bg4 5.Be2 Nf6 6.Nc3 Qh5 7.h3 O-O-O 8.O-O (8.hxg4? Qxh1{the h-pawn is pinned!}) 8...e5 9.d5 (9.hxg4 Nxg4 {and now the next task of black is to distract the Nf3, so black can play Qh2#. I won many game like this, but there are some variations, which are too long and many to give in a nutshell (see the main Nimzo-thread for this). At the end it can be assessed that white should not accept the sac with 9.hxg4}) 9...e4 {the d-pawn is pinned, too! White is now in some troubles, the initiative is clearly in black's hands, and it didnt cost you a pawn. Actually, white defended all the time as soon as with move 4.Nf3 which defends already the d4-pawn}

Of course white can play differently:

8

7

6

5

4

3

2

1

abcdefgh

( ? ) vs ( ? )

Date:

Event:

Site:

Round: ?

Result:

1. e4 Nc6 2. d4 d5 3. e5 but here you have an advantage in contrast to a french: your Bc8 is not locked in. Also you have not spend a tempo with c7-c6 like in the caro-kann, but instead developed Nc6. 3...  f6 4. f4 Nh6 5. Nf3 Bf5 6. Bd3 e6 7. c3 Qd7 8. O-O Be7 9. b4 O-O black has an equal game, here. The queens-wing expansion is nothing to worry much about, and I often do the knight-manoever Nc6-e7-g6 or Nc6-e7-f5 or Nc6-d8-f7 and by doing so I have one more piece available on the kings-wing, while usually white doesnt achieve much on the queenswing (The reason: the black pawns are just too flexible there and blunt white's pawn pushs efficiently enough to keep white at bay for a long time. The king's wing attack on the other hand get's a bit faster into swing. A program can hold the position balanced, but I believe that practically seen (for a human at least) the task to achieve s.th. is simpler for black than for white (which usually results in a huge time advantage for black). With overpowering white on the kingswing you do have statistically pretty good chances in the tactical struggles. I dont think black is standing better, but white is not standing better, too. And: you are not a pawn down for having a wild gambit style game. Furthermore, you can be sure white has zero experience in this position, while your experience will grow over time. Look at the position, and you will understand why I think black has good statistics here: black's pieces are all developed. One of the "crack-up-moves" (or "pry-open-moves") is also already achieved: f6xe5 at some point, whereas white needs yet to show to how to achieve such a pry-open-move. There is some space advantage for white, but the real good squares are already reached by black - so what is this space advantage good for? Chances are that black can make good use of the space behind white's pawns.

1.e4 Nc6 2.d4 d5 3.e5 {but here you have an advantage in contrast to a french: your Bc8 is not locked in. Also you have not spend a tempo with c7-c6 like in the caro-kann, but instead developed Nc6.} 3...f6 4.f4 Nh6 5.Nf3 Bf5 6.Bd3 e6 7.c3 Qd7 8.O-O Be7 9.b4 O-O {black has an equal game, here. The queens-wing expansion is nothing to worry much about, and I often  do the knight-manoever Nc6-e7-g6 or Nc6-e7-f5 or Nc6-d8-f7 and by doing so I have one more piece available on the kings-wing, while usually white doesnt achieve much on the queenswing (The reason: the black pawns are just too flexible there and blunt white's pawn pushs efficiently enough to keep white at bay for a long time. The king's wing attack on the other hand get's a bit faster into swing. A program can hold the position balanced, but I believe that practically seen (for a human at least)  the task to achieve s.th. is simpler for black than for white (which usually results in a huge time advantage for black). With overpowering white on the kingswing you do have statistically pretty good chances in the tactical struggles. I dont think black is standing better, but white is not standing better, too. And: you are not a pawn down for having a wild gambit style game. Furthermore, you can be sure white has zero experience in this position, while your experience will grow over time. Look at the position, and you will understand why I think black has good statistics here: black's pieces are all developed. One of the "crack-up-moves" (or "pry-open-moves") is also already achieved: f6xe5 at some point, whereas white needs yet to show to how to achieve such a pry-open-move. There is some space advantage for white, but the real good squares are already reached by black - so what is this space advantage good for? Chances are that black can make good use of the space behind white's pawns.}

8

7

6

5

4

3

2

1

abcdefgh

( ? ) vs ( ? )

Date:

Event:

Site:

Round: ?

