[Event "Riga ;MAINB"] [Site "Riga ;MAINB"] [Date "1957.??.??"] [White "Mikhail Tal"] [Black "Alexander Koblents"] [Result "1-0"] [Annotator "USER"] [Variant "Standard"] [ECO "B63"] [Opening "Sicilian Defense: Richter-Rauzer Variation, Classical Variation"] [Source "https://lichess.org/study/KJ5Vxrwu/qO55Rsrr"] [Orientation "white"] { Let's continue our journey through the magical world of Tal's games by examining his game against Latvian Master Alexander Koblents - who was incidentally Tal's coach since 1949 and very notably - his second in the 1960 and 1961 World Championship matches against Botvinnik. } 1. e4 { In this game, Tal opened with his favourite move. } 1... c5 { The Sicilian, leading to a sharp battle, was always a risky opening choice against a player as tactically gifted as Tal, but I guess Koblents had his reasons for choosing it. } 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 Nc6 { The Classical Sicilian has been overshadowed by Najdorf, Sveshnikov and some other Sicilians in the recent years, but it remains a very respectable opening to this day. Back in the day it was very popular - among other players, Mikhail Botvinnik used to play it on a regular basis at some point in his career. } 6. Bg5 { The Richter Rauzer attack is the critical way of handling the Classical Sicilian even today. } 6... e6 7. Qd2 Be7 { The old mainline, opting for quick kingside castling. } (7... a6 { Is the main move today, postponing kingside development in favour of rapid queenside development. }) 8. O-O-O O-O 9. Nb3 { This move avoids the typical exchange of the knights on d4. It is a very respectable and viable option, although less popular and arguably critical than the immedaite advance of the pawns on the kingside. } (9. f4 { Is what consitutes the mainline nowadays. }) (9. h4 { Also has the right to exist! }) 9... Qb6 { This is the most popular move in the position. The idea is to indirectly defend the d6 pawn and vacate the d8 square for the rook while avoiding any Nb5 ideas. Still, it is probably not to good, objectively speaking as the queen is put in front of the queenside pawns - and Black will very much want to attack on that side of the board. } (9... h6 { Is the immediate suggestion by the computer. The point is that Black can calmly give up the pawn on d6: } 10. Bxf6 Bxf6 11. Qxd6 Bxc3 12. bxc3 Qf6 { With more than ample compensation for the pawn. This is a relatively rare and unexplored continuation that deserves some more attention, for sure. }) 10. f3 { Removing the pawn from the attack and preparing the kingside play with g4-h4 and so on. Not the only way of handling the position, but definitely a very logical one. } { [%cal Gf3e4] } (10. Bxf6 Bxf6 11. Qxd6 Qxf2 { White can't really capture on d6 as it loses the f2 pawn and lands them in a lost position as the bishop pair and Black's dark-square control are too strong. }) 10... a6 { Black also prepares the queenside expansion slowly but surely. } (10... Rd8 { Defending the d6 pawn was also possible, but not necessary. }) 11. g4 { A very logical way of going for the attack, but maybe this is not the most precise way. } (11. Bxf6 Bxf6 12. Qxd6 { Now doesn't lose a pawn, but still provides Black with a ton of compensation after } 12... Rd8 (12... Bxc3 { is also quite reasonable. }) 13. Qg3 Qe3+ { As a matter of fact - Black regains the pawn at once and after } 14. Kb1 Bxc3 15. Rxd8+ Nxd8 16. bxc3 Qxc3 { is at least equal in this position. }) (11. h4 { Seems to be a better way - pushing the pawn to h6 is the priority here. Say, if Black continues with } 11... Qc7 { then the pawn on d6 is exposed: } (11... Rd8 { is achievement for White as after } 12. Be3 Qc7 13. Qf2 { Bb6 is threatend, the queen sidestepps trouble on the d-file and g4-g5 is coming. White has better chances here. }) 12. Bxf6 Bxf6 13. Qxd6 { And now the counterplay is missing and it is more difficult for Black