02 July 2022

Commentary: U.S. Women's Championship 2021, Round 9 (Krush - Nemcova)

This game features another ambiguous opening classification, since the database will tell you it's an English Opening (A10 ECO code), but one look at the board on move 7 will tell you that it's a Leningrad Dutch. GM Irina Krush as White kept her full intentions in the opening hidden until that point, but after her opponent WGM Katerina Nemcova committed to a full Leningrad setup, there was no reason not to play d4 and control the e5 square, especially after having done the early b3/Bb2 development.

From there Krush gains an small opening advantage, thanks to Black neglecting her development in favor of some premature demonstrations (7...Ne4 and 8...c5) that do not actually challenge White. Krush masterfully rides this advantage into the middlegame, although she seems to deliberately choose solid over sharper possibilities, in keeping with her general opening posture. Black's strategic weakness on d6 becomes the key feature of the game, leading White eventually to gain tactically. That said, it's worth observing that even when behind and under pressure, opportunities often present themselves - see move 31 - for the worse-off player to rally. Normally this is a feature of long endgames as well, but Krush never lets her opponent back into the game after entering a R+B v R endgame, which is instructive to see.

I did not post an evaluation chart this time, because the one generated was misleading and displays Black achieving equality (and more) around move 22, whereas longer engine analysis shows a persistent White plus. This sometimes happens with the "snapshot" type evaluation function of various programs/sites, so you should always be somewhat skeptical of anything insta-generated by a computer, until you can perform your own analysis.

