12 September 2021

Unforced errors

In the tradition of previous Chess vs. Tennis posts, here is another one with an insight for chessplayers, excerpted from The Economist (Sept 11, 2021) article "All Too Human", which leads off talking about Naomi Osaka's collapse after being ahead at this year's U.S. Open tennis tournament:

...Yet the way her fortunes turned at the US Open, with one mistake begetting another, turns out to be common. A recent paper by David Harris, Samuel Vine and Mark Wilson of the University of Exeter and Michael Eysenck of Royal Holloway University of London finds that top-tier tennis players are surprisingly prone to mistakes caused by situation driven anxiety...

The taxonomy of shots in tennis helps to isolate the impacts of consistency and risk appetite. When a player misses a routine shot and loses a point, it is scored as an "unforced error" (UE). Conversely, shots that bounce within the court without being touched by an opponent are "winners".

...In high-pressure contexts such as break and set points, UEs were 15% more common than under less stressful situations. Similarly, during points following UEs, the chances of a repeat blunder rose sharply. These effects reinforced each other, so that UEs were even more common during high-pressure points following a prior UE than you would otherwise expect.

I don't think it's very surprising to find that high-stress moments tend to lead to more mistakes, even at an elite level. However, I find it more interesting to see the study's observations that these mistakes do indeed tend to subsequently reinforce themselves, which can result in a downward spiral of performance. While this could also be considered "common sense", having an acute awareness of these types of situations can be important for a player, who may then choose to consciously focus on their mental toughness in an effort to combat a further downhill slide in their game.

This finding is consistent with what I've seen over the years in analyzing both my own games and those at the master level. Often I've found that a series of sometimes small mistakes - or perhaps even just failures to play better moves - can sometimes then suddenly lead to much more significant blunders when under pressure from an opponent. Blunder recovery is then extremely important and often quite possible, even when down material or under heavy positional pressure, again underlining the importance of mental toughness to the ultimate result of the game. Recently in Annotated Game #255 I had just such a moment for recovery, but was unable to find the drawing line at the board. That was just one of many opportunities that I've seen occur, even in "losing" games, so the old advice to never give up until you are actually beaten remains good.

