The Sicilian Berlin

After I finished my last game of the Belgium interclubs I was surprised to see in the pub that many players were busy following the ongoing game of Magnus Carlsen. A Belgian FM even showed me on his smartphone how that game finished by a beautiful and surprising mate see 2019 Grenke chess classic round 8. Players seemed (at least temporarily) more interested in Magnus' game than the ones of their teammates.

Unless you were completely disconnected from chess last couple of months, you could've not missed the news about the recent amazing successes of the reigning world-champion. While he defended last year with a lot of difficulty his world-title against Fabiano Caruano, suddenly wins are obtained again very easily. Magnus is again absolutely hot and that creates again speculations about the magical barrier of 2900 elo which can or can not be broken by a human player.

It just proofs to me that elo is relative. We all have fluctuations in our playing-strength. Sometimes there are clear reasons for that like e.g. new responsibilities but we should also not exclude luck (see my article the lucky one) especially in a short timeframe. I guess this last aspect can play here an important role as the gain of 40 elo by Magnus happened in only 31 games.

In the Chessbase-report of the last round in Grenke we can read that Carlsen can today benefit of the analysis made for the last world-championship. Magnus:" I can still use ideas and concepts which we have analysed." We notice this clearly in his shift from e5 to c5 as main-choice against 1.e4. In the last 10 years Magnus used e5 as his preferred weapon in standard-games but since the world-championship of last year he has another favorite. This is clearly shown in the table below which I created by screening the Big database 2019.

The switch brought some clear profit. After the world-championship Magnus achieved in 7 games, 4 win and 3 draws against (very) strong grandmasters. It is the new Berlin which everybody can't break. Meanwhile the solution is also known to the riddle which I asked in the article curieuzeneuzemosterdpot. Why nobody tries to play the mainline? Finally the Azerbaijani top-grandmaster Teimour Radjabov tried it in the last Tata Steel. I assume Teimour also got curious and he never minds a draw either.

[Event "81st Tata Steel Group A"] [Site "?"] [Date "2019.01.25"] [Round "?"] [White "Radjabov, Teimour"] [Black "Carlsen, Magnus"] [Result "1/2-1/2"] [ECO "B33"] [WhiteElo "2757"] [BlackElo "2835"] 1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e5 6. Ndb5 d6 7. Bg5 a6 8. Na3 b5 9. Nd5 Be7 10. Bxf6 Bxf6 11. c3 O-O 12. Nc2 Rb8 13. Be2 Bg5 14. O-O Kh8 15. a3 Ne7 16. Ncb4 Nxd5 17. Nxd5 Be6 18. Qd3 Qc8 19. Rad1 a5 20. Qg3 h6 21. h3 Qc6 22. Qd3 Qd7 23. Qg3 Qc6 24. Qd3 Qd7 25. Qg3 1/2-1/2

You probably wonder what is special about this line. Besides Teimour doesn't choose the most critical test. The answer you can find in the Ultracorr-x. If we select the games played with the position after 12....Rb8 by the very best players in correspondence-chess (both players having + 2500 elo) during the last 10 years then we discover below devastating statistic.

Indeed all 32 top-games played in the last 10 years were drawn. Not once white or black won. It is the Berlin in a much stronger version. By the way I also noticed that lower rated players rarely can win. I already discovered this in 2015 after I made a serious analysis of the opening which popped up in my game played in the clubchampionship of Deurne against Marcel Van Herck.