Result:

1. e4 Nc6 2. Nf3 that is played often, in the hope that black plays 2...e5. I admit I had trouble in finding s.th. good against this move, but finally I understood, that with the move 2.Nf3 white actually blocked his f-pawn. Thinking hard, I came to the conclusion, that the pirc and "The Modern" is suffering from opening-systems like the "Austrian attack" or the "Argentinian attack", where white goes f2-f4 or f2-f3 and throw his king side pawns against blacks short castle. So with the option gone of agressive f2-f3, or f2-f4 systems, I believe black has an excellent starting position to flank his bishop without worrying to get overrun: 2...  d6! 3. d4 Nf6 4. Nc3 g6! 5. d5 looks strong, but is not! 5...  Nb8 black will have a good bishop on g7, and the tempo loss Nb8 was worth it: white blocked the position (he cant jump Nc3-d5 anymore) and in blocked positions, a tempo loss isnt that important anymore. Instead it is more important that you will have good squares for your pieces. And you will have: Nb8-d7-c5 (together with a7-a5, so white can not harras the Nc5 with the pawn move b2-b4) AND a nice bishop on g7. You dont need to fear a severe attack, too (c-pawn and f-pawn are currently both blocked for white with the knights Nc3 and Nf3 in the way). Simply finish your development, and then you have a sufficient game. Nothing really exting for gambit freaks, but this is just the worst-case scenario, where white plays 2.Nf3 instead of 2.d4

1.e4 Nc6 2.Nf3 {that is played often, in the hope that black plays 2...e5. I admit I had trouble in finding s.th. good against this move, but finally I understood, that with the move 2.Nf3 white actually blocked his f-pawn. Thinking hard, I came to the conclusion, that the pirc and "The Modern" is suffering from opening-systems like the "Austrian attack" or the "Argentinian attack", where white goes f2-f4 or f2-f3 and throw his king side pawns against blacks short castle. So with the option gone of agressive f2-f3, or f2-f4 systems, I believe black has an excellent starting position to flank his bishop without worrying to get overrun:} 2...d6! 3.d4 Nf6 4.Nc3 g6! 5.d5 {looks strong, but is not!} 5...Nb8 {black will have a good bishop on g7, and the tempo loss Nb8 was worth it: white blocked the position (he cant jump Nc3-d5 anymore) and in blocked positions, a tempo loss isnt that important anymore. Instead it is more important that you will have good squares for your pieces. And you will have: Nb8-d7-c5 (together with a7-a5, so white can not harras the Nc5 with the pawn move b2-b4) AND a nice bishop on g7. You dont need to fear a severe attack, too (c-pawn and f-pawn are currently both blocked for white with the knights Nc3 and Nf3 in the way). Simply finish your development, and then you have a sufficient game. Nothing really exting for gambit freaks, but this is just the worst-case scenario, where white plays 2.Nf3 instead of 2.d4}

Last one:

8

7

6

5

4

3

2

1

abcdefgh

( ? ) vs ( ? )

Date:

Event:

Site:

Round: ?

Result:

1. e4 Nc6 2. d4 d5 3. Nc3 it does occasionally happen that white choses this knight move first, but it is not as common as the other 3 replies 3...  e5 my latest discovery, absolutely seldom, but sound.

1.e4 Nc6 2.d4 d5 3.Nc3 {it does occasionally happen that white choses this knight move first, but it is not as common as the other 3 replies} 3...e5 {my latest discovery, absolutely seldom, but sound.}

All this as a teaser. If you want deeper detail, I have written lots of guidance in my Nimzowitsch thread.

Errr, well yes, this is the latvian gambit thread. Sorry for the off-topic. I leave now as it is. This was more a reply for 16thompsong trouble. You may discuss the latvian gambit further, of course!

[16thompsong]

I saw that issue with a move such as d2-d4... I looked in Chess openings for White Explained, and that is the recommended antidote against my move order. I still have no answer. On a note, Black doesn't get equality in any of the Nxe5 lines. If it wasn't like that, The Latvian would be the refutation to 2. Nf3 because Black's better in every other line!

[16thompsong]

Well, now I shall try to answer the problem of 3. Nxe5.

8

7

6

5

4

3

2

1

abcdefgh

( ) vs ( )

Date:

Event:

Site:

Round:

Result:

 

Never mind. I found nothing. Against 5. d3, I had the idea of 5... Bc5, but 5. d4 ruins it. That's irritating.

