FreeRepublic wrote on 12/16/23 at 15:51:57:

I attempted to find other analysis on the exchange sacrifice line and largely came up empty. There's nothing on the diagrammed position in ChessPublishing or in the New in Chess Yearbooks. A search on GChess didn't turn up anything useful either. 

The exchange sacrifice was popular in correspondence chess, and ChessBase's Corr database shows a steady stream of games up until the late 90s. It has only four games since then, so perhaps the advent of engines has discouraged Black players from trying the line. One of the 21st century games  (Kryzanowski-Bongiovanni, 2002) was annotated by SIM David Myers. The analysis by Myers is close to that given by Sean Coffey in issue 10 of The New Winawer Report, which Coffey credits to Konikowski in the UltraCorr3 database. I don't have that particular database, so I can't tell if there have been any updates. Konikowski gave an improvement for Black that he evaluated as =/+. While believing this was too optimistic, Coffey thought the position was at least playable, with an evaluation of ∞ or +/=. Ten years later, Stockfish now finds that even this was too optimistic and evaluates the position as +-. 

Thus, I was unable to find any information to change the "does not look good for Black" conclusion. Another barrier to finding improvements in the exchange sac line is Black's success with 10...dxc3/11...dxc3. There isn't much reason to venture the exchange sac when its alternative is doing well.

Lauri Torni wrote on 12/12/23 at 15:45:16:


I still take a brief glance at the King's Indian Defense from time to time.

Now it's either the Nimzo or QGD. The Nimzo is livelier and probably better. The QGD has a "here I take my stand" stubborness which I like.

FreeRepublic wrote on 12/12/23 at 14:36:51:


I already stopped playing the Benkö in favour of Nimzo. Next natural step would be throwing the french away. My belowed 7. Qg4 0-0 was refuted years ago and I have played 3.-Nf6 instead of 3.-Bb4. I, however, don't like this line much. It has no soul.

The advance is currently popular and actually quite strong. Long time ago I  was looking forward to meeting it, not anymore. Then, it was a side line played by players avoiding the main lines. Today, White knows how to play it much better. 

Tarrasch is better for White unless Black plays 3...c5. However, the resulting positions are not very typical of the French defense. 

But what to play instead of the french? Caro-Kann would be quite nice if the classical (3.Nc3) would be forbidden.

I just remembered that the exchange sacrifice line is covered in two issues of The New Winawer Report. Issue 7 has information on its origins. Apparently it was not Uhlmann, as suggested by Moskalenko, but rather Lothar Schmid who introduced 10...Nd7 with the idea of the subsequent 11...Nxe5. See A.Paoli-L.Schmid, Venice 1953. Issue 10 gives a theoretical survey of the exchange sac line. There are limitations due to the state of the engines in 2013, the time of the issue's publication, but it's excellent work nonetheless. 

Sean Coffey, the author of these publications was extremely knowledgeable about the Winawer, and it's a shame that he stopped after 24 issues. Each issue is a treat for Winawer devotees. In 2016 he announced plans to continue the publication in some form, but sadly nothing appears to have developed in this regard. Amazingly, The New Winawer Report was free, so it was truly a labor of love.  For those interested in seeing the details on the exchange sac line in issues 7and 10 or just curious about the publication as a whole, see

http://www.irlchess.com/tnwr/

Sean posted on this forum in the 2012-2013 period, but doesn't seem to have been active since then.

Thanks for the excellent posts. My memories go back to a 1971 book, The Poisoned Pawn, by Zeuthen and Jarlnas. Sad to say, but theory moves on and the exchange sac line can no longer be recommended for Black. At least other ways have been found against Euwe's line.

Nernstian59 wrote on 10/13/23 at 20:24:44:


I added that game to my data base of games. "requires high precision from both sides." Indeed. My engine suggests that Black hangs in there with 21...Rc8. However, there is also 18Kxc2!? where White evaluates to being the equivalent of a pawn up. It looks like an unclear mess to my eyes, but White is an exchange up.