11 September 2021

Annotated Game #256: Back to the grind

This second-round game had me paired against a much lower-rated opponent. However, his rating was provisional and in practice he played significantly stronger. His main weakness was in not mobilizing his forces in a King's Indian Defense and letting me grab space and eventually win his unprotected d-pawn. After that I followed a strategy of safety first and felt confident in my strategic advantage, which became more apparent as material left the board. NM Dan Heisman's advice of pursuing a "go to sleep" strategy in the endgame paid off, as I covered all my weaknesses and then forced an imminent breakthrough, my opponent losing on time.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "ChessAdmin"] [Black "Class E"] [Result "1-0"] [ECO "E61"] [Annotator "ChessAdmin/Komodo Dragon 2"] [PlyCount "84"] 1. d4 Nf6 2. e3 g6 3. Nf3 Bg7 4. Be2 O-O 5. O-O d6 {transposing into the King's Indian formation, as I had expected.} 6. c4 Nbd7 7. Nc3 c5 {much less popular than the standard ...e5, but a good alternative. However, it's a less direct way of challenging in the center.} 8. d5 {database statistics show that White does better to maintain the central tension, for example after b3 or a3. I was comfortable immediately going for more space, however.} Nb6 {targeting c4 and opening the diagonal for the Bc8.} 9. Qc2 {this waiting move doesn't really do much for me, although it reinforces e4. I should do better by focusing on exploiting Black's slightly awkward setup, for example pushing a2-a4 immediately to gain further space.} Bd7 {this is slow, allowing me to recover the initiative, as the bishop is not placed to do anything useful here. Challenging in the center with ...e6 or ...e5 looks better.} 10. a4 Bg4 { despite the double bishop move, this is Komodo's choice as well, pointing out the flaw in developing to d7 in the first place.} (10... a5 $6 {would block ideas of a4-a5 but leave too many holes on the b-file for White to later exploit.}) 11. b3 {the idea here is to develop the dark-square bishop before doing anything else major. More active play is possible.} (11. Ng5 $5 {would offer to trade Black's most active piece and help increase White's light-square advantage.}) 11... a6 $16 {this keeps my knight out of b5, but now my structural and developmental advantages are more evident.} 12. Bb2 { although not winning yet, at this point I was content with what I felt was a strategic advantage here, with Black's pieces not cooperating well and me able to pursue longer-term ideas on both the queenside and in the center. For example, I can think about a4-a5 and Nc3-a4-b6, as well as the e4 advance and further play in the center.} Bd7 $6 {this third bishop move is just a time-waster.} (12... Bf5) 13. Rfe1 {with the idea of targeting play on the e-file, which I felt was the most dynamic option and also would keep squeezing Black.} Nc8 {my opponent continues time-wasting maneuvers.} 14. e4 Na7 15. h3 { a prophylactic move, taking away the g4 square from both Black's knight and bishop, to continue the squeezing motif.} (15. e5 $5 dxe5 16. Nxe5 {looks more active, but I was not in a rush to resolve the situation in the center.}) 15... Re8 {finally developing the rook to the e-file, to be able to aid in a future fight there.} 16. Bd3 {this is a bit of a wasted move, as the bishop accomplishes nothing on the new diagonal and e4 is already overprotected.} e6 { correctly challenging White's central pawn formation.} 17. dxe6 Bxe6 18. Nd5 { this obvious-looking move may let Black off the hook in the center. The point is that Black could then exchange off the bad light-square bishop for the centralized knight.} Nc6 $14 {getting the knight back into the fight, as b4 looks like a good place for it.} 19. Qd2 {proactively moving away from a ... Nb4 threat and to more open diagonals.} Rc8 (19... Bxd5 $5) 20. Rad1 {like my opponent, I try to develop my rook to a more potentially useful file.} Bd7 $6 $16 {while not immediately losing, this creates some targets for me to focus on and complicates Black's play. The d7 square is taken away from the Nf6, reducing the amount of space available to his pieces, and the d6 pawn is now unprotected, creating some tactical possibilities.} 21. Bb1 {getting out of the way of the d-file heavy pieces, while keeping up protection of the e4 pawn. } a5 $2 {ignores the threat to the pawn and thereby creates a fatal weakness.} (21... Be6) 22. Nxf6+ Bxf6 23. Bxf6 (23. e5 $5 {I did not even consider, instead going immediately for the win of the pawn. Komodo favors it because of the extra threats involved, although it is much more complicated.} Bg7 24. Qxd6 $18) 23... Qxf6 24. Qxd6 Qxd6 25. Rxd6 Be6 26. Red1 Rcd8 27. Kf1 {following the simple plan of activating the king.} (27. e5 $5) 27... Rxd6 {I was please to see this, as simplifying down material, especially from double to single rooks, normally makes the defender's situation worse in an endgame.} 28. Rxd6 Rd8 29. Rxd8+ {happy to exchange down further} Nxd8 30. Ke2 Kf8 31. Ne5 (31. Ke3 $5 {no reason to not keep centralizing the king and making it more powerful.}) 31... Ke7 32. Nd3 b6 33. Bc2 {played with the idea of establishing a defensive blockade first, then grinding away on the win.} Nc6 34. Bd1 { now ...Nd4 is no longer a threat.} Kd6 35. f4 {seizing space. although perhaps a little premature.} (35. Ke3 $5 {would better centralize the king and open up the diagonal for the bishop.}) 35... f6 36. g4 {my opponent by this point was quite low on time. I endeavored to play solid moves that kept my advantage and make the game easier on me, while difficult for my opponent to do anything useful.} Bd7 37. Ke3 Ke6 38. h4 $18 {by this point my 4v3 pawn majority is fully mobilized and supported by my pieces. The threat is to break through with g4-g5.} Nd4 39. f5+ {picking the simplest path to a win.} gxf5 40. exf5+ Kd6 41. Kf4 Be8 42. Nf2 Ke7 {and lost on time, although the knight maneuver Ne4-c3-d5 would decide things.} 1-0

