How A Grandmaster Analyzes His Games (Black vs. 4.Bf4 Grunfeld)
For today's game, I am annotating my Round 5 game from the Oceania Zonal, sharing my thought process during the game. Though to be honest, I did not think a lot (I had 77 minutes on my clock at the end of the game). 1. d4 I knew my opponent was a 1.d4/2.c4 player, who tended to like slow, building up/strategic play. 1... Nf6 2. c4 g6 3. Nc3 d5 Against such players, I find the Grunfeld works quite well, since the position will almost certainly be somewhat dynamic and open. 4. Bf4 This wasn't a huge surprise, as my opponent had played it in one game a couple of years ago, and it's a fairly safe and solid approach that matches my opponent's style quite well.
(4. cxd5 Nxd5 5. e4 Nxc3 6. bxc3 Bg7 7. Qa4+ is what my opponent had mostly played before. I was intending to avoid it with the interesting move order 6...c5!?, where I had done some preparation.)
4... Bg7 5. e3 c5 I decided to go against the trend of 5...0-0 to avoid a certain drawish line:
(5... O-O 6. cxd5 (6. Rc1 would be my preference as White, though Black has been doing well lately with 6... Be6 and even (6... c5!?) 6... Nxd5 7. Nxd5 Qxd5 8. Bxc7 Na6 (8... Nc6 9. Ne2 Bg4 10. f3 is tricky but better for White. 10... Bf5 11. Nc3 Qe6 12. Kf2) 9. Bxa6 Qxg2 10. Qf3 Qxf3 11. Nxf3 bxa6 12. Rc1 and a significant majority of games ended in a draw from here, which didn't really appeal to me. Also, such a position would suit my opponent's style fairly well.)
6. dxc5 Qa5 7. Rc1 The main move, anticipating ...Ne4.
(The main alternative is 7. Qa4+ Qxa4 8. Nxa4 I had forgotten what to do against this move, but fortunately the antidote is quite simple: 8... Bd7 9. Nc3 Ne4 10. Nxd5 Na6 11. f3 Nexc5 and Black's lead in development gives him full compensation for the pawn. See e.g. 0-1 (41) Wang,Y (2732)-Topalov,V (2803) Nanjing 2010)
(7. cxd5?! Nxd5 8. Qxd5 Bxc3+ 9. bxc3 Qxc3+ 10. Ke2 Qxa1 11. Be5 Qb1 12. Bxh8 Be6 13. Qd3 Qxa2+ 14. Kf3 f6 is obviously not in White's best interest - his extra bishop is trapped and his king is quite weak.)
7... dxc4 The clearly best move, releasing the central tension.
(7... Ne4 8. cxd5 Nxc3 9. Qd2 Qxa2 10. bxc3 Qa5 might score okay for Black, but it feels extremely dodgy if White develops normally.)
8. Bxc4 O-O It is important to play this before ... Qxc5.
(8... Qxc5? 9. Nb5)
9. Ne2 This knight move is somewhat trendy, as it gives Qb3/Nb5 ideas more sting when the rook on c1 is protected. Indeed, I had a feeling my opponent might go for this, as he played correspondence for many years and therefore would be a bit more theoretical than your average 1900.
(9. Nf3 Qxc5 transposes to the old 4.Nf3 Bg7 5.Bf4 0-0 6.e3 main lines, and Black is pretty comfortable here in general. A brief outline: 10. Bb3 (For comparison, Qb3 is less effective with the knight on f3: 10. Qb3 Nc6 11. O-O (11. Nb5? now fails to 11... Be6 - which would not work with Ne2 (covering the c1-rook) in place of Nf3!) 11... Na5 12. Qb5 Qxc4 13. Qxa5 b6 14. Qa3 Re8 15. Rfd1 Qa6 16. Qb4 Bb7 and Black has the bishop pair advantage. However, his queen is a bit awkward on a6, so the position is equal.) 10... Nc6 11. O-O Qa5 12. h3 Bf5 13. Qe2 Ne4 14. Nd5 e5 and there is some more theory, with White normally choosing between 15.Rxc6, 15.Bh2 and 15.Bg5, but in all cases Black is doing fine with his central space and fairly active pieces. The knight on d5 looks annoying, but can be undermined with ...Be6.)
9... Qxc5
(9... Nc6 10. O-O Qxc5 11. Nb5 Be6 was my choice in an Isle of Man game four years ago, but I had lost to some strong preparation by my opponent, so I didn't want to repeat it!: 1-0 (49) Basso,P (2428)-Illingworth,M (2517) Douglas 2015)
10. Bb3 My opponent hadn't faced this 5...c5 variation before and wasn't ready for it, which showed in this move. That said, the evaluation doesn't really change - the position remains equal.