[munich]

The main trouble in the Nxe5 line is not to find a playable variation.
The main trouble is, that it is structurally seen suspect:

You weakened your king with the f-pawn move. White players might not have seen the latvian gambit, but the move Nxe5 will free the diagonal for the queen d1-h5. It does not matter if that is at the end playable or not. It is the result that matters: most players will intuetively play Nxe5 with or without understanding.

The move NxNc6 is also very natural. White is a pawn up - he likes to trade pieces. And double pawns are supposed to be bad.
So knowing theory or not - the Ne5 is attacked, and instead of moving back to f3 (which would block the Qd1-h5 diagonal) there is this obvious solution for white: trade the knights!
And indeed, it is (co-incidently) again the strongest move.

Now back to the structural problem black has: the f-pawn isnt that useful, and the lead in development is only one tempo move (it is actually two, but black is not white, and black starts with a tempo down in the first place, so now he is only "white" with one tempo move gained, only. This is by rule of thumb never enough for a pawn. 3 tempo moves are considered to be the equivalent of a pawn.)
White has no weaknesses. Black has no center pawn. This is a problem, because without structural support from pawns, the attack is a bit difficult to get going. No doubt black will have good squares for his pieces. But then? There is not much potential of tearing white's position appart, because there is no center pawn that you could throw against white's position. All there is, is this bloody f5-pawn, and that is a bit "thin".

In practice you will face the position after 4.NxNc6 very often, and after that you may have analyzed heavily with a computer program and will have found a good sequence. However, realistically white wont play the best moves in the next moves, and thus you will be out of book, too. What is left is a position for black with no white weakness, and black is a pawn down.
If the position is positionally not sound for black, I would not like to rely on some shaky variations that "maybe" work, but dont come up. What is left is a positionally inferior position.
I believe the Latvian gambit is losing a pawn for nothing. There are risky openings for black, where black does get severe counterplay but does not need to be a pawn down.
The trick is not in finding a gambit. The real trick is in finding a gambit, that is not a gambit, because you received plenty of play without sacrificing a pawn.

When you start learning an unsound opening, you will hate it after a year. Then you will look for something else. So best is: start with a reasonable opening right from the beginning. Then any problems can get fixed. You will have frustrating outcomes, but with a structural sound opening you will always be able to find a suitable variation that is still acceptable for you. And thus you dont need to throw everything you learned "over board", but can stick to your opening and grow with it.
Hopping between openings is one problem chess players have. At the end they wont have deep knowledge of anything, and not surprisingly they wont have a lot of success with openings on the long run. Hours and hours are wasted, just to skip the opening later once again. So my advice:
Look for an opening with good statistics to beginn with. Some variations might be not so good, but the good statistics will remain, and you will always be able to revive an unloved variation.
But the latvian gambit can not repaired. It is interesting, but not worth the effort on the long run. If you fall into the trap in falling in love with it now, you will be sorry in a year to come, after which you will start with s.th. completely different once again.

And off-topic again: look at the chesstempo database: what are the statistics for black in this opening? 1.e4 1...Nc6!
The move 1...Nc6 scores 34.2% wins for black in the "all players universe". That is a very high winning percentage for a black opening. However, it does not come with a disadvantage: white scores 40.4% wins, also slightly above average.

So it is 40.4% white wins vs 34.2% black wins. (40.4% - 34.2% = 6.2% difference)
Compare this with 1.e4 e5:
Here it is 39.8% white wins vs 28.1% black wins. (39.8% - 28.1% = 11.7% difference)

The best opening is probably the sicilian with:
36.6% white wins vs 33.7% black wins. (36.6% - 33.7% = 2.9% difference).