06 September 2021

Annotated Game #255: Too eager to return

Earlier this year I had my first OTB tournament in some time and was quite eager to return to the board. Usually the first game of a tournament after a long layoff is dedicated more to cleaning out mental rust and getting back into fighting form; this one is no exception. The opening phase is not bad, despite my relative unfamiliarity with the position, but I am too aggressive in my intentions and delay kingside castling for too long. As a result, my opponent gets a significant strategic advantage and I decide to sacrifice material twice in order to try and get counterplay. The second sac is a better practical one and in fact my opponent did go wrong in the ensuing king hunt, allowing what would have been a drawing continuation, but alas I did not find it either. 

Although disappointed in the loss, I did not worry about it too much, given the ratings differential (over 100 points in my opponent's favor) and the fact I had another three games to go. At least I had made my opponent sweat, which gave some satisfaction.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "Class A"] [Black "ChessAdmin"] [Result "1-0"] [ECO "D31"] [Annotator "ChessAdmin/Komodo Dragon 2"] [PlyCount "67"] 1. d4 d5 2. c4 c6 3. Nc3 e6 {the Triangle/Wedge formation. This can transpose to various different openings and my opponent opts for an aggressive continuation.} 4. e4 Bb4 (4... dxe4 {is most common here and may be superior, even with the ...Bb4 idea. One recent example with Carlsen playing it:} 5. Nxe4 Bb4+ 6. Nc3 c5 7. a3 Ba5 8. Be3 Nf6 9. Nge2 cxd4 10. Qxd4 Qxd4 11. Bxd4 Nc6 12. Bc5 Ne4 13. Be3 Nxc3 14. Nxc3 Bxc3+ 15. bxc3 e5 16. Rb1 Kd8 17. f4 Re8 18. Kf2 Kc7 19. Bd3 h6 20. Rb5 f6 21. Rhb1 Re7 22. Be4 Kd8 23. Rd5+ Kc7 24. fxe5 Nxe5 25. Bf4 Rb8 26. Ke3 Bd7 27. Rc5+ Bc6 28. Bxc6 b6 29. Bxe5+ fxe5 30. Rd5 Kxc6 31. Rf1 Rc8 32. Rf5 Kb7 33. Rfxe5 Rxe5+ 34. Rxe5 Rxc4 35. Kd3 Rg4 36. g3 Kc6 37. c4 Rg5 38. Re7 a5 39. Kd4 Kd6 40. Rb7 Rg4+ 41. Kd3 Kc6 42. Re7 h5 43. Re6+ Kc7 44. Re5 g6 45. Re6 Rg5 46. Kd4 Kb7 47. Rf6 h4 48. gxh4 Rg4+ 49. Kc3 Rxh4 50. Rxg6 Rxh2 51. c5 Rh3+ 52. Kc4 bxc5 53. Kxc5 Rxa3 54. Kb5 a4 55. Rg7+ Kc8 56. Rg4 Kc7 57. Rxa4 Rxa4 58. Kxa4 {1/2-1/2 (58) Xiong,J (2709)-Carlsen,M (2863) Lichess.org INT 2020}) 5. e5 {my opponent thought for a while here, probably expecting that I would take on e4. The text logically grabs space and prevents the knight development to f6, so} Ne7 6. Nf3 c5 {the correct reaction. This break is a major theme in the line, challenging White in the center and also with the idea of developing the other knight with Nb8-c6.} 7. dxc5 Bxc5 { here I thought for a while, not familiar with the position. It's better to at least temporarily sacrifice the pawn, rather than be so concerned with immediately re-establishing material equality.} (7... Nbc6) 8. cxd5 { psychologically this continued the pressure, with me feeling that my position was rather open once White's light-square bishop is freed. However, I didn't see anything real for White and this is objectively the case.} exd5 (8... Nxd5 {seemed a little loose to me.}) 9. Bd3 Bg4 {may be a little premature, as following the usual rule about developing knights before bishops seems better. However, in practical terms it is probably equal, since the knight development is coming anyway.} 10. O-O Nbc6 11. Re1 {this overprotects the e-pawn, but developing the dark-square bishop would be a more effective move.} Qd7 $6 { at this point I had to consider middlegame plans and also how to finish my king development. I was too afraid of castling kingside and instead kept my options open for castling queenside with the text move. However, it is too passive a square for the queen.