[Event "U.S. Women's Chess Championship 2021"] [Site "http://www.chessbomb.com"] [Date "2021.10.16"] [Round "09"] [White "Krush, Irina"] [Black "Nemcova, Katerina"] [Result "1-0"] [WhiteElo "2429"] [BlackElo "2331"] [EventDate "????.??.??"] [ECO "A10"] [PlyCount "135"] [BlackClock "0:03:47"] [BlackFideId "322750"] [TimeControl "5400+30"] [WhiteClock "0:25:41"] [WhiteFideId "2012782"] 1.c4 f5 2.g3 Nf6 3.Bg2 g6 {committing to the Leningrad Dutch formation.} 4.b3 {this commits the bishop to develop to b2 and then to immediately fight its Black counterpart on the long diagonal, including for the key square e5.} 4...Bg7 5.Bb2 O-O 6.Nf3 d6 7.d4 {White finally commits her pawn to d4, but only after Black was threatening to get in ...e5. We now have a reasonably standard Leningrad Dutch position, where White has chosen an early b3/Bb2 development scheme.} 7...Ne4 {This is the third most played choice in the database, although the standard ...c6 and ...Qe8 are more popular. It doesn't score very well for Black, though, and seems premature.} 8.Nbd2 {the usual location for this knight in the variation, leaving the Bb2 unblocked.} 8...c5 {Black challenges the center, but is falling behind in development.} ( 8...Nc6 $5 ) 9.Qc2 {the usual spot for the queen, protecting b2 and pressuring e4.} 9...cxd4 {this exchange helps White, who has everything covered and more pieces developed.} 10.Nxd4 Qb6 11.e3 {a solid choice preserving White's central edge and restraining f5-f4.} ( 11.Nxe4 $5 {is the engine's preference and here seems a relatively simple path to advantage, limiting Black's counterplay. The knight is inadequately supported on e4, a fact which informs the next few moves of the game.} 11...Bxd4 ( 11...fxe4 {this cedes the pawn for no real compensation.} 12.Qxe4 Qa5+ 13.Kf1 $16 {the White king is well-protected and the Rh1 can still participate in the game by supporting an h-pawn push.} ) 12.Bxd4 Qxd4 13.Nc3 $16 {and after some straightforward exchanges, White has a significant lead in development and active play in both the center and kingside against Black's weak squares.} ) 11...e5 {the rule in the Leningrad Dutch is to play ...e5 whenever Black can get away with it. Here it is a little premature, because of the unstable Ne4.} ( 11...Nf6 {the engine assesses the knight retreat is best.} ) 12.Nb5 {this avoids complications and leaves White with a small plus.} ( 12.Nxe4 {immediately still works, but is messier than in the previous variation, since now the Nd4 is attacked by a pawn.} 12...exd4 13.Ng5 {and now} 13...dxe3 $2 {looks dangerous, but it is Black that is in trouble after something like} 14.Bd5+ Kh8 15.O-O-O exf2 16.Qe2 $18 {White is threatening to penetrate with Qe7 and Nf7 cannot be defended against, so Black will have to give up the exchange.} ) 12...a6 {the obvious choice, kicking the knight.} 13.Nxe4 {now White pulls the trigger on the Ne4.} 13...axb5 14.Nc3 bxc4 15.Nd5 {an excellent outpost for the knight.} 15...Qd8 16.Qxc4 Be6 {Black starts repairing her development, While pinning the knight and blocking White's tactical ideas along the diagonal.} 17.O-O Nc6 18.Rfd1 {naturally White wants to exert latent pressure on the d-file, eyeing the backward d6 pawn.} 18...Rf7 {the immediate ...e4 may have been better, waiting to commit the rook to a course of action.} 19.b4 {White intends to mobilize her queenside pawn majority.} 19...e4 {cutting off the Bg2 and opening up the long diagonal.} 20.Bxg7 {the exchange is best, as otherwise White will have to take time to defend the bishop.} 20...Kxg7 21.a4 {White could have done some other things here, for example immediately reactivating the bishop with Bf1. The engine points out a nice tactical maneuver possibility by Black that could have lead to equality, in response.