However, the sicilian is one of the most frequently played openings, whereas if you look for seldom variations (like the latvian) to have an opening advantage, then the pretty seldomly played 1...Nc6 gives you a very good statistic for a black opening. If you learn it well, it is really rewarding. A club mate of mine complained ones: why do you always get so nice positions with black? We have the same pawn structures with black, but your pieces are active, mine are passive. (he plays caro can, french and aljechin, all openings push the black d-pawn. So does the nimzowitsch defense. That explains why the pawns structure is similar. The active position of my Nimzowitsch defense in contrast to the other d5-push openings is structural: black developed a piece instead of moving a pawn. In the aljechin the knight on b6 stands structurally passive (=bad). In the french the Bc8 is locked in. In the caro black is behind in development. So naturally - it is no surprice the move 1...Nc6 is the best in the category of all d5-push opening. That has nothing to do with a certain variation - it is the health of the position that gives its favorable statistics. And now look at the latvian gambit  - the statistics are scary! 52.7% vs 30.8% (= 52.7% - 30.8% = 21.9% difference). If both players are out of book, black has a lost position, structurally seen. There might be tactics here and there (that explains the high winning percentages on both sides) - but at the end, by tendency the outcome will favor white. 

[16thompsong]

When you start learning an unsound opening, you will hate it after a year. Then you will look for something else. So best is: start with a reasonable opening right from the beginning. Then any problems can get fixed. You will have frustrating outcomes, but with a structural sound opening you will always be able to find a suitable variation that is still acceptable for you. And thus you dont need to throw everything you learned "over board", but can stick to your opening and grow with it.
Hopping between openings is one problem chess players have. At the end they wont have deep knowledge of anything, and not surprisingly they wont have a lot of success with openings on the long run. Hours and hours are wasted, just to skip the opening later once again. So my advice:
Look for an opening with good statistics to beginn with. Some variations might be not so good, but the good statistics will remain, and you will always be able to revive an unloved variation.
But the latvian gambit can not repaired. It is interesting, but not worth the effort on the long run. If you fall into the trap in falling in love with it now, you will be sorry in a year to come, after which you will start with s.th. completely different once again.

I play a sound, solid opening. If you notice, I never said I dropped an opening for the Latvian. I still rely on the Sveshnikov as my main defense against 1. e4. The idea with the Latvian was to play it against a specific opponent who would have played 3. exf5.

[perkovsky]

Which opening against d4 you think is sound ? I did not like  Benoni , I spent hours studying  Grunfeld but I am not happy with that , tried Slav-Semislav but not happy with Slav , I am thinking going back to Nimzo-Indian , Bogo-Indian , but not yet sure. I have not tried Kings-Indian. I like somewhat unbalanced position , but against stronger opponents  that may not be wise.
Against  e4 I try  to learn Najdorf Sicilian. As White I always open  e4.I am most comfortable against  Ruy-Lopez , Sicilian especially Najdorf ,not very comfortable against Taimanov-Paulsen , Kan,moderately comfortable against French and least comfortable against Scandinavian and  Pirc. Possibly because I have studied  least the last two.
Sometimes I feel better with Black pieces.
I would want to hear  opinion of others.

[munich]

1.d4 is not so much of problem to me than 1.Nf3 (as a result I started playing 1.Nf3 myself, but the problem I have with black against it remains).

Against 1.d4 you have at least some counterplay, especially if white goes 2.c4.

There is always the option to play 1...g6. But I admit the trouble is, that you have to learn quite a lot ofter 1...g6.
There are 3 pawn attacks with d4/e4/f4 to be learned, as well as d4/e4/f3 attacks, and then there are calmer d4/c4/Nf3 or stonewall like d4/e3/f4 systems possible.

"My way" is 1...d6, which is the attempt of an improved 1...g6 move order:
 
If white plays 2.c4 then I play 2.e5. It is good to know that white has no advantage after 3.dxe5 dxe5 4.Qxd8 Kxd8, but only gets himself into trouble on the long run, which is why black's statistics are so overwhelming good here.
However, white can of course play 2.Nf3, but then I still have my 2...g6. In that way I could get away quite economically, because I dont need to learn f2-f4 or f2-f3 systems against 1...g6.
So my move 1...d6 is a different move order to get into 1...g6, but with the advantage of white having blocked his f-pawn with the knight after 1.d4 d6 2.Nf3 g6: I reduced the amount to be learned considerably, while still enjoying very good statistics which the move 1...g6 indicates.

Still, it is a lot to learn. I found that 1.d4 d6 2.c4 e5 3.e3 can be simply answered with 3...exd4 4.exd4 and black has already solved all troubles in the opening. It really plays easily after that (=no need to learn anything here), and the "threat" d6-d5 hangs in the air, resulting in an isolated d4-pawn. So even though it is balanced, the black chances are pratically seen promissing as it is more difficult to play correctly as white than it is for black. Not a problem for a chess engine, but for a human below master level, the chances for white are grim here.