(10. Qb3 is more critical, when I had quickly checked the following variation before the game: 10... Nc6 (10... Qa5 11. O-O Nbd7 would have been my 'shortcut' of choice to get a game if he'd bashed out the first 10 moves (suggesting preparation).) 11. Nb5 Threatening Bxf7. 11... Qh5 12. Ng3 Qh4 13. Nc7 e5 14. Bg5 (14. Bxe5 Nxe5 15. Nxa8 is greedier, which may be why the engines initially consider it more promising. But the following forcing line seems OK for Black: 15... Bd7 (15... b6 and ...Bb7 is another way, but I trust it less.) 16. Nc7 Bc6 17. Nd5 Kh8 A nice move, to meet Nxf6 with Bxg2! 18. Nf4 g5 White is missing the tempo he needs to consolidate with castling. 19. Nd3 Nxd3+ 20. Bxd3 Bxg2 White has a few tries here, but with the White kingside collapsing, the position seems dynamically balanced, e.g. 21. Nf5 Qh3 22. Nxg7 Kxg7 23. Bf1 Ne4 and Black has enough piece activity, especially with White's king stuck in the centre, to claim full compensation. This type of very concrete play and sacrificing material for the initiative is quite typical for the Grunfeld, which is why I tend to recommend that my students don't play this opening until they are already quite strong (2200+ Elo). Well, if you really, really like it, ignore my warning and go ahead!) 14... Qxg5 15. h4 (15. Nxa8 e4 sets up ...Qe5-b8 to trap the knight (or 16.Nc7?? Qa5).) 15... Qg4 16. Nxa8 Na5 17. Qb5 Nxc4 18. Rxc4 Bd7 19. Qb4 e4 20. Nc7 Bc6 21. Qd6 h5 and I stopped here, trusting that I would be able to find the right moves from here now that the position is settled. The litany of correspondence games suggest that Black has full compensation for the exchange, as White's c7-knight is stuck and White can't easily castle because of ...Qxh4.)
10... Qa5 I wasn't sure where my knight belonged yet, so I moved my queen prophylactically.
(10... Nc6 11. O-O Qh5 is also decent - Black has no weaknesses, so he has more decent options.)
11. O-O Na6 This move wasn't my initial intention - I realise that sounds weird to say on a move you spent one minute on! Initially I was going to play 11... Nc6, but I had a feeling my opponent would see ghosts and go astray if I kept the option of both ...Nc5 and ...Nb4. This proved to be true in the game.
(11... Nc6 is what I would probably do against a player my own rating or higher. The position should be equal, but Black has scored better in practice, with a good example being: 0-1 (47) Bryakin,M (2407)-Afanasiev,N (2460) Voronezh 2018. Incidentally, what would you play as Black after 12. Nd4 ? That's right, we should keep the tension: 12... Bd7 , and now our bishop gets to a better square if White releases the piece tension with Nxc6?!)
12. Nd5?! (83 minutes to 73) I had considered this option, but it was still a bit of a surprise to me. It's easy to explain in retrospect - my opponent wants to trade all the pieces in the hope of making a draw.
(12. Qc2 is what I expected, though it's a bit inaccurate because of 12... Nb4 13. Qb1 Nh5 , acquiring the bishop pair advantage.)
(12. Bc4 is probably best, to meet ... Rd8 with Qb3, although the position remains equal after simple development like 12... Bd7)
12... Nxd5 When there is piece tension and the opponent has a threat, it's often easiest to just trade and get out of the threat in the process. 13. Qxd5 Qxd5 14. Bxd5 Nb4 In this position I trusted my gut and quickly played the attacking move.
(14... Bxb2 15. Rb1 Bf6 16. Bxb7 (16. Rfd1!?) 16... Bxb7 17. Rxb7 seemed hard for Black to play for a win in the game. I probably underestimated Black's chances here, but it doesn't matter too much, as White can also delay taking on b7 and play for (quite decent) development compensation.)
15. Bc4 (74 to 63 minutes) I expected this move, but I thought 15.Bf3 was probably better, to keep the pressure on the b7-pawn and leave the c-file open for White's rook (for Rc7 invasions). Indeed,
(15. Bf3! e5 (or 15... Nxa2 16. Rc7 e5 , but White has the extra and probably better option now of 17. Bg5!) 16. Bg3 Nxa2 17. Rc7 should be close to equal, with Black keeping some slim chances due to the passive g3-bishop and e2-knight.)
15... Bf5 I was a little bit nervous about 16.Nd4 when I played this move, but nothing else really made sense for Black, so I played this fast. Then, as I thought more about 16.Nd4, I realised that this is a move I would be happy to see! 16. Nc3?! This surprised me and indeed the move is a mistake. Ironically, this was the one time I had a significant think (90 to 81 minutes) in the game, and yet my decision was not the best! Maybe you can do better?