} 21...Ne5 $6 {now White's queen gets out of the pin and is still powerfully centralized.} ( 21...Qf6 $5 {the point being that} 22.Nxf6 Bxc4 23.Nd5 Bb3 {and Black is fine, being able to effectively trade the weak d-pawn for one of White's queenside pawns. Otherwise, Black as a follow-up can play ...Qe5 and/or trade on d5 and be equal.} ) 22.Qd4 $16 Bxd5 23.Qxd5 Rd7 {apparently an idea behind the original ...Rf7, but Black is getting cramped defending her weaknesses.} 24.Bf1 Qg5 $6 {superficially aggressive but Black does not have an attack. White calmly defends, allowing Black to overcommit her forces, then strikes back in the center and queenside.} ( 24...Qg8 ) 25.Be2 h5 26.h4 Qe7 27.Kg2 {covering some holes and allowing the first rank to be cleared.} 27...Rc8 28.Rd2 {intending to double rooks on the d-file.} ( 28.Rac1 {seems more straightforward. Exchanging off a pair of rooks would be advantageous to White, who will have a winning 2-1 queenside majority in the endgame. If Black avoids the trade, then White has the file and can start pushing the pawns anyway.} ) 28...Qf7 29.Qd4 {White has the better queen so avoids the trade.} 29...Qf6 30.Rad1 d5 {at least this gives Black a little more room for maneuver.} 31.a5 ( 31.b5 $5 {would prevent ...Nc6.} ) 31...Rcd8 $2 {going for static defense, which does not work.} ( 31...Nc6 {gains a tempo and solves some of Black's problems in the center. For example} 32.Qb6 Ne7 $14 {and Black is close to equality, as the knight can hold d5 by itself, freeing up both rooks.} ) 32.b5 {now the Ne5 is out of the important central fight.} 32...Nf7 33.Qc5 Qd6 34.Qxd6 {White chooses to simplify closer to a winning endgame.} 34...Rxd6 35.a6 bxa6 36.bxa6 $18 {The Be2 plays a key supporting role for the pawn, even if not active otherwise.} 36...Ne5 {the knight tries to get back in the fight, but it is too late.} 37.a7 {passed pawns must be pushed!} 37...Nc6 38.Rxd5 $1 {White finds the correct tactical follow-up. With the pawn about to queen, Black cannot take twice on d5.} 38...Rxd5 39.Rxd5 Ra8 40.Rc5 Nxa7 41.Ra5 {the knight is now inevitably lost, since White can bring around the bishop to attack the Ra8.} 41...Kh6 42.Ra6 Kg7 43.Bc4 Rc8 44.Rxa7+ Kf6 {the game is now a theoretical win for White. Black knows that it still takes practical skill to win the R+B v R ending, so she keeps playing. Krush is up for it, however.} 45.Ra6+ Kg7 46.Bb5 Rb8 47.Ba4 {the bishop is now shielded from the rook's attention, as ...Rb4 is met by Be8, winning the pawn.} 47...Kf7 {the king is tied to the defense of the g-pawn.} 48.Rd6 Kg7 49.Kf1 {White decides to march her king around, which Black permits. However, she cannot really stop the idea.} 49...Kf7 ( 49...Rb1+ 50.Rd1 ) 50.Ke2 Rb4 51.Bc2 {bringing it back as a shield for the king.} 51...Rb2 52.Kd2 Kg7 53.Kc3 Ra2 54.Rb6 {note how Black's rook is beginning to get cramped.} 54...Ra1 55.Rb1 {naturally Black has zero chances if she gives up the rook.} ( 55.Rb7+ {it's interesting to see how White could have been more aggressive if she chose. The problem for Black is that her king is potentially vulnerable, without g5 and h4 as escape squares. So for example} 55...Kf6 56.Kd4 Rf1 $2 57.Bb3 {and Black has to give up material.} ) 55...Ra2 56.Rb2 Ra3+ 57.Bb3 {a good opportunity to place the bishop on this key diagonal.} 57...Ra1 58.Ra2 Rb1 59.Bc4 Rd1 60.Ra6 Rd8 61.Bb5 Kf7 62.Ra7+ Ke6 63.Rg7 Kf6 64.Rd7 Ra8 65.Kd4 {the centralized king is the significant difference from when the piece maneuvering all began. White is patiently constricting Black's maneuvering room and bringing up her king as a strong reinforcement.} 65...Ke6 66.Kc5 Ra5 $2 {this allows a tactical finish, forcing the rook trade.} 67.Rd6+ Kf7 68.Ra6 1-0