But white could play 1.d4 d6 2.e4. I play then the "bad" 2...e5, hoping that white goes for the "good" 3.d4xe5 dxe5 4.Qxd8 Kxd8.
This is statistically very good for white, and 2...e5 is in this respect really a bad move. However, I found out that if you do it the right way in the next 2-3 moves, black has at least an equal game and a very similar position to the good black positions that emerge after 1.d4 d6 2.c4 e5 3.dxe5 dxe5 4.Qxd8 Kxd8.
Black's position is only bad if he goes into 1.d4 d6 2.e4 e5 3.dxe5 dxe5 4.Qxd8 Kxd8 5.Bc4 Bxe6?, a move that is very often done by black, and the results are awful. However 5...f7-f6 is logical and not difficult to find and black has an easy life after that. Strange this isnt played often here, but my computer analysis shows me, that it is really that simple, and it shows that statistics can be misleading sometimes.

remains one problem variation:
1.d4 d6 2.e4 e5 3.Nf3! - after which I play currently 3...exd4. Black is slightly worse here. However, how likely is this comming up if white started with 1.d4 in the first place? Such players dont play 2.e4, and if they do, I have a good chance that they jump at the chance to destroy my right to castle with 3.dxe5 instead of playing the modest 3.Nf3.
But even in the worst case: the philidor exchange variation is not a bad opening for black if you compare it with the statistics of other black opening options after 1.d4. The slav and the orthodox have much worse statistics, so even my worst case scenario (getting into a philidor exchange variation) can be considered a success for black: I managed to get into 1.e4 waters, and they are on average all less dangerous for black. In case I become unhappy with the philidor exchange variation - I have lots of other options after 1.d4 d6 2.e4, which is the advantage of having a solid opening. 

But your question is very off-topic: "what to play after 1.d4?"
As this is about the latvian gambit, and that is an opening after 1.e4

Really, I believe the latvian gambit is just a "Patzer" move. Even if black gets away unpunished, he opened up the game early, s.th. that is always suspicious for black.

[16thompsong]

No, it's ok that he asked the question. He would have a more fully explained answer if he had created his own thread, but this works too.

Well, I would have suggested the Benoni to anyone asking about an answer to 1. d4. However, since you don't like that, I have a few other suggestions:

You say you like unbalanced positions, so you could try to go for a Benko gambit. It is considered solid, but it might not fit into your chess philosophy

Othrwise, I would say that you should look into the King's Indian before you return to the Nimzo/Bogo. Also, Munich's suggestion is not to be ignored. You can play it against both 1. e4 and 1. d4.

[munich]

I believe the benoni or the wolga/benko-gambit are good choices. I do like them! And the benoni would certainly fit to my style, too. I played it for some time, and in Blitz games I occasionally still do.

However, the benoni does not fit into my current opening repertoire for black, which is made of similar black set-ups in order to reduce the amount to be learned.
I do have in mind that I might come back to the benoni or wolga-gambit, as an alternative. For the moment I am very predictable as I always play the same moves. Having an alternative would make me more unpredictable.

For the moment I need to go economical about it, though: I completed my opening repertoire (even though there is still plenty of room to improve on it). I believe my time is currently better invested if I concentrate on rook endgame studies.

About the blocked kings indian - I dont like much! Instead: I like dynamic positions where the tension is in place all the time. With blockades the tension appears to be gone, but the truth is, that it is only temporarily blocked. 10 moves later when the blockage falls appart the whole board is suddenly on fire. I feel like rolling the dice in such positions.

At least I found an interesting approach for white against the KID. It is hardly played and thus very unknown to black KID players (my approach: fianchetto Bg2 with the other bishop on f4. Considered to be doubtful, but I trust stockfish and Rybka here, which both suggest placing the bishop to f4 as best move despite its little popularity. I wrote about it in the threat "reti vs kings indian").

[gm darksquare]

I was having a brainstorm one day about opening, and I considered looking at openings against 1.e4. Being a Sicilian player, I didn't those lines. However, I don't like the French or Caro-Kann. So after (briefly) looking at the Alekhine, I turned to the open games. I didn't like the idea of playing 2... Nc6 because of somebody playing some random opening that I don't feel like looking at. So I finally found a great opening which, I bet you can guess it, is

THE LATVIAN GAMBIT!