(16. Ng3 Bd3 (16... Bxb2? 17. Nxf5 gxf5 18. Rb1 Ba3 19. Rb3 is annoying!) 17. Bxd3 Nxd3 18. Rc7 with at least some active piece play is the way to limit the damage. It's easy to think White loses a piece to 18... e5 19. Bg5 f6 , but after 20. Rd1 Rad8 21. Bh4 things aren't that simple, as ...g5 runs into Nf5, and otherwise White is quite active with Kf1-e2 in the air to exploit the d-file pin.)
(16. Nd4? Bd3 17. Rfd1? (17. Bxd3 Nxd3 18. Rcd1 Nxb2) 17... Rac8 is something White should avoid.)
16... Nd3?! Unfortunately, I made the wrong choice, which loses most of my advantage. I realised that the arising position wouldn't offer much for Black, but I underestimated my chances after the alternatives. In hindsight, I was also a bit too focused on the result, as I was thinking about this endgame was so quiet and that it would be not so easy to convert it into a win, even against a much lower-rated player.
(16... Bd3 17. Bxd3 Nxd3 18. Rc2 is an instructive position, since here I assumed that Black had to play in a direct way. In fact, Black can just keep the position as is, leaving the strong d3-knight in place: 18... Rac8 and with ...Bxc3! coming up, White is really feeling the heat. (18... Nxf4 19. exf4 ⩱ is the only option I considered, when indeed White is holding without too much trouble.))
(16... Rac8 17. Nd5 Nxd5 18. Bxd5 Bxb2 felt strong during the game, but I couldn't make it work: 19. Rxc8 Bxc8 20. Rb1 and what I missed is that Black doesn't have to move the b2-bishop, but can play 20... Rd8! 21. Rxb2 (21. Bxf7+?? Kxf7 22. Rxb2 Rd1#) 21... Rxd5 , with an extra pawn and quite decent winning chances. Still, I think I like 16...Bd3 best of all.)
17. Bxd3 Bxd3 18. Rfd1 Rfd8 19. f3 Now White is very solid, and I already felt that I would have my work cut out to win. 19... Ba6?! Once again going against my intuition, which considered 19... f5 to be the obviously best move.
(19... f5 I was ready to play this, but just after my opponent moved, I got worried about 20. Rd2 and Rcd1. But then 20... Ba6 21. Rcd1 Rxd2 22. Rxd2 Rc8 keeps a comfortable advantage for Black, thanks to the bishop pair.)
20. Kf2 f5 21. Bc7? White plays to exchange everything and make a draw, but it has a major flaw.
(21. Nd5! is a move I saw, but I underestimated it: 21... Bxb2 22. Rc2 Ba3 23. Bc7 Re8 24. Rdd2! and suddenly the a3-bishop is a bit short of squares. White will obtain compensation with Rc3 next move. A sample line is 24... Rec8 25. Rc3 e6 26. Nf6+ Kf7 27. Rd7+ Be7 28. Be5 Rxc3 29. Bxc3 Rd8 30. Rc7 and I don't see how Black untangles.)
(21. e4 would also fit well, now that ...Bd4 can be met with Be3.)
21... Rxd1 22. Rxd1 Kf7 I knew I couldn't avoid my opponent's simplifying Rd8 idea anyway, so I played an improving move. 23. Rd8? Surprisingly, this move leads to real problems for White.
(23. Ba5 Rc8 24. Rd2 would still be quite solid, although Black has serious winning chances with the bishop pair and the possibility of grabbing space with ...b6/..a5 or ...g5/...h5.)
23... Rxd8 24. Bxd8 Bc4 I initially wanted to play ...Be5, but then I thought I could improve the move order. My decision was justified in the game by my opponent playing to automatically save his pawns, which proved a decisive mistake.
(24... Be5 I rejected this move because I wanted to discourage Ba5, but 25. Ba5 b6 26. Bb4 Bd3 and ...a5 is still pretty tough for White. I have a feeling it's holdable, but only if he plays very well.)
25. a3? Now White is just lost - can you see how?
(25. Ba5? b6 26. Bb4 a5 27. Ba3 b5 and ...b4 is winning for Black.)
(25. Bc7 is what I expected, since White does not have to fear the opposite-coloured bishop endgame. In that sense, my bluff was successful. I would have played 25... Ke6 26. Bb8 a6 , with the plan of steadily advancing with ...Kd7-c6, ...b5, ...a5 and ...b4. I would most likely go for ...g5/...h5/...g4 too, to try and create weaknesses in White's kingside structure. Again, White is probably within the drawing margin, but it's not so trivial.)