01 July 2022

Book quote #2: Pale Gray For Guilt

From Chapter Seven of Pale Gray For Guilt by John D. MacDonald:

Meyer came over on Christmas morning with a cumbersome vat of eggnog and three battered pewter mugs. We had a nice driving rain out of the northwest and a wind that made the Flush shift and groan and thump. I put on Christmas tapes because it was no day to trust FM programming. Sooner or later daddy would see mommy kissing Rudolph. Meyer and I played chess. Puss Killian, in yellow terry coveralls, sat and wrote letters. She never said who they were to, and I had never asked.

He won with one of those pawn-pressure games, the massive and ponderous advance that irritates me into doing the usual stupid thing, like a sacrifice that favors him, just to get elbow room on the board. 

30 June 2022

Book quote: Pale Gray For Guilt

 From Chapter Two of Pale Gray For Guilt by John D. MacDonald:

Meyer came out of a long and somber contemplation, hunched like a hirsute Buddha, reached a slow ape arm and picked up his queen's bishop and plonked it down in what at first glance seemed like an idiotic place, right next to my center pawn. A round little lady who was one of his retinue that week beamed, clapped her hands and rattled off a long comment in German.

"She says you give up now," said Meyer.

"Never!" said I. I studied and studied and studied. Finally I put a knuckle against my king and tipped the poor fellow over and said, "Beach-walking, anyone?" 