Immediately I discounted such sources as the internet, because every single person seems to say it's bad. Even I thought it sucked a week ago. I also jumped over such opening books as SCO and WCO both by Eric Schiller. They irritate me a lot. So I turned to MCO. What I (surprisingly) found was that Black was better in every line except for 3. Nxe5. These lines are the lines given by MCO that I prefer. However, I am playing a side line against 3. Nxe5, as I believe Black's horrible in the main line.

8

7

6

5

4

3

2

1

abcdefgh

( ? ) vs ( ? )

Date:

Event:

Site:

Round: ?

Result:

1. e4 e5 2. Nf3 f5 3. Bc4 ( 3. exf5 e4 4. Ne5 Nf6 5. Be2 d6 6. Bh5+ Ke7 7. Nf7 Qe8 8. Nxh8 ( 8. Nc3 g6 9. fxg6 ( 9. Nxh8 gxh5 ) 9... hxg6 10. Nxh8 gxh5 ) 8... Qxh5 9. Qxh5 Nxh5 10. g4 Nf6 11. Nc3 ( 11. Rg1 Nc6 12. Rg3 Nd4 13. Kd1 g6 ) 11... Nxg4 12. Nd5+ ) ( 3. d4 fxe4 4. Nxe5 Nf6 5. Bg5 d6 6. Nc3 ( 6. Nc4 Be7 7. Ne3 c6 8. Qd2 d5 9. c4 O-O 10. Be2 Be6 ) 6... dxe5 7. dxe5 Qxd1+ 8. Rxd1 h6 9. Bxf6 gxf6 10. Nd5 Kd7 11. Nb6+ Kc6 12. Nxa8 fxe5 13. Bc4 Bd6 ) ( 3. Nxe5 Nc6 4. Nxc6 ( 4. Qh5+ g6 5. Nxg6 Nf6 6. Qh3 hxg6 7. Qxh8 Qe7 ) 4... dxc6 5. Nc3 Bc5 6. d3 Nf6 ) 3...  fxe4 4. Nxe5 d5 5. Qh5+ g6 6. Nxg6 hxg6 7. Qxg6+ ( 7. Qxh8 Kf7 8. Qd4 Be6 9. Be2 Nc6 ) 7...  Kd7 8. Bxd5 Nf6 9. Nc3 Qe7 10. d3 exd3+

1. e4 e5 2. Nf3 f5 3. Bc4   ( 3. exf5 e4 4. Ne5 Nf6 5. Be2 d6 6. Bh5+ Ke7 7. Nf7 Qe8 8. Nxh8   ( 8. Nc3 g6 9. fxg6   ( 9. Nxh8 gxh5  )  hxg6 10. Nxh8 gxh5  ) Qxh5 9. Qxh5 Nxh5 10. g4 Nf6 11. Nc3   ( 11. Rg1 Nc6 12. Rg3 Nd4 13. Kd1 g6  )  Nxg4 12. Nd5+  )   ( 3. d4 fxe4 4. Nxe5 Nf6 5. Bg5 d6 6. Nc3   ( 6. Nc4 Be7 7. Ne3 c6 8. Qd2 d5 9. c4 O-O 10. Be2 Be6  )  dxe5 7. dxe5 Qxd1+ 8. Rxd1 h6 9. Bxf6 gxf6 10. Nd5 Kd7 11. Nb6+ Kc6 12. Nxa8 fxe5 13. Bc4 Bd6  )   ( 3. Nxe5 Nc6 4. Nxc6   ( 4. Qh5+ g6 5. Nxg6 Nf6 6. Qh3 hxg6 7. Qxh8 Qe7  )  4... dxc6 5. Nc3 Bc5 6. d3 Nf6  ) 3... fxe4 4. Nxe5 d5 5. Qh5+ g6 6. Nxg6 hxg6 7. Qxg6+   ( 7. Qxh8 Kf7 8. Qd4 Be6 9. Be2 Nc6  )  Kd7 8. Bxd5 Nf6 9. Nc3 Qe7 10. d3 exd3+ 

I apologize for the 'un' user-friendly text that it's in, but that's generally how I think.

Because of my aggressive form of play, I found some variations over the years and even with some experienced Latvian players, haven't found these little gems.

