Jon Speelman's Agony Column #20

by Jonathan Speelman
9/21/2016 – After a short interlude in Baku, we return this week to this column's normal business, with a tale of Agony and revenge in the regional Bundesliga. Sent by Andreas Beyerlein they comprise two games against the same opponent, Miroslav Kalous. In 2010 Andreas got a winning position but then dived into the abyss; two years later he easily converted another won game. Analysis and training by GM Jonathan Speelman.

Jon Speelman's Agony Column #20

Andreas Beyerlein

Andreas Beyerlein writes: "I am 34 years old, married, and have two children (aged two and five years). I have an MSc in Statistics and a PhD in Human Biology, and I live in Munich where I am working as a researcher in diabetes epidemiology. Apart from spending time with my lovely kids and playing chess, I am interested in football (but I am not a supporter of Bayern München), history (of chess and in general) and science.

I currently play for SC Roter Turm Altstadt, a club from Munich, but the games against Miroslav Kalous come from league matches with my former team, SG Siemens Amberg, against SC Windischeschenbach, a club from close to the Czech border."

He annotated the games in some detail and I've added my comments as JS as usual.

[Event "Regionalliga Bayern Nordost "] [Site "?"] [Date "2010.03.21"] [Round "8.2"] [White "Kalous, Miroslav"] [Black "Beyerlein, Andreas"] [Result "1-0"] [ECO "D02"] [WhiteElo "2066"] [BlackElo "1958"] [Annotator "Speelman,Jonathan"] [PlyCount "67"] [EventDate "2010.??.??"] [SourceDate "2015.07.13"] 1. d4 d5 2. Nf3 Nc6 3. Bf4 Bg4 4. e3 e6 5. Be2 Bd6 6. Bg3 Nf6 (6... Nge7 { JS is more common} 7. Nbd2 Nf5 {they often castle but this makes White take a decision about the bishop immediately} 8. Bxd6 cxd6 {and Black has equalised.}) 7. Nbd2 Bxg3 8. hxg3 Qd6 9. c3 O-O (9... O-O-O {AB maybe risky JS You can play this way if you like since White is a long way from creating a serious attack though} 10. b4 {prevents e5 for the moment}) 10. Qa4 $5 {[#]} e5 $2 {This should lose a pawn} (10... Bf5 11. Nh4 Bg4 $11) (10... a6 11. Bd3 {preparing e4 }) (10... h6 {stops tactics against h7 though it could encourage White to try to organise g3-4-5 in the future.}) 11. dxe5 $2 (11. Nxe5 Bxe2 (11... Nxe5 $4 12. dxe5 Qd7 13. Qxd7 Nxd7 14. Bxg4 $18) 12. Nxc6 bxc6 13. Kxe2 $16 {JS Black can certainly fight but shouldn't have enough for the pawn.}) 11... Nxe5 12. Nxe5 Bxe2 13. Kxe2 (13. Nxf7 {JS White also wins a pawn with this internezzo though after} Rxf7 (13... Qb6 14. Kxe2 Qxb2 15. Ng5 {is nonsense}) 14. Kxe2 Re8 15. Nf3 {Black has excellent compensation with a nice position and much the safer king. Play might then continue} Qb6 16. Qb3 Qa6+ 17. Ke1 c6) 13... Qxe5 14. Rh4 {JS I like this move playing to use the h-file but Black is very solid and should be fine.} (14. Qf4 $11) 14... Rfe8 15. Rah1 Rad8 $6 (15... Qf5 { JS is Houdini's suggestion, the idea being that if} 16. g4 Qg6 17. Kd1 d4 18. cxd4 Nxg4 {when the White king is a bit unsafe}) 16. Nf3 Qf5 17. g4 Qg6 (17... Qe4 {AB does not work well} 18. g5 (18. Qxe4 dxe4 $11) 18... Qxa4 19. Rxa4 Ne4 20. Rxa7 Ra8 {[#] JS This is a line which Black would be very unlikely to go into but it is a typical position in which you have to decide on which endgame you want. If} 21. Rxa8 (21. Rxb7 {My initial instinct was that this would lead to more but after} Rxa2 22. Rc1 g6 23. Rc2 c5 24. g3 {it remains very complicated so I think I'd prefer the first line.}) 21... Rxa8 22. a3 $14 Ra5 23. Rc1 c6 (23... Rb5 $2 24. b4 {misplaces the rook}) 24. Nd2 $1 {returning the pawn to get active} Nxg5 25. c4 {with quite a nice advantage}) 18. g5 {[#]} Ne4 (18... Re4 19. Qb3 (19. Qc2 Rxh4 20. Qxg6 hxg6 (20... fxg6 $5 21. Rxh4 Ne4 22. Rg4 Rf8 23. c4 c6) 21. Rxh4 Ne4 22. Rg4 Kf8 23. Nd2 Nxd2 24. Kxd2 Ke7 { JS is pretty equal.