FreeRepublic - Your post inspired me to do some research of my own. Recent books don't cover this variation, which I'll call the "exchange sacrifice line", adopting Moles' terminology. Some (like Miedema as you note) don't mention it at all. Berg says that he investigated it, but was less satisfied with it. Watson, in the 4th edition of Play the French, gives the line up to the 14th move, saying that "the analysis of both 14...Rf8 and 14...d3 goes on in depressing length; I believe there are draws there for Black if you desperately need one." His statement suggests that heavy duty analytical prep may allow Black to grovel for some draws.

The most extensive coverage of the exchange sacrifice line is found in much older books.  These all seem to prefer 10...Nd7 to your 10...Nbc6, though the two moves obviously transpose after 11...Nxe5. Moles explains the reasoning behind 10...Nd7: if White tries 11.f4 to prevent 11...Nxe5, Black brings the knight to e4 via c5. In his Wonderful Winawer, Moskalenko calls 10...Nd7 "Uhlmann's option", but he doesn't give any analysis

Of the older sources, Moles appears to have the most thorough coverage. RHM's The French Defense has a bit of analysis and Pachman in Semi-Open Chess provides somewhat more. Eric Schiller in his French Winawer Poison Pawn is close to Moles. Despite his less than sterling reputation, Schiller's analysis doesn't seem to be too bad, though this could be because he quotes Moles rather liberally. In one non-Moles line, it's fascinating to see White's strongly +- position degrade to +/- and then to = and finally to -/+.  And Schiller is supposedly writing from White's POV! 

Despite his book's analytical shortcomings, Schiller does provide a bit of a historical nugget when he states that "10.Kd1 was all the rage in the 1960s, receiving as much attention as the Najdorf Poison Pawn does today" (which would be 1988, the year that Schiller's book was published). The exchange sacrifice line was apparently the main line back then, and the Mega Database shows numerous games by top players in the 1965-1967 period. However, by 1975 Moles was referring to 10...Nd7 as "the old move, previously held to refute 10.Kd1, but now thought inferior". He concludes, "On the whole, despite the enormously complicated play, the exchange sacrifice line favours White, whose K-side pawns count for more than Black's central push." With respect to the two 14th moves for Black, 14...Rf8 seems a bit more highly regarded because Black is able to return some material and start rolling his central pawns forward, producing some counterplay. The other move 14...d3 is given a ?! by most commentators, who reference Bronstein-Uhlmann, Zagreb 1965, where White overlooked a winning move and allowed a draw by perpetual check. Moles' analysis stands up fairly well to engine scrutiny. Stockfish finds some improvements for Black, but Moles' overall verdicts aren't overturned because the engine also finds better moves for White.

Finally, the most recent book to treat the exchange sacrifice line in any detail is Pederson's The Main Line French: 3.Nc3 (2001).  Pedersen's analysis is mostly similar to Moles'. He offers a few improvements but nothing to overturn the overall evaluation of the line. 

While reviewing Moles' work with Stockfish, the following engine line line came up: 14...Rf8 15.Bd3 Bd7 16. Ke2 (Best according to Moles) and now the untried 16...Nf5, with the following continuation: 17.Re1 0-0-0 18.Nxf7 Rde8 19.Ne5 Re7 20.Qh3 Qc3. Note that Black rejects Moles' human strategy of returning the exchange and just hangs on with typical engine stubbornness. Stockfish evaluates the position after 20...Qc3 as +/-, which is definitely better than the winning +- given for all other lines where White doesn't play an obvious inaccuracy.  However, it seems unlikely that a position where White is "merely" better instead of winning will be much of an incentive for Black players to revive this line.

Ender wrote on 06/30/09 at 18:01:52:


According to J. Watson in PtF4, it's playable (i.e. White doesn't lose) but "you probably wouldn't want to be White in a real game". His main line starts with 10. - Nbc6 11. Nf3 dxc3 12. Ng5 Nxe5 13. f4 f6!.

The line is complicated enough that you may have some success against unprepared opponents, but it's forcing nature means that you will likely struggle somewhat against well-prepared ones.

Personally, I'd rather punt 8. Bd3!? instead of the Euwe variation in the move order you're giving (some players play 7. - cxd4 instead of 7. - Qc7, most likely to avoid this attempt).