[gm darksquare]

Needless to say, my preferred attack is

8

7

6

5

4

3

2

1

abcdefgh

( ? ) vs ( ? )

Date:

Event:

Site:

Round: ?

Result:

1. e4 e5 2. Nf3 f5 3. Nxe5 Qf6

1. e4 e5 2. Nf3 f5 3. Nxe5 Qf6

[munich]

...if you play it. But do you play it other than in Blitz games?

Here a variation, which is not forced, but could be considered to be the main variation here:

8

7

6

5

4

3

2

1

abcdefgh

( ? ) vs ( ? )

Date:

Event:

Site:

Round: ?

Result:

1. e4 e5 2. Nf3 f5 3. Nxe5 Qf6 4. d4 d6 5. Nc4 fxe4 6. Nc3 Qg6 7. f3 exf3 8. Qxf3

1. e4 e5 2. Nf3 f5 3. Nxe5 Qf6 4. d4 d6 5. Nc4 fxe4 6. Nc3  Qg6 7. f3 exf3 8. Qxf3

I believe that white has a development lead that counts for something. The idea of black was to get an active game, but looking at the position I'd say it is now white who has all that black wanted.
(needless to say that I recommend the Nimzowitsch defense to surprise white. The surprise will be as big as in the latvian gambit, but in contrast to the Latvian gambit, the Nimzo is sound and actually statistically one of the best black could play. If it wasnt for the popularity, I'd say the sicilian O'Kelly is pretty tough to meet, too, but it isnt so sharp like the nimzo, nor is it unknown.)

[16thompsong]

In the search for the truth on 3. Nxe5, I have looked at every source I have.

The Chess Tempo database confirms what Munich says about Qf6, probably because that's where he got his information. Besides, Qf6, the only other move of considerable games (which in this case aren't a lot) given is Nc6, where the statistics do not favor Black.

MCO gives Nc6 only as a side note, and it doesn't include Munich's critical line 5. d4, instead playing 5. Nc3 Bc5 6. d3 Nf6, where now Black has got some play. The main move is 3... Qf6, and it then follows Munich's line, and then further out, again concluding that White is better.

Next on the list was Chess Openings for White, Explained. This follows the same idea as Munich, but recommends the immediate Nc4. It also recommends the move 5. d4 against 3... Nc6.

8

7

6

5

4

3

2

1

abcdefgh

( ? ) vs ( ? )

Date:

Event:

Site:

Round: ?

Result:

1. e4 e5 2. Nf3 f5 3. Nxe5 Qf6 ( 3... Qe7 4. Qh5+ g6 5. Nxg6 Qxe4+ 6. Be2 ) ( 3... Nc6 4. Nxc6 dxc6 5. d4 fxe4 6. Qh5+ ) 4. Nc4 fxe4 5. Nc3 Qg6 ( 5... Qf7 6. Ne3 Nf6 ( 6... c6 7. Nxe4 d5 8. Ng5 Qf6 9. h4 h6 ( 9... d4 10. Nc4 ) 10. d4 ) 7. Bc4 Qg6 8. d3 Bb4 9. Bd2 Nc6 10. Nxe4 Nxe4 11. dxe4 Bxd2+ 12. Qxd2 Qxe4 13. O-O-O ) 6. d3 Bb4 7. Bd2 Nf6 ( 7... exd3 8. Bxd3 Qxg2 9. Qe2+ Kf8 10. O-O-O ) 8. Nxe4 Bxd2+ 9. Qxd2 Nxe4 10. dxe4 Qxe4+ 11. Ne3

1. e4 e5 2. Nf3 f5 3. Nxe5 Qf6   ( 3... Qe7 4. Qh5+ g6 5. Nxg6 Qxe4+ 6. Be2  )    ( 3... Nc6 4. Nxc6 dxc6 5. d4 fxe4 6. Qh5+  )  4. Nc4 fxe4 5. Nc3 Qg6   ( 5... Qf7 6. Ne3 Nf6   ( 6... c6 7. Nxe4 d5 8. Ng5 Qf6 9. h4 h6   ( 9... d4 10. Nc4  )  10. d4  ) 7. Bc4 Qg6 8. d3 Bb4 9. Bd2 Nc6 10. Nxe4 Nxe4 11. dxe4 Bxd2+ 12. Qxd2 Qxe4 13. O-O-O  ) 6. d3 Bb4 7. Bd2 Nf6   ( 7... exd3 8. Bxd3 Qxg2 9. Qe2+ Kf8 10. O-O-O  )  8. Nxe4 Bxd2+ 9. Qxd2 Nxe4 10. dxe4 Qxe4+ 11. Ne3  

Now you have SCO and WCO. They both say that Nc4 directly against 3... Qf6 is the better plan, and that White gets the better of the position.