} (11... Qb6 {is the logical reaction to White's last move, which left the f2 pawn vulnerable. White would now have to spend a move covering it, losing the initiative and allowing ...O-O.}) 12. h3 { White calls my bluff on the c8-h3 diagonal. The bishop sac on h3 would get me nothing, so now I have to retreat.} Bh5 (12... Bxf3 $2 13. Qxf3 {and now White really does have an excellent attacking game on the kingside.}) 13. Be2 Bg6 $6 {unnecessary and time-wasting. White does not follow up as strongly as he could, at least.} 14. Bd3 (14. Na4 $1 {one way or another will result in the disappearance of the Black dark-squared bishop, currently my most effective piece, making White's counterpart a monster. For example} Bb4 15. Bd2 O-O 16. Rc1 Rac8 17. a3 Bxd2 18. Qxd2 {with holes on the queenside (especially c5) that can be exploited by White's knight light-square bishop.}) 14... Rd8 $6 { again postponing castling, to my detriment.} 15. a3 $16 {taking away the b4 square from my pieces and preparing the next pawn advance.} O-O 16. b4 { I underestimated this. Now there is no good alternative to retreating the bishop to b6, but I felt that would just consign me to a hopeless strategic position. I was feeling aggressive so decided to sac a piece for two pawns instead. This is almost never the right decision, but it seems I need to better remember that.} Nxb4 $2 17. Bxg6 $18 fxg6 {in the hopes of eventually generating counterplay on the f-file.} 18. axb4 Bxb4 {a good example of a one-move threat that goes nowhere.} 19. Qb3 a5 {if my pieces were better placed, the two connected passed pawns on the queenside might provide compensation. However, White's pieces are much more active and mine still have to cover too many weaknesses.} 20. Rd1 Rxf3 $5 {given the desperate nature of the position, this is not a bad try, as it forces White to do some calculating on the defense, after his king position gets opened.} 21. gxf3 {one problem is that there is no good immediate follow-up. Here I decide to get my king off the diagonal and thereby free up the pinned d-pawn. This takes too long to allow me to make effective threats against White's king, however.} Kh8 22. Bg5 Rf8 23. Nxd5 Nxd5 24. Rxd5 Qxh3 {a last-gasp effort, although White could theoretically still go wrong here.} 25. Rd8 Rxd8 26. Bxd8 Bc5 27. Ra2 Qg3+ 28. Kf1 Qh3+ 29. Ke2 Qh2 30. Kd3 Qxe5 31. Re2 $2 {and he does go wrong! But under pressure, I miss how to use my queen.} (31. Bxa5 Qf5+ 32. Ke2 $18) 31... Qd6+ $2 {now it is effectively over.} (31... Qf5+ $1 32. Kd2 {I thought this would give White an out, but it doesn't. I suffered from "tunnel vision" and did not consider the possibility of a bishop check, only another queen check.} (32. Kc4 Qf7+ 33. Kxc5 Qxb3 34. Re8+ Qg8 35. Rxg8+ Kxg8 $11) 32... Bb4+ $1 {now White either has to repeat moves with Ke3 or lose the Bd8 to a queen fork.}) 32. Kc2 Qxd8 33. Qf7 $1 {White avoids the trap of being greedy.} (33. Qxb7 $2 $11 { now Black can escape the back-rank threat various ways, including ...Bf8.}) 33... Bf8 34. Re8 1-0

05 July 2021

Developing board sight


There are various definitions of "board sight" or "board vision" in chess, but fundamentally it means the ability to (mentally) see all of the various possibilities for moves by your pieces (and those of your opponent). While it is closely related to your level of tactical skill, it is not quite the same thing. In fact, I would say that, once you have been exposed to the full range of tactical ideas and patterns, whether through books like Understanding Chess Tactics or sites like Predator at the Chessboard, more often it is a failure of board sight in a particular position that will trip you up and cause you to miss a tactic. There are examples of this at all levels, including Kramnik's infamous missed mate-in-one.