}) 19... Rxh4 20. Rxh4 Ne4 21. Qxb7 h6 $1 22. Qxc7 Qa6+ 23. Ke1 Rf8 {JS Black is getting the queen in here and White must equalise with} 24. gxh6 Qd3 25. hxg7 Qb1+ {with perpetual check.}) 19. Rh5 $6 {Giving Black a very dangerous initiative.} (19. Rxh7 Qxh7 20. Rxh7 Kxh7 21. Qxa7 {Computers tend to overassess this type of position for the queen since they value her relatively more highly than humans but here White does look better because he has a couple of pawns, and - at least as important - the Black king is exposed.} Ra8 22. Qxb7 Rxa2 (22... Rab8 $2 23. Qxd5 Rxb2+ 24. Kf1 Rb1+ 25. Ne1) 23. Nd2 { JS This is where Andreas ended his line and I agree that White is better though it's certainly not that clear for example if} (23. Qxd5 $4 Nxc3+) (23. Qb5 $4 Nxc3+) (23. Qxc7 $2 Rxb2+ 24. Kd3 Nxf2+) 23... Nxd2 24. Kxd2 Re6 25. Qb3 Ra1 26. Ke2 Rb6 27. Qxd5 Rxb2+ 28. Kf3 Raa2 29. Qf5+ Kg8 30. Kg3 g6 31. Qf6 Re2 32. f3 Rxg2+ 33. Kf4 Rgd2 {is still very much a battle}) 19... Nc5 20. Qd1 (20. Qd4 Ne6 $1 {JS and it turns out that the queen doesn't have a decent square.} 21. Qg4 (21. Qa4 b5 $1) 21... f5 $1 22. Qg3 (22. Qh4 Qxh5 23. Qxh5 Nf4+) 22... f4) 20... Qa6+ 21. Ke1 $2 {JS This should lose} (21. Kd2 Qxa2 22. Rxh7 { JS was a total mess.} Qxb2+ 23. Qc2 (23. Ke1 Kf8 24. Nd4 Re5 $1) 23... Ne4+ 24. Ke1 $1 Qa1+ 25. Ke2 Qa6+ 26. c4 $1 Kf8 27. R1h4 {[#] JS whatever the true evaluation - and Houdini is giving it as equal - in a game either side could lose this position within a couple of moves.}) 21... Nd3+ 22. Kd2 Nxf2 {[#]} 23. Qc2 (23. Rxh7 $4 Qd3+ 24. Kc1 Qxd1+ 25. Rxd1 Kxh7) 23... Nxh1 $2 ({AB I can't remember why I did not play} 23... g6 24. Rxh7 Nxh1 25. Rxh1 $19 { JS Of course this would have avoided a world of trouble and in some sense it was the critical moment in the game.}) 24. Qxh7+ Kf8 25. Rxh1 Qg6 $6 (25... Qxa2 26. Qh8+ (26. Nd4 Qxb2+ 27. Kd3 Rxe3+ 28. Kxe3 Qxc3+ 29. Kf2 (29. Qd3 Re8+ ) 29... Qxd4+ $19) 26... Ke7 27. Qxg7 Qxb2+ 28. Kd3 Kd7 29. Qxf7+ Kc8 30. Rh7 Qb6 31. Nd4 $19 {AB but still difficult to play JS Agreed. Of course Black should be totally winning but could still go wrong.}) 26. Qh2 $1 {AB a bit nasty JS It's now really easy to go wrong.} Qb6 $6 (26... Rd6 {JS blocks the diagonal and keeps the queen defending}) (26... c6 27. Qc7 Kg8 28. Ne5 Qe6 29. Nxf7 $1 {and Black can't do better than take perpetual check.}) 27. Kc2 Rxe3 ( 27... Qxe3 28. Qxc7) 28. Nd4 Qg6+ (28... g6 {JS would have prevented Nf5 at a nasty moment and kept some control.} 29. Rf1 (29. Qh6+ Ke7 30. Rf1 Qa6 31. Rf2 (31. Rf4 Qd3+ 32. Kb3 Rd6) 31... Qd3+ 32. Kb3 Qxd4) ({and} 29. Qh8+ Ke7 30. Qg7 Qa6 $1 {are both simple but}) 29... Qa6 30. Rxf7+ Kxf7 31. Qf4+ Kg8 32. Qxe3 Qa4+ 33. Kd2 Re8 {would still leave Black with some work to do.}) 29. Kb3 {[#]} c5 $4 (29... Rd6 $2 30. Qh8+ Ke7 31. Qc8 $1 {attacking c7 and f5} Rb6+ 32. Ka4 Ra6+ 33. Kb3) (29... Qd3 $1 30. Qxc7 Qc4+ $17 {JS and Black should win in the end - rooks are better than knights.}) 30. Qc7 $18 c4+ ({AB Here I realized with horror that} 30... Qb6+ {does not work because of} 31. Qxb6 axb6 32. Nf5 $1) 31. Kb4 a5+ (31... Qd6+ 32. Qxd6+ Rxd6 33. Rh8+ $1 Ke7 34. Nf5+ $18) 32. Kc5 b6+ 33. Kb5 Ree8 ({AB I remember that we were both in time trouble, but I think my opponent had only one second (with old time control!) left on his clock, so either} 33... Rd7) ({or} 33... Qh6 {would probably have won the game for me, although both moves lead to forced checkmate!}) 34. Rh8# {AB One of my most painful defeats ever! JS Andreas outplayed his opponent in the opening and early middlegame but allowed him back into the game after winning the exchange. The intermezzo 25...g6! would have left him with a safe king and easy play but after failing to play this he faced serious practical problems, even if engines sneer at White's compensation. He became confused and in time trouble it was a lottery.} 1-0