26 June 2022

Video completed: A sharp Slav vol. 1

I recently completed Andrew Martin's "A sharp Slav, vol. 1" 60-minute ChessBase video. This particular Slav Defense variation - responding 3...dxc4 after 1. d4 d5 2. c4 c6 3. Nc3 - I've in fact always played, but was never fully satisfied with the different lines I used previously. Martin focuses on the aggressive 4...b5 response to anything (normally 4.e3 or 4. e4) except 4. a4, which I believe is the correct way to play. Here's the full contents list:

One of the reasons 4...b5 has been demonstrated as preferable is super-GM Hikaru Nakamura's use of it. He in fact is featured in the first model game, which contains an unusual and possibly busted sideline for White (4. e3 followed by 5.a4 b4 and the strange-looking 6. Nce2). The general point of 4...b5 isn't to hold onto the extra c-pawn forever, but to chase the knight on c3 and make White have to work hard to recover the pawn, while Black can play actively and develop. The first game (Mamedyarov-Nakamura) is a great example of this. White in fact had several options to secure a draw, but passed them up and Black won with sharp play (going back to the title of the video).

The essential soundness of this aggressive approach with the queenside pawns is illustrated by the next model game provided, featuring Botvinnik as Black in 1933 (using an older line for Black), as well as the remainder of the lines in all main variations (4.e3, 4.e4 and 4. a4). Martin looks at them in a fair amount of (rapid) detail, while pointing out common themes/ideas, including things like Black's need to watch out for tactical threats from White on the long diagonal after playing ...Qd5. 

There are a lot of similarities between the various lines, whose differences hinge largely on where White chooses to put the Nc3 after it is attacked. White's 4. e4 would seem to be the obvious choice, seizing the center, but unlike with the 4. e3 variations, the Nc3 no longer has e4 available to go to; that is a significant trade-off, as the knight is not very happy either retreating to b1 or a2, having to lose significant time to get back in the game.

Some other commentary on the contents:

  • This is not a repertoire for Black (or White), as Martin presents multiple options for variations, and within them as well. He usually signals what he prefers, but he also mentions which ones he considers playable if not preferred.
  • Martin presents each game somewhat quicker than normal, I would say, which I expect is due to the limited total time format. I found myself going back several times in each individual video to review particular lines, rather than being able to keep up in real time, but that's not a terrible thing necessarily.
  • It's very helpful to see the full games in each case. This is not an opening theory product, rather one that's intended for training, familiarization and understanding. In addition, this is not an opening where either side is going for an early knockout blow, so it's important to see how the middlegame (and sometimes endgame) can flow from the opening.
  • Favorite quote: "4. a4 might be played by cowardly opponents who do not want to brave the complications after ...b5."

19 June 2022

Commentary: U.S. Women's Championship 2021, Round 8 (Eswaran - Tokhirjonova)

One finds that analyzing master-level games often leads to multiple insights. This one, at 148 moves, is no exception. A few top-level takeaways:

  • The game's opening could be classified several different ways, which to a purist would be horrific. This however helps illustrate how openings are often fluid, rather than rigid constructs. One of the things that has helped increase my strength over the years is a less-rigid view of opening play and an acceptance of the fact that your opponents will follow "proper" book lines much less often than opening book writers imply. Understanding different opening structures and their characteristics is much more important than adhering to a specific move order - except, of course, when you can play some useful tricks with different move orders.
  • Positional advantages - in this case, White gets one out of the opening - are nice to look at, but in themselves are not decisive. They can evaporate, as White's does, after which Black finally seizes the initiative.
  • Patient maneuvering can be the key to winning without the presence of forcing play. Here this was evident in both the middlegame and endgame phases, with White first losing an advantage, then getting herself trapped into an unfavorable rook exchange, which was decisive.