One of the benefits of analyzing your own games is to see how particular issues repeatedly appear in them, which allows you to better correct them in future games. When looking at missed tactics, it is therefore very important to understand why you missed them - both for yourself or your opponent. In the latter case, I have long had a bias toward focusing on my own plans over the chessboard and not looking hard enough at the possibilities for my opponent; I have been working on this using a better thought process with tactics training. This expansion of mental focus, while helpful in general, will not necessarily eliminate board sight problems - although it should at least increase my chances to spot additional threats.

Here are some specific examples of "hard-to-see" tactical moves that can fool our board sight:

  • Backwards piece moves, especially by bishops and queens. It is more natural to focus on forward / attacking moves by long-range pieces, and miss the ability to retreat, even short distances. 
  • Horizontal piece moves, especially by rooks and queens. Again, there is a natural bias towards forward movement (toward the opponent), which may lead us to miss pins and attacks along the chessboard ranks.
  • "Impossible" moves by a piece onto a protected square. This can be a "naked sacrifice" where the moving piece looks like it will simply be captured, or a breakthrough sacrifice where for example a protected pawn is taken by a piece. In both cases, doing the usual math about how many times a square/piece is protected does not work, due to other tactics being present on the board.
The diagram at the top is a convenient illustration of all three of the above phenomena. Black has just snatched a pawn on c2 with his bishop, rarely a good idea in these kinds of positions. The key to the resulting tactic is a backwards queen move (Qf3-e2) which blocks in White's bishop. Moving a major piece a second time in the opening, while preventing the development of a minor piece, is not a normal occurrence, which may have helped bias Black when missing the move. The result is a double attack, horizontally against the Black Bc2 and vertically against the e7 pawn, which as a result is pinned. Afterwards Nc4-d6 is possible, threatening mate. So Black must guard against the mate threat and lose the bishop.

The obvious question, then, is how can we improve our board sight. Much of the general advice in that regard is along the lines of "practice more tactics" and "play a lot of games". It's hard to argue with these ideas, since greater experience normally translates in practical terms into better play - at least for a while. However, I would offer up the following practices as addressing more directly the issues involved with board sight failures, when you are actually playing a game:
  • CCT (checks, captures and threats). Incorporating a CCT scan into your thinking process will make you consider everything that is a forcing move on the board.
  • Status examination, as outlined in Understanding Chess Tactics. Essentially this means you will have an explicit awareness of what each piece can do and understand which are weak pieces and squares on the board.
  • Thinking on your opponent's time. This is more of an overall mental strategy, but the efficiency gains from doing the above two practices while your opponent is thinking, rather than doing things like calculating variations, will be significant.

13 June 2021

Commentary: FIDE World Cup 2017, Round 5 (Jobava - Yu)

Here we continue with the general theme of a b3/e3 development scheme for White, which appears here with an early c2-c4. White (GM Jobava) pursues a deliberate strategy of restraint, adopting a Hedgehog structure and maneuvering while waiting for his opponent to create a weakness. GM Yu Yangyi as Black does not oblige for most of the game, apparently being largely content to maneuver as well, with typical Hedgehog characteristics: enjoying a space advantage and eyeing the typically weak d-pawn, but not having any truly weak targets to go after. As can often occur, this relatively quiet maneuvering period ends with an explosive tactic, after some of Black's choices weaken his position subtly.