to prove the compensation for the pawn. }) 11... Rd8 { This move is not necessary! } (11... Qc7 { Was possible. With the pawn on g4 instead of the pawn on h4 } 12. Bxf6 Bxf6 13. Qxd6 { Is not as good as after } 13... Qxd6 14. Rxd6 Be7 15. Rd1 { Black manages to play the shocking move } 15... g5 { Securing the dark-squares and the e5 square for the knight. This is, of course, pointed by the computer, but it is amazing to me how deeply the engines understand chess strategy these days - this line is perfectly understandable from a human viewpoint! } { [%csl Gg5][%cal Gg5f4,Gc6e5] }) 12. Be3 { Now the bishop has done its job on g5 and there is no point in keeping it there, so Tal retreats it with the gain of tempo and clears the way for the kingside expansion. } 12... Qc7 13. h4 { But this is a blunder that allows the immediate counterplay in the center. } (13. Qf2 { It was necessary to remove the queen from the d-file, threaten Bb6 and only after } 13... Nd7 { push the pawns with } 14. h4 { When White has better prospects in the upcoming fight with the opposite-side castling. }) 13... b5 { This is a mistake that allows White's attack to gain the momentum. } (13... d5 { Black could have executed the thematic Sicilian blow, exploiting the position of the rook on d8. The line could continue } 14. Qf2 dxe4 15. Rxd8+ Qxd8 16. g5 Nd5 { And Black has no problems, as the crisis in the center diverts White from executing the attack on the kingside. }) 14. g5 { Now the knight gets chased away and three is no d5 break any more. } 14... Nd7 15. g6 { A very interesting, although not necessary pawn sacrifice, typical of the Sicilian. The idea is to play h5 next and open the lines on the kingside. } (15. h5 { Continuing the attack with the push of the h-pawn, }) (15. Qf2 { Or moving the queen and calmly improving the pieces, were both possible, although Tal's choice is definitely more enterprising and thematic. }) 15... hxg6 { But this logical move is already a decisive mistake, as it opens up the h-file completely. } (15... fxg6 { For better or for worse, taking with the f-pawn should have been preferred. Usually, it is not positionally desirable to capture "Towards the center", but in this particular case, Black simply doesn't want to open up the h-file. Play might continue } 16. h5 gxh5 17. Rxh5 Nf6 18. Rh1 d5 { And here there is no direct attack along the h-file, so Black is much safer than in the game. }) 16. h5 gxh5 17. Rxh5 Nf6 18. Rh1 d5 { Now the same variation but with the pawn on f7 instead of h7 is losing for Black due to the open h-file. } 19. e5 { A good reply, closing the action along the d-file and preparing to reposition the bishop on f4 and the queen on h2. } (19. Bf4 { The immediate transfer of the bishop is pointed out by the engine as even stronger, but only due to some inhuman lines, such as } 19... Bd6 (19... e5 { also leads to a fascinating line. } 20. Qh2 Kf8 21. Qh8+ Ng8 22. Rg1 Bf6 23. Bh6 { And White manages to break through. Not easy sequence to see, even for the great Mikhail Tal. }) 20. Bxd6 Qxd6 21. f4 { When e5 is threatened and Black can't take on e4 due to the pin on the d-file. }) 19... Nxe5 20. Bf4? { But this is actually a significant mistake that allows Black to consolidate by moving the bishop from e7 to d6. } (20. Qh2 { Was necessary as now mate in 1 is threatened so Black doesn't have time to move the bishop from e7. Thus after } 20... Kf8 21. Qh8+ Ng8 { Black's pieces are not as coordinated and White can exploit it with the super precise } 22. Bd4 { Threatening f4. } 22... Bf6 23. Bc5+ Be7 24. Rh7 { When g7 will fall and the position is completely won after } 24... Bxc5 25. Qxg7+ Ke7 26. Nxc5 { As the e5 knight is exposed and the king is in the center of the board. Some chess still remains to be played, though, but this is definitely much better for White compared to the game. }) 20... Bd6 { Now the king gains the