[Event "Regionalliga Bayern Nordost"] [Site "?"] [Date "2012.02.05"] [Round "6.4"] [White "Beyerlein, Andreas"] [Black "Kalous, Miroslav"] [Result "1-0"] [ECO "C03"] [WhiteElo "1947"] [BlackElo "2035"] [Annotator "Speelman,Jonathan"] [PlyCount "49"] [EventDate "2012.??.??"] [SourceDate "2015.07.13"] {AB I was able to take revenge about two years later.} 1. e4 e6 2. d4 d5 3. Nd2 b6 $6 4. Ngf3 (4. Bd3 Ba6 {JS Of course this isn't in any way special for Black but there's no reason to move the bishop from f1 and lose a tempo}) 4... Nf6 5. e5 (5. Ne5 $5 {JS is a suggestion of Andreas's or perhaps Fritz. I'd want to reply} Bb7 6. exd5 (6. Bb5+ c6 7. Bd3 {doesn't look bad for Black}) 6... Qxd5 {if possible and Houdini doesn't demur}) 5... Nfd7 6. c4 $1 dxc4 ( 6... Bb7 7. cxd5 Bxd5 8. Bc4 {JS Black has a nice square on d5 if he can get organised but lags in development} c6 {was my first thought though quite possibly wrong} 9. O-O Be7 10. Qe2 {If the b8 knight were somewhere sensible I'd quite like Black's position but he's going to have to work to organise it and this will give White an appreciable amount of time to try to create something}) 7. Bxc4 Ba6 (7... Bb7 {JS is obviously also possible and given that Black is lagging in development and will find it hard to prevent d5 a better idea.}) 8. O-O Be7 (8... c5 9. d5 $1) 9. Re1 Bxc4 10. Nxc4 Nc6 $6 { AB causes even more trouble to Black} (10... O-O) (10... b5 {JS makes sense of the position aining to put the knight on b6 to control d5.} 11. Ne3 (11. Na5 $1 {is very testing} Bb4 (11... c5 12. Bd2 O-O 13. dxc5 Nxc5 14. b4 {and the knight on a5 is very unsettling}) 12. Bg5 Qc8 13. Bd2 Bxa5 14. Bxa5 Nc6 15. Bd2 Nb6 {or possibly Ne7. Black is in danger on both sides of the board and something awful could easily happen to his king. But he will hope to lose a pawn in return for establishing a knight on d5 in which case he would have decent chances.} (15... Ne7)) 11... Nb6 12. Qe2 {JS followed by Bd2. Again Black would be absolutely fine if the knight on b8 were developed but since it isn't will have to be very careful. Still this looks much less scary than Na5}) 11. a3 {JS Preventing ...Nb4} (11. d5 exd5 12. Qxd5 Nb4 13. Qd2 (13. Qe4 Nc5 14. Nd6+ cxd6 15. Qxb4 O-O 16. Rd1 Qc7 {JS Certainly no more than +=}) 13... Nf8 $1 (13... O-O 14. Ne3 $16) 14. Nd6+ cxd6 15. exd6 Ne6 16. dxe7 Qxd2 17. Bxd2 Nd3 {should be okay for Black.}) 11... O-O (11... b5 {made sense again to get some squares.} 12. Ne3 Nb6 13. Qd3 O-O $1 {JS possible because if Qxb5 Nxd4 } 14. Rd1 (14. Qxb5 Nxd4) 14... a6 15. Bd2 Qd7 16. Rac1 Rfc8 {JS Of course White must be a bit better but Black is fighting.}) 12. d5 $5 {JS This gives Black a tactical opportunity. Slower play would have retained a very nice advantage.} (12. Qa4 Ndb8) (12. Qc2) 12... exd5 13. Qxd5 {[#]} Ndb8 $2 (13... Ndxe5 $1 {AB only chance} 14. Rxe5 (14. Qxd8 $2 Nxf3+) 14... Nxe5 15. Qxe5 Qd1+ 16. Qe1 (16. Ne1 $4 Bf6 $19) 16... Qxe1+ 17. Nxe1 Rad8 18. Be3 $14 {JS White is certainly better here but rooks are powerful in the endgame and Black has decent fighting chances.}) 14. Qe4 {JS Black simply hasn't got enough development} b5 (14... Qd7 {JS was also very bad.} 15. Bg5 Na6 16. e6 $1 Qxe6 ( 16... fxe6 17. Rad1 Bd6 18. b4 {White will regain the pawn with heavy intewrest }) 17. Qc2 Qd7 18. Rad1 Bd6 19. Bf4 Rae8 20. Rxe8 Rxe8 21. Bxd6 cxd6 22. Ng5) 15. Ne3 Qd7 16. Nf5 $18 Na6 $4 {[#] JS Losing immediately but White is winning anyway.} (16... Qe6 17. N3d4 Nxd4 18. Nxd4 Qb6 19. Qxa8 Qxd4 20. Be3 $18) ( 16... Re8 17. Bf4 Qe6 18. Rac1 Qg6 19. Red1 {JS and Black still retains material equality but his position is ghastly.}) 17. Qg4 Qxf5 (17... g6 18. Nh6+ Kg7 19. Qxd7) 18. Qxf5 Nc5 19. Qc2 Na5 20. b4 Ncb3 21. Rb1 Nxc1 22. bxa5 Bxa3 23. Ng5 g6 24. Rexc1 Bxc1 25. Qxc1 {JS In the opening, there is always a tension between getting your forces developed and setting up the pawn structure you want. In this game, Black arranged an exchange of light squared bishops and created a potentially excellent post for a knight on d5. But he lost too much time doing so and was crushed right out of the opening. Given the grim alternative, 13...Ndxe5! was forced as a bail out even if the resultant endgame ought to be fairly unpleasant.} 1-0