[Event "Klubkampioenschap Deurne r6"] [Site "?"] [Date "2015.??.??"] [Round "?"] [White "Brabo"] [Black "Van Herck, M."] [Result "1/2-1/2"] [ECO "B33"] [WhiteElo "2318"] [BlackElo "2100"] [PlyCount "68"] [EventDate "2015.??.??"] [CurrentPosition "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"] 1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e5 {(I was sure that Marcel would try to surprise me. I never saw Marcel trying the Svechnikov before so my preparation was useless.)} 6. Ndb5 d6 7. Bg5 a6 8. Na3 b5 9. Nd5 Be7 10. Bxf6 Bxf6 11. c3 O-O 12. Nc2 Rb8 {(Ach so Marcel prepared the opening by using my game played against Gilles in Open Gent of 2013.)} 13. h4 Be7 14. Nce3 {(Marcel told me after the game that he was slightly disappointed because I deviated from my game against Gilles. Obviously I didn't want to lose the same game twice. Besides I had recommended on my blog Nce3 already as an amelioration but Marcel had missed this in the preparation.)} 14... Be6 15. a4 {(This is very popular in correspondence-chess but almost unknown in standard-chess.)} 15... Qd7 16. axb5 {(Again I forget my analysis as Be2 is only played after axb5. Anyway it is unclear to me if any move can guarantee any advantage for white. Besides often the moves just transpose.)} 16... axb5 17. b4!? {(The critical move is for sure Be2 but even then white has very little to play for.)} (17. Be2!? Bd8 18. h5 b4 {(H6 and Bg5 have been tested a lot in correspondence chess with acceptable results for black. I concentrate upon b4 as the engines can't find the smallest advantage for white against that continuation.)} 19. h6!? {(Immediately Ra6 is also possible but I don't see anything special for white.)} 19... g6 20. Ra6 bxc3 21. bxc3 Rb7 22. O-O Kh8 {(This position has been tested already 3 times by very strong white-players in correspondence-chess. Each time black made a draw effortlessly.)}) 17... Rfc8 18. Bd3 Bd8 19. Qf3 Ne7 20. h5 Bxd5 21. Nxd5 Nxd5 22. exd5 g6 23. hxg6?! {(More accurate is Be2 to maintain the balance.)} (23. Be2! e4 24. Qxe4 Bf6 25. O-O Rxc3 26. Bd3 ∞) 23... fxg6?! {(It is surprising that this position was already analyzed in Lets check. I got access to the tool quite recently as I bought the engine Komodo 8. Here hxg6 is stronger as the menacing g4 is not so good.)} 24. Ra3 Ra8 25. Rxa8 Rxa8 26. O-O Ra3 27. Rb1?! {(I still play for a win but that is not smart. Rc1 is better.)} 27... Bb6?! {(This seems very attractive but the engines prefer Rxc3 immediately as now white gets tactical counterplay.)} 28. Qe2?? {(I am too optimistic. I miss the 29th move.)} (28. Qf6! Rxc3 29. Ra1 Rc8 30. Bxb5 Qc7 31. Bc6 Bd4 32. Ra8 Rxa8 33. Bxa8 = {(After this sequence of typical engine-moves, the queens will be exchanged and we get an endgame of opposite bishops which is a draw.)}) 28... Rxc3 29. Bxb5 Qf5 30. Rf1 Rc2 {(After the game I showed Rg3 to my opponent which the engines evaluate as winning. Likely Rc2 is also still sufficient for the win.)} 31. Qf3 Qxf3?! {(It is hard to resist such moves against a 200 higher rated player but this jeopardizes the win.)} (31... Rd2! 32. Qxf5 gxf5 33. Bd7 f4 34. Bf5 Kg7 35. Be4 Bd4! -+ {(Stockfish and Komodo can't hold this for white so I guess this should be winning but it is still a serious technical job.)}) 32. gxf3 Rb2!? {(Rd2 and Rc3 were more challenging especially as white had little time.)} 33. Bd3 Rxb4!? {(Again more critical is Bd4 to keep the rooks on the board.)} 34. Rb1 Rxb1+ {(Black proposed a draw which I accepted of course immediately. I would've continued the game with black as it is not yet a dead draw.)} (34... Rxb1+ 35. Bxb1 Kf7 36. Bc2 Kf6 37. Be4 (37. Ba4 Kg5 38. Bd7 h5 39. Kg2 h4 40. Be6 {(Necessary was Be8 to transpose to the other line as now white loses.)} 40... Kf4 41. Bf7 g5 42. Be6 (42. Bg6 h3+ 43. Kxh3 Kxf3 -+ {(Black keeps the g-pawn and that makes a vital difference.)}) 42... e4 43. fxe4 g4 44. Bf5 h3+ 45. Kf1 Kf3 46. Kg1 {(Otherwise the h-pawn will cost a piece.)} 46... Bxf2+ 47. Kh1 Kg3 48. Be6 Kh4 49. Bf5 g3 -+) 37... Kg5 38. Kg2 h5 39. Bb1 h4 40. Bc2 Kf4 41. Bxg6 h3+ 42. Kxh3 Kxf3 43. Bh5+ Kxf2 44. Bg6! Kf3 45. Bh5+! {(With this accurate move white can just hold the draw.)}) 1/2-1/2

It is the reason why I don't play anymore the mainline of the Svechnikov. So there will be no follow up anymore of a theoretical duel in the Svechnikov and the scientific approach. However what else I will play is something you will need to find out for yourself. I already surprised somebody with it and I hope to do it a few times more. I already share a lot of information on this blog, likely more than what a dozen of players share together so I don't think somebody can blame me of being unfair.

Anyway I still try to keep track of the developments in the mainline. This is new for me as before when I stopped playing a line, I lost interest. Nowadays I do also regression-analysis. So at some points of time I look up if something has changed which maybe makes the line again playable. Since 2015 I've analyzed some recent small discoveries for white  but simultaneously there were also some shifts favoring black which cancelled out any possible advantage for white.

Today Magnus has an edge but I am sure his opponents won't rest. Sooner or later people will find anti-dotes which will decrease the scores of Magnus in the Sicilian. Openings are for many players something boring but a few can also enjoy the eternal fluctuations.

Brabo