25... Be5! Suddenly White has no way to save his bishop. 26. f4
(26. b3 could be tried immediately, though the same reply as in the game still works: 26... Bd3 (The prosaic 26... Bxb3 also wins.))
26... Bb8! It is important for the bishop to go here and not to d6!
(26... Bd6 27. Ba5 b6 28. Bb4 is the reason why, when Black would have to win the game all over again.)
27. b3 An attempt to confuse.
(27. Ba5 b6 28. Bb4 a5)
27... Bd3!? I really liked my idea after Nd5, so I went for this.
(27... Bxb3 28. Nb5 Ba4 29. Nc7 Bc6 is the other line I saw, when Black is winning because the knight and bishop are stuck. However, I couldn't see an easy way to win these pieces, so I went for the forced win of material instead.)
28. Nd5
(28. Ba5 b6 29. Bb4 a5 traps the bishop once again.)
(or 28. a4 b6 29. a5 Ke8 30. axb6 Kxd8)
28... Ke6 I was really proud of this move, although
(28... Bd6 would also win (but only a pawn).)
29. Nb4 My opponent tries to be tricky.
(29. Bc7 Bxc7 30. Nxc7+ Kd6 31. Ne8+ Kd7 32. Ng7 e6 and ...Ke7-f7 winning the knight on g7 is what I'd seen. It is not unusual for fianchettoed knights to get stuck, as they only have four legal moves!)
(29. Nxe7 Kd7 30. Nxg6 hxg6 is easy.)
29... Be4 30. Ba5
(or 30. Na2 b6)
30... b6 31. Na6 This is only a temporary solution to the trapped bishop. 31... bxa5 32. Nxb8 Kd6 White's knight can't move anywhere, as Na6 fails to Bb7. 33. b4 axb4 34. axb4 Bd3 White resigned, as he's about to lose the knight for nothing. I hope this game has shown you how to defeat players rated about 1900! Now I will share with you what I would say to my opponent if he were my student: 1) Your opening was okay, you knew the first few moves got an equal and fairly safe position. 2) Your most common mistake was making some concessions in the form of exchanging pieces, borne out of trying to play for a draw against a much higher-rated player. 12.Nd5 handed Black the initiative - maybe you assumed Black had to take on b2, and missed the intermediate 13...Nb4. 3) Continuing from this point, you got fixated on forcing the exchange of rooks with 21.Bc7? and 23. Rd8?, when playing more actively, or even a solid move to improve the position, would have actually put pressure on your opponent to keep good winning chances. It's well worth asking 'does this trade favour me or my opponent?' before committing. 4) The game was decided in the end by losing your sense of danger - you assumed there were no real tactics once the rooks came off and forgot about the bishop on d8 being short of squares. It was almost study-like how your bishop got trapped - looking for the opponent's idea before making your move will help you minimise these oversights. 5) A more advanced point, but I would also work on looking beyond your initial assumptions - on moves 15 and 16, you played the most obvious move, but could have ended up in trouble had Black been a bit less forcing in his thinking, and appreciated the long-term strength of his position in the 16...Bd3 variation. Ask yourself 'can I ignore the opponent's threat?' (be it ...Nxa2 or ...Bxb2) before delving deeply into ways to defend against the opponent's threat. Thanks for taking the time to learn from this game! Let me know your questions and feedback! 0-1
I hope you enjoyed reading my thought process during each move of the game! This serves as a good example of how you should go about analyzing your own games.
In the past, when people found this too hard to do, they would need to find a coach to go through the game with them, so that they could extract the key lessons for next time.
However, I'm about to share a video lesson called 'How to Analyse Your Games' to show you how I analyze a game 'live', and give you a step-by-step process to how you can analyze your games yourself, and get the maximum learning and value out of this work!
To access this video lesson, you need to be a member of 'The Chess Improvement Group' - my paid peer-to-peer mentoring group, where we work together to resolve any challenges we face in becoming stronger chess players, while I share frequent chess content (including 7+ hours of video lessons, many annotated games, and opening files) to direct your study and give you some of that 'magic' Grandmaster chess understanding
In the past, when people found this too hard to do, they would need to find a coach to go through the game with them, so that they could extract the key lessons for next time.
However, I'm about to share a video lesson called 'How to Analyse Your Games' to show you how I analyze a game 'live', and give you a step-by-step process to how you can analyze your games yourself, and get the maximum learning and value out of this work!
To access this video lesson, you need to be a member of 'The Chess Improvement Group' - my paid peer-to-peer mentoring group, where we work together to resolve any challenges we face in becoming stronger chess players, while I share frequent chess content (including 7+ hours of video lessons, many annotated games, and opening files) to direct your study and give you some of that 'magic' Grandmaster chess understanding
