A very interesting new idea and concept. White tries to remain flexible as regards the development of his knight. White challenges Black to find a usefule other than the critical dxc4 for which Re1 is well suited to protect e4. If Black is going to take c4, then White;s knight is better placed on b1 than c3 as potentailly it will not be attacked by black's pawn pushes b5-b4.7.Nc3dxc48.h3Na69.e4b510.Qe2Bb711.Rd1Nc712.Bf4Ne613.Be5Qb614.d5Nc515.Bd4Nfd716.b4cxb317.axb3a518.Rac1Rac819.Be3Bxc320.Rxc3cxd521.exd5b422.Rcc1Ba623.Qa2a424.bxa4b325.Qd2Qd626.Bf4Qf627.Bg5Qd628.Bf4Qf629.Bg5½-½ Romanishin,O (2585)-Anand,V (2715) PCA-Wch Candidates m2 New York 1994 (1)7...dxc428:148.e4N7Bg45:049.Nbd29c5!?4:10 A very nice idea, where Black is transforming the structure into a Benoni where he has exchanged his problem piece- light squared bishop.9...b510.h3Bc8is an idea, which Ding would have considered, but ofcourse would not have felt fully comfortable.10.d57:37e63:4611.h36:27Bxf31212.Bxf31:46exd54:2313.exd525Nbd78:36 Black transformed a Neo Gruenfeld into a Benoni14.Nxc41:25b5315.Na3!1:19 outwardly not a great square. But White holds back black's queenside majority and later on the knight gets activated via the c2-square. Anish Giri suggested that all this could have very well been part of Gukesh's prepration!Qb65:0416.Bf42:19 White's Bishop Pair make a formidable impression.Rfe813:4017.Qd252Rad82618.Nc217:55Nf836 Usually with a pawn on d5 for the opponent, this is a not a good square for the Knight was the observation of top players.18...Rxe1+19.Rxe1Ne8was suggested by Anish and seems more natural.19.b4!18:30 This was a typical idea in the Benoni, but as Magnus pointed out, allowing a protected passed pawn in the centre is a committal decision and was not easy at all.c414:2620.Be32:51Qa62:1721.Bd46:06Rxe1+7:0821...N8d7is apparently Black's best move, but it is a very hard decision to admit one's mistake and come back to the earlier square!22.Qf4Qa4!as pointed out by Anish gives Black counterplay22.Rxe15Qxa2!?5 For the World Champion a big time trouble was looming and he decided to force things through trusting his intuition and grab the pawn on a2!23.Ra16:42Qb34Apparently the complex23...c3!retains a sort of balance as various commentators pointed out.But with limited time on the clock, this was extremely hard to come up with.24.Ra337Qb1+525.Kg236Rd71:2026.Ra5!20:12 Gukesh took 20 minutes for this wonderful move, which apparently the computer was screaming from the word go! The position was very complex and Gukesh came up with objectively the best move, even though it was far from obvious for various commentators.I was thinking about26.Bc3threatening Ra1 and Nd4 trapping the Queen, but Black hasNe6!and after which it is not easy to make progress.27.Ra1Qb328.Nd4Nxd429.Bxd426...Qb35027.Ra31:11Qb1428.Ra55:41 Gukesh goes back to Ra5 after repeating once, but Nakamura criticized this decision. He was not gaining time as there was no increment, but he let Ding come one move closer to move 40.Qb3429.Rxb510Qd37
30.Qf431 The most natural move and a good one for sure!But White has a much stronger even a decisive continuation at his disposal it seems with30.Be3!Qxd231.Bxd2+-White is not even ahead in material. But his pieces are dominating. He has the Bishop Pair and more importantly than that, Black's pieces, especially the Knight on f8 is totally passive and lacking any prospects. Black's pawn on a7 particularly and to a certain extent the one on c4 also can be a weakness, but White's pawn on d5 and b4 are not weaknesses at all. But playing such non forcing moves, when one side is having a big initiative and looking for a killer blow, is extremely difficult in human chess!30...Qxc26:5431.Bxf65Qf5432.Qxf5?!1:22A very important and instructive moment pointed out by Anish. Here, more than the doubled pawns, the mobility of the pieces is the priority. The knight on f8 needs a chance to come into the game, and this is helped by doubling the f-pawns and freeing up the g6-square for the knight.32.Bg5!Qxf433.Bxf4c334.Rc532...gxf5233.Bxg7?!533.Bg5!Even here White had to maintain the bishop pair, and the automatic capture on g7 was criticised by all commentators. But, in the heat of the battle and in time trouble, some moves are done automatically, and this might be one such moment.33...Kxg7234.Rc525Ng6135.Rxc41:13Ne5136.Rd449Nc630 The knight has become totally activated and even though Black is down a pawn, there is a definite improvement from his earlier situation.37.Rf42:2037.Rd2Nxb4Perhaps Black should still be looking for a blockade on d6.37...Ne7!?38.d6Kf639.Be2!was a suggestion of Maurice Ashley at the press conference. Clearly this was also far from obvious and missed by players too. Black is not able to come closer with his King to e6 or e5 and attack the pawn on d6. And White's Bishop aims for b5/a4 squares from where it supports its own advanced passed pawn and block's black's a-passed pawn on the same diagonal.37...Ne7338.b525Kf63:0939.Rd41:06h655Black should go for a blockade on d6-square. This is usually effective when dealing with a light square bishop of