Swapping a minor piece for 3 pawns?

What is your opinion: is it a good idea to swap a minor piece for three pawns?

I lost a game against a stronger player recently where I had the option of aggressively making the swap on move 19 or of sitting back and defending. I chose the latter and lost. Chess.com’s analysis engine rates the two approaches about the same, but I was concerned that a strong opponent would find a way to pick the pawns off one by one as they went down the board.

Here is the game. I am sort that I don’t know how to make a chessboard appear in the post.

[Site "Chess.com iPhone"]
[Date "10/26/2017 09:43pm"]
[White "usignolo (1403)"]
[Black "DavidEricAshby (1261)"]

1.d4 d5 2.c4 c6 3.Nf3 Bg4 4.e3 e6 5.Be2 b6 6.O-O Bd6 7.Ne5 Bxe2 8.Qxe2 Nf6 9.Nc3 O-O 10.f4 c5 11.Rf3 cxd4 12.exd4 dxc4 13.Qxc4 Nbd7 14.Qc6 Qc7 15.Qxc7 Bxc7 16.Bd2 a6 17.Rc1 Rac8 18.g4 Nxe5 19.fxe5 Nd7 20.Ne4 f6 21.Bb4 Rf7 22.Nd6 Rff8 23.Nxc8 Rxc8 {usignolo won by resignation}

 

Dierdre and off topic rocker, thanks for your input. Yes, the alternative line that I wonder about is 19...nxg4. 20 h3 attacks the knight and it has no safe square to run to, so 20...nxe5. 21 d4xe5. Bxe5.

After that, black has connected passed pawns, but with a long way to go to queen.

Thanks for the comments.
After h3, Nh6, Bxh6 ruins the King’s castle and destroys the kingside pawn structure, so I didn’t see that as a sensible escape route for the knight.

A minor piece is always stronger than 3 pawns up to the very late endgame(just 1 or 2 pieces each side), unless those are quality pawns, i.e. connected, defended, advanced ones, passers.

 

So I guess the general rule is: never trade a minor piece for 3 opponent pawns, unless you get

very specific and obvious compensation in return.

[Low-key self-advertising alert]

I created a YouTube video analyzing the idea of trading pieces for several pawns in some depth, which may be helpful.

Cole's Notes: Sacrificing pieces for several pawns is a good idea if

1. You gain the initiative;

2. The enemy King is  / becomes open;

3. It becomes an endgame and the pawns can advance very quickly.

Every position needs to be evaluated, but the above checklist is a useful guide.  Hope that helps.

 

Great video.

However, we should ask ourselves what is initiative?

On 2, that would be true only if there are available concrete tactics, otherwise the surplus minor piece

can shelter the king sufficiently well.

3, I guess, is half-true. The piece is still stronger in the endgame, if those are not

connected advanced passers.

 

PS. sorry for posting so much today, I just discovered the discussion going on here,

and I kind of liked it, so I don't want to stop.

 

 

I would have grabbed the g4 pawn, then if h3 moved my N to h6 where, if White plays Bxh6, gxh6 opens up the g-file for Blacks's R's to go after the unprotected White K.

I don't know if it would have won, but it's a plan!

I disagree with your statement that a minor piece is always better than 3 pawns until late endgame because when less pieces are involved (maybe somewhere between late middle game and the endgame) since pawns can tie up the extra minor piece and threaten to make many passed pawns which can be hard to stop (even a rook pawn by itself can be annoying for a knight to stop) and I don't think you would be able to just sacrifice a minor piece and guarantee that you'll receive three pawns back to restore material balance (though you could win the pawns outright). Something I feel is interesting is the sacrifice of a minor piece for two pawns but with something else as compensation like activity, king safety etc

It depends on the situation though. The three pawns are often equivalent in power to a bishop or knight, but depends on many factors including whether the pawns can be picked up easily or not, and whether the pawns are all in the sixth rank or elsewhere. In the game I think either choice can be made as long as you know the subsequent plans and attacking ideas, but my personal preference is to go ahead with the plan (I love my pawns very much). Pawns combined together not only forms an effective fighting force, but also a strong defence force (imagine three pawns stampeding forward to bully a knight). Though, in the actual game, Black will have to be mindful of the unprotected queenside Black pawns in the game.

 

Many times I have sacrificed a bishop or knight for a mere solitary pawn (or even for nothing in return - I am very generous at times) for fun, and I always challenge myself to use my pawns and other forces to justify the sacrifice. Of course this ends up in me having more failed attempts than successful ones, but the successful attempts are very pleasing. Given the option, I will probably perform the sacrifice for three pawns without much thought.

That is the problem, the pawns could be outright won.