[Event "U.S. Women's Chess Championship 2021"] [Site "http://www.chessbomb.com"] [Date "2021.10.14"] [Round "08"] [White "Eswaran, Ashritha"] [Black "Tokhirjonova, Gulrukhbegim"] [Result "0-1"] [WhiteElo "2244"] [BlackElo "2322"] [ECO "A48"] [Annotator "ChessAdmin/Dragon 2.6.1 by Komodo"] [BlackClock "0:04:44"] [BlackFideId "14203626"] [TimeControl "5400+30"] [WhiteClock "0:00:58"] [WhiteFideId "2080788"] 1.Nf3 Nf6 2.d4 g6 3.e3 Bg7 4.c4 c6 {a rare setup from Black at this stage, who now goes for a Slav Defense structure with a fianchetto, which is called the Schlechter variation. However, the opening can be reached by different paths and can be classified in various ways.} 5.Be2 d5 6.O-O O-O 7.Nc3 {for example, this is now classified as a Grunfeld Defense, Burille Variation by the Chess.com Explorer, while the PGN download of the game calls it a King's Indian, East Indian Defence (ECO code A48). The key points of the structure, regardless, of how it is classified, is that White has developed with e3/Be2 instead of fianchettoing the bishop, while Black has the c6/d5 pawns and has fianchettoed her dark-square bishop.} 7...Ne4 {another relatively rare move. Black moves the same piece twice before developing further, which is against general opening principles and probably explains why it is less popular. The two main choices are:} ( 7...Bg4 ) ( 7...dxc4 ) 8.Qc2 ( 8.Qb3 {this scores better and is more active than the text move, placing the queen on the more useful b-file and a2-g8 diagonal.} ) 8...Nxc3 {Black chooses a safe and somewhat passive route.} ( 8...Bf5 $5 {is a more interesting and active choice. However, Black would need to be prepared to have an uglier pawn structure.} 9.Bd3 Nxc3 10.Bxf5 gxf5 ( 10...Nxa2 $6 11.Rxa2 gxf5 12.Qxf5 e6 13.Qh3 {and now that White's queen has been freely transferred to the kingside, it may only be a small advantage according to the engine, but White certainly has the easier game and more winning chances.} ) 11.bxc3 e6 12.Ba3 Re8 13.cxd5 cxd5 14.c4 {with a slight edge for White.} ) 9.bxc3 {the doubled pawns are only temporary, as the c4 pawn can always be exchanged for the d5 pawn.} 9...e6 {trying to be ultra-solid, apparently. First time this has appeared in the database, with ...b6 and ...dxc4 previously played.} 10.Ba3 {certainly a very logical way to develop the bishop, as the diagonal is excellent and it has nowhere else useful to go. Komodo Dragon prefers plans involving an a4 pawn push, exploiting White's space advantage on the queenside and further restricting Black.} 10...Re8 11.Rab1 $14 {putting the rook on the half-open file and restraining Black's b-pawn from advancing to b5. By this point White's advantage in development has given her a small but persistent plus.} 11...Nd7 12.cxd5 {this may be a little premature and also helps un-cramp Black's structure to some degree. However, White still has a freer game afterwards.} ( 12.h3 $5 {is an interesting prophylactic idea by the engine, the idea being to prevent an eventual ...Bg4.} ) 12...exd5 13.c4 dxc4 {this abdicates Black's central pawn presence without a fight.} ( 13...Nb6 14.cxd5 cxd5 {is the engine's preference, ending up in an IQP position. This may have been what turned Black off to the idea.} ) 14.Bxc4 {White's lead in development and positional plus is obvious. The two bishops have excellent diagonals and the Rb1 is well-placed. That said, White does not have any immediate threats.} 14...Nb6 {clearing the way to develop the Bc8 and kick the Bc4.} 15.Bd3 Be6 ( 15...Bg4 {seems more natural, threatening to exchange off the knight and compromise White's kingside pawn structure.} ) 16.Bc5 {increasing pressure on the queenside.} ( 16.Rfc1 $5 {getting the other rook into play might be better first.} ) 16...Qc7 {this seems like a normal move, protecting b7 and connecting the rooks, but there are lurking issues with it.} ( 16...Nd7 17.Rxb7 Nxc5 18.dxc5 Re7 {is a pawn sacrifice suggested by the engine, where Black now has the two bishops and a freer game in compensation. If eventually a 4v3 pawn rook endgame is reached, it could be drawn.} ) 17.Ng5 {looking to obtain the two bishops after a piece exchange.} 17...Bc8 {awkward, but the best way of preventing the exchange.} 18.Ne4 {Eyeing the weak d6 square. White has a comfortable game and a positional plus, with better control of the center and more active pieces.