Did you ever play agonising/ecstatic games that you would like to share? Send them in to! If your games gets published, you get a free three-month ChessBase Premium Account.

Do you want to avoid agony in games? Let Nicholas Pert help you.

Nicholas Pert:
Typical mistakes by 1800-2000 players

GM Nicholas Pert about his DVD: “After the success of my previous DVD Typical mistakes by 1600-1900 players I decided to produce a follow up DVD aimed at players of a slightly higher level. The examples all come from games played by players with a rating between 1800 and 2100. This DVD offers slightly more complex material than the previous DVD, and will hopefully provide an insight into what I believe are the main errors which stop players of this level to be more successful. I divided the material into categories such as “when to exchange pieces”, “how to convert an advantage”, “passive pieces”, “anticipating your opponents plan”, “openings and pawn structure”. Each section contains several examples which illustrate the theme and practical examples which allow the viewer to test his skills. This DVD provides a useful training tool for ambitious players and may highlight mistakes that the viewer should be aware of.”

  • Video running time: 3 hours 50 min. (English)
  • With interactive training including video feedback
  • Extra: 50 additional examplesIncluding CB 12 Reader
  • ISBN: 978-3-86681-513-1
  • Delivery: download, post
  • Price: €29.90; €25.13 without VAT (for customers outside the EU); $28.41

About the author

Jonathan Speelman was born in 1956 and graduated from Worcester College Oxford, where he read mathematics. He became an IM and a professional chess player in 1977, a GM in 1980 and was a member of the English Olympic team from 1980–2006. Three times British Champion he played twice in the Candidates, reaching the semi-final (of what was then a knockout series of matches) in 1989 when he lost 4.5-3.5 to Jan Timman. He was twice a second to a world championship challenger: Nigel Short and then Viswanathan Anand against Garry Kasparov in London 1993 and New York 1995.

Jon has written for The Observer (weekly) since 1993 and The Independent since 1998. With its closure (going online, but without Jon on board) he is expanding online activity and is also now offering online tuition. He likes puzzles, especially (cryptic) crosswords and killer sudokus. If you'd like to contact Jon, then please write to


Jonathan Speelman, born in 1956, studied mathematics but became a professional chess player in 1977. He was a member of the English Olympic team from 1980–2006 and three times British Champion. He played twice in Candidates Tournaments, reaching the semi-final in 1989. He twice seconded a World Championship challenger: Nigel Short and then Viswanathan Anand against Garry Kasparov in London 1993 and New York 1995.
Feedback and mail to our news service Please use this account if you want to contribute to or comment on our news page service


Rules for reader comments


Not registered yet? Register