That tests good in practice. Top engine code would prefer pieces more than pawns.

A minor piece is generally worth close to 4 pawns, so way above 3.

 

Endgame tablebases/practical examples also support this.

For example, RRB vs RR is a tablebase win(Fischer did not know that).

Adding 2 pawns for the weaker side, that are not connected, i.e. RRB vs RRPP,

is also a win for the stronger side, proven in many games, so the 2 additional pawns

don't seem to matter at all.

What is much more important is the better coordination among the higher number of pieces.

Quality pawns, advanced, passed, connected, though, might have the upper hand even over a rook.

Thank you all for your inputs! In summary, it would have been great fun and educational (Deirdre Sky and Eric) to have made the trade of the minor piece for the three pawns. However as the pawns are not particularly advanced and it does not give the initiative, it may not have been a game winning move. From the specific example Mickin MD's idea of stealing just one pawn, allowing white to open up black's castle and then sallying from the castle with the rook sounds interesting too. Lyudmil introduced the idea of endgame tablebases, which I was not familiar with previously and will investigate later. The video looks great, and I will have to watch it when I wouldn't wake up the rest of the house in the process. Out of interest, I ran the game through chess.com's code, choosing the best move for each side and with the three pawns, black eventually won, though it was a very close run thing and I am sure that there was far more scope for black to blunder than for white. Which code did you run to have white out on top Lyudmil?

Cheers all :-)

David

 

I like to sacrifice pieces where possible and reasonable - the temptation is just too irresistible for me. Simply trading bishop for bishop or rook for rook can be boring at times.

 

The computer must have felt that the three extra pawns are really the superheroes in the game, but the win must have taken very long. As a human player playing Black though, this will be harder, since the pawns are nowhere near promotion and some of them are hanging.

Komodo on 16 cores.

Try the same position with a top engine on good hardware and longer TC,

if you want, run a statistical tour of 10 or more games from the same position with

different engines, and you will see white will get on top.

The black e6,f7,g7 pawns are connected, but unadvanced, so white should have the advantage.

I did not say it was a win, though, It might well still be a draw with perfect play.

What is that chess.com code?

I guess it is Stockfish running on limited time or to a fixed depth, right?

In that case, this is not trustworthy, a single game is also not trustworthy

statistically. You need a much larger sample.

Lyudmil, yes, I think that it was Stockfish running for a couple of seconds or so per move.

Wow, SmithyQ, I just watched your video and it is pretty awesome! I am adding the YouTube link: https://youtu.be/HrEBIyMTInQ because the link you gave shows up on my computer, but not on my phone.

Having watched the video, your answer definitely matches the position and the question that I had and did so at a nice pace and in a manner that a 1200s to 1300s player like myself can understand. Good luck with the YouTubing.

David

I made a trade of 2 knights for a rook and three pawns once I probably would have done it even if the rook was a piece because the three pawns were opposing my four pawn majority so I tried to trade down to queen pawn with my opponent having a minor piece endgame.  Ended up losing one of my pawns before getting 1 pawn on the 6th and 2 on the 7th which made queening 1 pawn and taking their bishop inevitable.

Big daddy, Eric 22, Lil boat, do you have any games that you can post which illustrate a won or a lost game following trading pieces for pawns?

In general I sacrifice a knight for bishop for up to two pawns; sometimes one, sometimes zero.

Here's one of my games, where I messed up (in the first place I had insufficient attackers to attack the White king) after attempting 14...Nxg3 15. hxg3 Qxg3 and should have lost (I was completely toast around move 30+), but I won due to exchanges of inaccuracies, mistakes and blunders, some of which involve time trouble. I am unable to put annotations inside because the annotations always disappear every time I posted the game.

 

 

-----------------------------------------------------------------------------------------------------------

 

Here's my second game which I lost after trying sacrifices on move 13.

 

 

-----------------------------------------------------------------------------------------------------------

 

This time my opponent performed a failed sacrifice.

 

 

-----------------------------------------------------------------------------------------------------------

 

In this fourth example, my opponent performed the sacrifice in the opening. Against the player, I have mixed results in this opening as Black - some wins, some losses, and perhaps, some draws as well.

 

 

-----------------------------------------------------------------------------------------------------------

 

Another sacrifice which I attempted and obviously unsound, but I managed to win eventually because of weird stuff unrelated to the sacrifice.

 

 

-----------------------------------------------------------------------------------------------------------

 

This time both players performed sacrifices at different moves, one on move 13 and one on move 20; we just played the sacrifice for fun.

 

 

I have many more games, too many to count, that involve sacrificing a knight or bishop for pawns, with mixed results.

 

 

Thank you Eric,

will review these later. Teaching my daughter just now.

David