} 18...Nd5 19.Rfc1 ( 19.Nd6 $5 {White could continue the previous idea and force the piece exchange now. However, with the text move she keeps her own centralized knight.} ) 19...Rd8 {regaining control over d6.} 20.a4 {not having an obvious way to make further progress, White decides to grab some space.} ( 20.h3 {is again a prophylactic idea.} ) ( 20.Rb3 $5 {would allow for potential doubling of the rooks on the b-file and also prevent a tactical idea of Black's, as we can see in the next variation.} ) 20...Bf5 {the bishop re-activates itself.} ( 20...b6 $5 {is the tactical idea.} 21.Ba3 Bf5 {Black can allow capture of the c-pawn, due to the pin on the Ne4 once the queen is gone.} 22.Qxc6 ( 22.Rb3 $5 ) 22...Qxc6 23.Rxc6 Re8 24.Rd6 Nf4 25.exf4 Bxe4 26.Bxe4 Rxe4 $10 ) 21.Ba3 {without the Nb6 as a target, the bishop was not doing as much, so retreats.} 21...Rab8 {Black does not find the ...b6 idea and more passively defends.} 22.Rb2 ( 22.Rb3 {seems like a better placement, providing protection along the third rank.} ) 22...a5 {this freezes the White a-pawn and gives Black a potential outpost on b4. However, as with all pawn advances, it leaves some weaknesses behind.} 23.h3 {restricting the Bf5.} ( 23.Nc3 {an alternative would be to initiate exchanges and highlight White's pressure on the queenside.} 23...Bxd3 24.Qxd3 Nxc3 25.Qxc3 $14 ) 23...h5 {preventing g2-g4, but again leaving some holes behind.} ( 23...b6 $5 ) 24.Nc5 {now one pair of pieces are exchanged.} 24...Bxd3 25.Qxd3 Bf8 {putting the bishop on a better diagonal.} 26.Qb1 {it might have been more flexible for the queen's placement to use Rc1-b1 to double on the file. This would also help deter Black from opening the b-file.} 26...Qc8 {the point of this move is to cover the a6 square to allow Black's b-pawn to advance, otherwise there would be a knight fork on a6.} 27.Qa2 {this protects the Ba3.} 27...b5 {now Black finally gets some counterplay rolling, although if anything White still has a slight edge.} 28.axb5 $6 {this immediate resolution of the tension is unnecessary and makes Black's queenside pawns much more threatening. White may have been uncomfortable with the threat of ...b4, but Black will have that possibility anyway.} 28...cxb5 29.Nd3 $6 {now Black firmly takes the initiative, exploiting White's cramped pieces.} ( 29.Rbc2 ) 29...Nc3 30.Qa1 ( 30.Rxc3 $5 {Komodo Dragon favors this exchange sacrifice, leaving White more active after} 30...Qxc3 31.Ne5 Qc7 32.Bxf8 $15 {with threats to f7, g6 and a fork on c6 helping White regain the initiative, even if down material.} ) 30...b4 $19 {although Black does not immediately win the Ba3, White is in serious trouble.} 31.Rbc2 Qf5 {targeting the hanging Nd3 and moving out of the pin on the c-file. Now White avoids losing material, but Black gets a rook on the 2nd rank, with a stranglehold on the position.} 32.Ne5 bxa3 33.Rxc3 Rb2 {the most immediate threat is to f2, but the advanced a-pawn combined with Black's rooks is also a major problem.} 34.Rf1 Rdb8 35.Nd3 Rd2 ( 35...Bb4 {would more aggressively activate the bishop.} 36.Nxb2 Bxc3 37.Qxa3 Bxb2 $19 ) 36.Ne5 Rbb2 37.Nf3 Rdc2 {White has nothing better than to give up material.} 38.Rxa3 Bxa3 39.Qxa3 Qb5 40.Qe7 {the best try. White hopes to combine her queen and knight and take advantage of the weak square complex around Black's king.} 40...a4 {Black goes with the straightforward plan of ramming the a-pawn down the file.} 41.d5 {an excellent practical move by White, who must have hoped to distract Black, and succeeds. The engine shows White cannot queen the d-pawn, but it must have looked menacing.} 41...Qxd5 $2 ( 41...Rb1 42.Rxb1 Qxb1+ 43.Kh2 Qb8+ 44.d6 Qf8 {and Black wins.} ) 42.Qe8+ {now White is back in business, forking the a-pawn with her active queen.} 42...Kg7 43.Qxa4 {White now has high hopes for a draw, with a pawn for the exchange and a reasonable position.} 43...Ra2 44.Qf4 ( 44.Qd4+ Qxd4 45.Nxd4 {after simplification there seems little prospect for a Black breakthrough. Perhaps White did not trust her endgame skills, however.} ) 44...Qc4 45.Qe5+ $10 {the engine evaluates the position as equal. For Black, though, there is no reason not to play on, as long as she is careful not to blunder.} 45...f6 46.Qe7+ Qf7 47.Qd6 Qc7 48.Qd5 Ra5 49.Qd1 Rca2 50.Nd4 {Black cannot expel the knight from its central outpost.