the opponent. This concept is seen in various games, one that comes to my mind is from the Kramnik-Grischuk match in 2011.39...Nc8!40.Be2Nd6Here the Knight does a very effective job of blocking the d5-passed pawn, protecting the f5- weakness and attacking the b5-pawn as well!41.Rh4Kg742.Ra4Kf643.Ra6Ke5=40.Kf1!0 White forces Ding to come up with a decision on the 40th move. And it did work, Gukesh was putting pressure on Ding and he seemed to have cracked(almost!) here.Ke5?0Clearly DIng had considered40...Nc8!but decided against in trying to be more active intuitively.41.Rh4Kg7!Anish pointed out the importance of stepping back to g7 with the King.41...Kg642.Rc4Nd643.Rc6!This pin gains the requsite time to improve White's king.Kg744.Ke2Nxb545.Kd3 Once the King joins the party as an active fighting unit things get hard for Black.42.Rc4Nd643.Rc5Kf6and Black is faster with his king and holds the balance.41.Rh46:00Nxd55:3142.Rxh62:15Nc3743.Rc6!4:35Ne42:49
44.Ke1?!5:04 This move was criticised by Magnus, who even went on to call this move an eyesore! But Ding felt this was a very strong move on the part of Gukesh during the game. Just goes to show how complex this endgame is!The computer's reccomendation to win was44.h4+-when White clearly does not fear the exchange of knight for a bishop.Nd2+45.Ke2Nxf346.Kxf3And White is winning this rook endgame. Well once again one has to understand this during the game- not easy.A fascinating concept pointed out by Anish Giri was44.Bxe4!fxe4to go into a Rook endgame and even undoubling Black's pawns in the process. If the Queenside pawns are exchanged it would lead to a three versus two on the same flank and would most likely result in a draw. However, that is not so easy to achieve from Black's viewpoint since White's pawn on b5 is pretty advanced. Apart from that there are various scenarios in which White's active rook could cut off Black's king along the rank which might be a decisive factor in many a rook endgame. In the given case White carries on with a variation which shows lot of subtleties along the way.45.Rc5+!Ke646.g4!From the view of economy this is brilliant. White gets a square for his rook on f5, from where it will protect both the b5 pawn on f2 pawn. The other pawns are protected by pawns, which inturn are protected by the king! Rubinstein would have surely loved this! Apart from the above listed features, this move also disconnects Black's e and f pawns, but having said all this the variation goes on and wins in a very concrete fashion as showed by Anish.Rd347.Kg2Rd248.Rc6+Ke549.Rc7Kf649...Ke650.Rxa7e351.Ra3+-50.Kg3Rd3+51.Kf4Rxh352.Rxa7Rf3+53.Kxe4Rxf254.Kd5Rf455.b6Rxg456.b7Rb457.Kc5!+- with the idea Ra6+ and Rb6 .A typical idea known from similar Rook endgames.44...f6?!22:18 Objectively a mistake after twenty two minutes of thought in the final time control! But as Carlsen put it, its a fantastic practical move, Black moves the pawn away from the seventh rank where it can be a target for opponent's rook, and at the same time increases the attacking scope of his rook. Ding said he thought this position was lost and came up with this as a last resort fighting resource.44...Rd5±45.Be2Rd4is suggested by the computer and is apparently fine for Black!45.h4!?9:4545.Rc8to try and check from behind is a resource and trying to check opponent's activity by our own is a resource. But once again, not at all obvious!45...Rd3646.Bd1?1:20 This crucial moment was where White gave up the advantage and let black eventually draw the game.Anish pointed out a win for White here with46.Ke2!Rd546...Rb347.Bxe4fxe448.Rc5+Ke649.g4+-Once again the Rook on the fifth rank is super strong for White and this position is winning for White.This is probably due to too many weaknesses in Black's camp, namely a7 and e4 pawns and also White's h-passed pawn is a formidable force.47.b6axb648.Rxb6± Even though the pawns are on the same flank, White has an outside h-passer and Black's king is not well placed in the centre surprisingly! This means that White is perhaps winning this endgame.46...f4!3247.gxf4+!1:40Kxf49 Black fought ferociously keeping all his pieces including his king very active.48.Bc210:21The natural48.Rc4does not win because ofRh3!49.f3Rh1+50.Ke2Rh2+=is the line that Gukesh had missed earlier, as he pointed out.48...Rd59
49.Rc4!43f5350.Rb41450.Bd1Rxb5!?50...Rc551.Rxc5Nxc5is also not easy for White to make progress.51.f3Re552.Be2Kg353.fxe4Kxh454.exf5+Kg5=would lead to Rook Bishop versus Rook position!50...Kf3=1:5351.Bd1+35Kg231 In time trouble phase for both sides, Ding played very very actively. The king going deep into opponent's camp and maintaining coordination.Gukesh kept trying his every chance, but it did not prove enough to unsettle Ding.52.Rb333Re5!653.f438Re71:5554.Re336Rh73855.h531Nf62156.Re540Nxh51857.Rxf51Ng31058.Rf85258.Rd5Rf7Black is not winning the pawn immediately but eventually will do so thanks to his active pieces.58...Rb74859.Ba425Kf3=1:2360.f51Kf42861.f68Ne4462.Bc238Nd64063.Rd81:19Ke52164.Bb37Nf72565.Rd5+57Kxf6566.Kd22Rb61767.Bc414Rd6568.Kc35Rxd5369.Bxd53Nd6570.Kb43Nxb5471.Kxb53a6+372.Kxa63½–½