[Event "FIDE World Cup 2017"] [Site "Tbilisi GEO"] [Date "2017.09.08"] [Round "2.5"] [White "Jobava, Baadur"] [Black "Yu, Yangyi"] [Result "1-0"] [ECO "A01"] [WhiteElo "2687"] [BlackElo "2744"] [Annotator "ChessAdmin/Komodo Dragon 2"] [PlyCount "83"] [EventDate "2017.??.??"] 1. b3 d5 2. Bb2 Nf6 3. e3 g6 4. c4 {challenging the central pawn, taking advantage of White's control of d4 to prevent its advance.} dxc4 5. Bxc4 (5. bxc4 {is also possible, with the trade-off of controlling d5 with a pawn, but no longer having the half-open c-file to use, and isolating the a-pawn.}) 5... Bg7 6. Nf3 O-O 7. O-O c5 {asserting some central control.} 8. Be2 (8. d4 { is usually played here, directly challenging in the center. This leads to a very equal position. However, not necessarily a draw, as we can see from this high-level encounter:} cxd4 9. Nxd4 Bd7 10. Nd2 Nc6 11. N4f3 a6 12. Rc1 b5 13. Be2 Qb6 14. a3 Rfd8 15. Qc2 Rac8 16. Qb1 Ne8 17. Bxg7 Nxg7 18. Qb2 Bf5 19. b4 Ne6 20. Rc3 Bg4 21. Rfc1 Na7 22. h3 Bxf3 23. Nxf3 Rxc3 24. Rxc3 Rc8 25. Rd3 Rd8 26. Ne5 Rxd3 27. Nxd3 Qd6 28. Qc3 Qc6 29. Qe5 Qc7 30. Qd5 Kg7 31. Bf3 Nc8 32. Bg4 Nb6 33. Qe4 Nf8 34. Qd4+ f6 35. Nc5 e5 36. Qd3 f5 37. Bd1 Nc4 38. Bb3 Nd6 39. Qd5 e4 40. Qg8+ {1-0 (40) Nepomniachtchi,I (2784)-Dominguez Perez,L (2758) Lichess.org INT 2020}) 8... b6 9. d3 {now we see that Jobava is choosing to play with a Hedgehog type structure, which required the bishop retreat.} Bb7 10. Nbd2 Nc6 11. a3 {taking away the b4 square from Black's knight and completing the typical Hedgehog pawn structure.} Nd5 {centralizing the knight and leading to the exchange of White's better bishop.} 12. Bxg7 Kxg7 13. Qc2 { the queen gets off the first rank, connecting the rooks, and heads for the now-open long diagonal.} Qd7 (13... e5 {would be a more aggressive and imbalancing approach, setting up a more traditional central pawn presence immediately. Instead, Black maneuvers and prepares it for later.}) 14. Qb2+ f6 {blocking the diagonal and controlling the e5 square.} 15. Rfd1 {the Hedgehog structure typically calls for patient maneuvering and improvement of piece placement, which is what we see in the next several moves.} Rfd8 16. Rac1 Rac8 17. h3 {gives the king "luft" and controls g4. Also a useful waiting move.} e5 {Black now commits in the center, seizing space.} 18. Ne4 {centralizing the knight. It cannot be kicked with ...f5, due to the e5 pawn's weakness.} (18. Ne1 {is liked by the engines, as it reinforces d3 and threatens to activate the bishop via g4.}) 18... Qe7 19. Re1 {White continues patiently maneuvering. The rook overprotects the Be2 and lines up on the Black queen.} Rd7 (19... f5 { would no longer lose a pawn and would gain space, but would be committal on Black's part.} 20. Ned2 Qf6 $11) 20. Bf1 {gets the bishop out of the way of the Re1. It is not needed to protect the Nf3, so this helps activate the rook more.} Qd8 {moving the queen off the e-file and doubling up pressure on the d-file. This however seems slightly awkward, cutting off the Rc8 from moving along the 8th rank, but Black has nothing in particular to do at this moment.} 21. Rcd1 {reinforcing d3 again.