} 50...Qc4 51.Qf3 Qd5 {Black would be happy to trade off White's mobile queen, which has better access to her opponent's king.} 52.Qd1 {perhaps hoping for a repetition. However, Black is not interested.} 52...Kh6 53.Qb1 Qa8 54.Qb6 {an inaccuracy. This is a characteristic of queen endings especially, where both sides can go astray more easily.} ( 54.Qb4 Rg5 55.Nf3 Qxf3 56.Qf8+ {and draws.} ) 54...Rg5 55.Qc6 {forced} 55...Qxc6 56.Nxc6 Rb5 ( 56...h4 $15 {is the plan preferred by the engine, with the idea of following up with ...g5 and ...Kg6, to work Black's space advantage and increase the pressure. With the queens off the board, Black has nothing to fear.} ) 57.Nd4 Rbb2 {looking to safely tie up White with pressure along the 2nd rank, but Black has nowhere to go from there.} 58.h4 Kg7 {Black chooses to swing her king around, the long way, without moving her pawns.} 59.g3 ( 59.Ne6+ Kf7 60.Nf4 $10 {White's knight is very active and holds the kingside steady from here.} ) 59...Kf7 60.Kg2 Ke7 61.Nf3 Ke6 62.Ng1 {White effectively is marking time with the knight.} 62...Kf5 63.Nf3 Ke4 64.Nd4 Rd2 65.Nf3 {this allows Black to up the pressure and for White to lose the thread. Other knight moves were less risky.} ( 65.Nb3 ) ( 65.Ne6 ) 65...Re2 66.Ng1 Rxe3 {the engine shows no change in evaluation, but now White has a pawn's less worth of a margin of error.} 67.Rb1 Ke5 68.Rb5+ ( 68.Nh3 $5 ) 68...Kd6 69.Rb6+ Ke7 70.Rb7+ Kf8 71.Rc7 Re7 ( 71...g5 $1 {The point is that Black will be able to play ...g4, threatening to shut out the knight and preventing the king from escaping the rooks. For example} 72.hxg5 fxg5 73.Nh3 g4 74.Nf4 Re1 $19 {White's R+N cannot mate Black's king and eventually Black will break through.} ) 72.Rc6 Kf7 73.Nh3 Re5 ( 73...Ra5 {would accomplish the same thing on the 5th rank, while keeping the 7th rank covered.} ) 74.Nf4 ( 74.Rc7+ ) 74...Ra7 $17 {now preventing annoying rook checks. Black still has a clear edge in material, but a lot of maneuvering now ensues, with her trying to patiently make something of it.} 75.Rb6 Rf5 76.Rc6 Rb7 77.Ra6 Rd7 78.Ra2 Kg7 79.Rb2 Ra7 80.Rc2 Kh6 81.Rc6 Ra2 82.Rb6 Ra4 83.Nh3 Re4 84.Ra6 Kg7 {Black gets tired of making rook moves, but this should allow White to equalize.} 85.Ra7+ Kg8 86.Ra8+ ( 86.Nf4 {is pointed out by the engine, since Black cannot protect the g-pawn.} 86...g5 87.hxg5 Rxg5 {Black now has two weak pawns to look after.} 88.Ra8+ Kg7 89.Ra7+ Kh6 90.Kh3 {and White appears to hold.} ) 86...Kf7 87.Ra7+ Re7 88.Ra6 {trading rooks would be bad, since White's rook activity is one of the main things keeping Black in check.} 88...Kg8 89.Nf4 Kh7 90.Rb6 Ra7 91.Rc6 Kg7 92.Rb6 Kf7 93.Rb2 Rfa5 94.Rb6 Rd7 95.Rb2 Ra4 96.Rb5 Ra2 97.Rc5 Rdd2 98.Nh3 Rdc2 99.Rb5 Re2 100.Rb7+ Re7 101.Rb5 Re5 102.Rb7+ Ke6 {Black does not want to keep doing the same maneuvers and brings her king forward.} 103.Nf4+ {getting the knight back into the game and reminding Black about the g6 weakness.} 103...Kf5 104.Rf7 Rc5 105.Rf8 Rcc2 106.Nh3 Ke6 107.Rb8 Rcb2 108.Rc8 Rc2 109.Rb8 Rc7 110.Rb5 Rac2 111.Rb8 Kf5 112.Rg8 $2 {White finally loses patience with her own rook maneuvers and allows it to be traded off, leading to a won endgame for Black.} ( 112.Nf4 ) 112...Rc8 $1 113.Rg7 R2c7 114.Rxc7 Rxc7 $19 115.Nf4 {although the knight is no longer tied to defending f2, without White's rook available to harass Black's king, it now becomes a formidable piece in combination with the rook. After some maneuvering, Black can then support a pawn break.} 115...Rc2 116.Nd5 Ke6 117.Nf4+ Kf7 118.Kf3 Kg7 119.Kg2 Kh6 120.Nd5 Rc6 121.Ne3 g5 122.hxg5+ fxg5 123.Kf3 Rf6+ 124.Ke2 Kg6 125.Ng2 Kf5 126.Ne3+ Ke4 {the power of the centralized king is evident.} 127.Ng2 Ra6 {the rook can do more against White targets operating from the side.} 128.Kf1 Ra1+ 129.Ke2 Rb1 130.f3+ Kd4 131.Kf2 Rb2+ 132.Kf1 Kd3 133.f4 g4 {the f-pawn poses no threat, while control of f3 and h3 is now had.} 134.Nh4 Ke3 135.Kg1 ( 135.Nf5+ Ke4 136.Nh4 Rb3 {and White is soon in zugzwang.} ) 135...Ra2 136.f5 Ke4 ( 136...Rf2 ) 137.Kf1 Ke5 138.Kg1 Rb2 139.Kf1 Rh2 {forcing a winning position for Black.} 140.Kg1 Rxh4 {when you can force a winning endgame position, material balance is irrelevant.} 141.gxh4 Kxf5 142.Kf2 Kf4 143.Kg2 Ke4 144.Kf2 ( 144.Kg3 Ke3 145.Kg2 Kf4 {and wins.} ) 144...Kd3 145.Kg2 Ke2 146.Kg3 Kf1 147.Kf4 Kf2 148.Kg5 g3 0-1

Evaluation chart generated by HIARCS Chess Explorer Pro