However, I let Fritz have a say. Fritz gives 3... Nf6 its stamp of approval. The only other move it considers is 3... Qe7 (So we assume that the other lines were discarded because I checked the box saying cut bad lines?), where it gives a 1.98 advantage to White in the line given by Chess Openings Explained. In its line White only gets .66 of an advantage, but a practical issue arises. White usually has various moves available to keep at least a .4 advantage, while Black has only one move in most positions, or he is suddenly close to 2 pawns down...

8

7

6

5

4

3

2

1

abcdefgh

( ? ) vs ( ? )

Date:

Event:

Site:

Round: ?

Result:

1. e4 e5 2. Nf3 f5 3. Nxe5 Nf6 4. Bc4 ( 4. exf5 Qe7 5. Qe2 d6 6. Nf3 Nc6 7. Nc3 Bxf5 ) ( 4. Nc3 d6 5. Nd3 fxe4 6. Qe2 Qe7 7. Nf4 Na6 ) ( 4. Qe2 d6 5. Nf3 fxe4 6. d3 Qe7 7. dxe4 Nxe4 ) 4...  Qe7 5. d4 ( 5. Bf7+ Kd8 6. d4 Nc6 7. exf5 Nxe5 ) ( 5. f4 d6 6. Nf7 Qxe4+ 7. Qe2 d5 ) ( 5. Nf7 Qxe4+ 6. Qe2 Qxe2+ 7. Kxe2 Rg8 ) 5...  Nc6 6. Nxc6 Qxe4+ 7. Kf1 Qxc6

1. e4 e5 2. Nf3 f5 3. Nxe5 Nf6 4. Bc4   ( 4. exf5 Qe7 5. Qe2 d6 6. Nf3 Nc6 7. Nc3 Bxf5  )    ( 4. Nc3 d6 5. Nd3 fxe4 6. Qe2 Qe7 7. Nf4 Na6  )    ( 4. Qe2 d6 5. Nf3 fxe4 6. d3 Qe7 7. dxe4 Nxe4  )  4... Qe7 5. d4   ( 5. Bf7+ Kd8 6. d4 Nc6 7. exf5 Nxe5  )    ( 5. f4 d6 6. Nf7 Qxe4+ 7. Qe2 d5  )    ( 5. Nf7 Qxe4+ 6. Qe2 Qxe2+ 7. Kxe2 Rg8  )  5... Nc6 6. Nxc6 Qxe4+ 7. Kf1 Qxc6

I only gave lines for white, to show the easy variations he has. All of these lines range from .35 to .66.


We seem to have 4 candidate moves in this line after 3. Nxe5. There is 3... Qf6, 3... Nc6, 3... Qe7, and 3... Nf6.
The Greco Counter Gambit (3... Qe7), according to all sources, sucks. 3... Nc6 gives white an advantage in Explained and the Database, but in MCO, gets a little play in a non critical line which it gives as a mainline for the move. 3... Qf6 gets a bad game in the Database, SCO, WCO, and MCO. As for why it's considered the mainline, "the world may never know." (Sorry, bad pun). Finally, there is 3... Nf6 which isn't considered in any of the sources, and only makes a 5 game appearance in the database. However, it has the support of Fritz, which might be saying a lot. The issue is that with almost any moves, White can keep a .35 advantage, while Black has to play a series of "only" moves. But as Munich has said before:

However, even in the "refutation line", black is not worse than 0.5 - 0.6 pu. However, this "refutation" is others openings "normal", such as the kings indian, the Pirc, the scandinavian center counter, and the Aljechin Defense.

So the idea is that Black gets this disadvantage in most regular openings. And the study time used to find the "only" move and memorize will be more work than your opponent has given. That is: none. While your opponent has been learning the Spanish, the Sicilian, the Caro-Kann, the Alekhine, the French, and who knows what else, all of the work you have done is on mastering these few positions in a sideline opening that most don't bother with.