} Nde7 $6 {the first slip-up by Black. Now White can advance his b-pawn and gain some space and activity.} (21... Nc7 22. b4 cxb4 23. axb4 a6 {and here the Nc7 controls b5, preventing the b-pawn from advancing further. As can be seen in the game continuation, this advance plays a significant role, attacking and driving away the Nc6.}) 22. b4 cxb4 (22... c4 $2 23. Qa1 cxd3 24. b5 $1 {this wins the knight, since if it moves, White will have the Nxf6 sacrifice, similar to the game.} Na5 $2 25. Nxf6 $18 Kxf6 (25... Bxf3 26. Nxd7 Bxd1 27. Nxe5 $18 {and White will regain the piece while having a much superior position.}) 26. Nxe5) 23. axb4 Nf5 {activating this knight and increasing control of d4.} 24. Rc1 {the rook is now free to do this, since the bishop covers d3, the minor piece serving a useful if limited purpose.} Qe7 { targeting the b-pawn, but essentially forcing White to play the best move.} 25. b5 $14 Na5 {the knight is now largely out of the action. The main effect is to relinquish control of the d4 square.} 26. Rxc8 Bxc8 {Black's pieces are now rather awkwardly placed and are not coordinating well.} 27. d4 {Black now has to exchange, in which case the game is still even, but he misses the threat.} ( 27. g4 $5 {is an alternative leading to some simplification.} Nd6 28. Nxd6 Qxd6 29. d4 $14) 27... Bb7 $2 (27... exd4 28. exd4 (28. Nxd4 $4 Qxe4 $19) 28... Qd8 29. d5 Rxd5 30. g4 Nd6 31. g5 {and the position is rather awkward for Black, despite being temporarily a pawn up, but it is better than the game continuation.}) 28. Nxf6 $1 $18 {a difficult-to-spot tactic, since it visually appears that the pawn is properly defended.} Kxf6 (28... Qxf6 29. dxe5 $18 { and now either the queen blocks on e6 or the discovered check wins the rook.} Qe6 30. Ng5 Qb3 {attacking the Qb2 almost works, but} 31. e6+ Qxb2 32. exd7 { and amazingly the d-pawn cannot be stopped without losing the queen.} Qd2 33. d8=Q Qxd8 34. Ne6+ $18) 29. dxe5+ Kf7 (29... Ke6 30. e4 $18) 30. e6+ (30. e4 { first would also work.}) 30... Kxe6 {now Black's king is exposed in the center, but it is still not so simple to convert the win.} (30... Qxe6 31. Ng5+ $18) 31. e4 {the only winning move, kicking the knight and seizing more squares in the center.} Ng7 (31... Nd6 32. Qa2+ Kf6 33. e5+ Kg7 34. exd6 Qxd6 35. Re6 Qf4 36. Qa1+ Kf7 37. Re3 {and Black cannot stop all the threats to the king, including Ne5+}) 32. Qa2+ Kd6 (32... Kf6 $2 33. e5+) 33. Rc1 $1 {this is a good example of how to conduct a king hunt properly. White first cuts off the escape route, before attempting to directly attack again.} Bxe4 $2 {this looks like a reasonable try, but now White goes after the king.} (33... Qd8 {is the best defence, but it also loses to} 34. Qa3+ Ke6 35. h4 h6 36. Qe3) 34. Qd2+ Ke6 35. Ng5+ Kf5 (35... Kf6 36. Nxe4+ Qxe4 37. Qxd7 $18) 36. g4+ Kf6 37. Nxe4+ Qxe4 38. Qxd7 {White is now up the exchange with a commanding position.} Ne6 39. Bg2 {the bishop finally sees some action.} Qd4 {hoping that White will exchange into a (still winning) endgame, but Jobava can simply increase his advantage.} 40. Qxh7 Nf4 {hoping for a cheapo fork on e2.} 41. Qh8+ Ke6 42. Re1+ {White does not fall for the fork trap, so Black gives up.} 1-0