e7 square and Black's bishop is more active on d6. } 21. Qh2 { Now this infiltration comes a move too late. } 21... Kf8 (21... Ng6 { Was even stronger, when White doesn't have anything better than } 22. Bxd6 Qxd6 { When there is no attack and Black has two extra pawns. White does have some practical combination, but it should be objectively lost. }) 22. Qh8+ { Going from bad to worse - but this might be the best practical chance. } (22. Kb1 { Is the computer move, but you know you are sad if you have to suddenly switch to making such defensive moves in a position where you were attacking. }) 22... Ng8? { As usual in Tal games, it is very difficult for the opponent to withstand the pressure and Koblents goes very wrong here, allowing Tal to seize the initiative once again. } (22... Ke7 { Would have been the logical continuation now that the e7 square is available. Koblents was potentially worried about leaving the g7 hanging, but after } 23. Qxg7 Rg8 { The queen is simply trapped. Maybe Koblents didn't want to beat his student? A bit shocking, to be honest, that he didn't play this simple and obvious move. }) 23. Rh7 { Now White is fully back in the game and concrete threats start to appear. } { [%cal Rh7g7,Rh8g7] } 23... f5 { The only way to defend the h7 pawn and the best move in the position, although at first, I didn't fully understand why it is better to have the pawn on f5 as opposed to f6. However, it transpires that it is important to control the g4 square - precisely because of one specific line! } (23... f6 24. Bh6 Rd7 { It transpires that there is this strong move } 25. Nd4 { Threatening Nxe6. } 25... Rf7 26. Bxb5 gxh6 27. Rg1 { And in this position there is no Ng4 as the pawn is on f6 and not on f5, so Whiteis clearly better after some crazy line } 27... Ng4 28. Rxg4 Bf4+ 29. Kb1 Bg5 30. Rxh6) 24. Bh6 Rd7 25. Bxb5 { Adding some more fuel to the fire and creating a typical Tal chaos. The idea of this move is not so much to recapture on b5 as to clear the 1st rank for the d1 rook to join the attack. } { [%cal Gd1g1] } (25. Nd4 { Now doesn't work as after } 25... Rf7 26. Bxb5 gxh6 27. Rg1 Ng4 { wins for Black. This is why it was better to push the pawn to f5 instead of f6. }) 25... Rf7? { And immediately, Koblents commits a fatal mistake. } (25... axb5 { First of all, taking on b5 is not good as it - just like the text move - gives White time to play } 26. Rg1 { Increasing the pressure on the g7 pawn. } { [%cal Rg1g8,Rh6f8,Rh8g7] } 26... Ng4 { The only defence. } 27. Nxb5 { Getting ready to eliminate the bishop on d6. } { [%cal Rb5d6,Rb5c7] } 27... Bf4+ 28. Kb1 Qe5 29. Bxf4 Qxf4 30. fxg4 { And White retains material equality and has a raging attack. }) (25... Ng6 { Attacking the queen in the corner, was the only move. Play continues } { [%cal Rg6h8] } 26. Nd4 { Literally the only move to stay in the game, creating a counter threat of Nxe6. } 26... Re7 { Defending the e6 pawn. } (26... Nxh8 27. Nxe6+ Kf7 28. Nxc7 { Turns out to be better for White after some crazy complications } 28... Rxc7 29. Rxg7+ Kf8 30. Rxc7+ Nxh6 31. Bc6 { and when the dust has settled - White has slightly better chances in the incoming imbalanced endgame. Crazy and beautiful how even the most complicated lines can end up in some imbalanced, but technical endgames. }) 27. Bxg7+ { Absolutely the only move - but one has to see the following point. } 27... Rxg7 28. Bd7 { Marvelous interference! White creates the threats of Nxe6 and Qxg7 and it is very difficult to deal with both! } 28... Qxd7 { Absolutely the only move! } (28... Bxd7 29. Qxg7+ { Loses on the spot. }) (28... Nxh8 { also loses by force: } 29. Nxe6+ Ke7 30. Rxg7+ Nf7 31. Nxd5+ Kxd7 32. Nexc7) 29. Nxe6+ { Diverting the queen or the king from the defence of the rook on g7. } 29... Qxe6 30. Qxg7+ Ke8 31. Nxd5 { And the engine claims this position is equal, although it most certainly wont end in equality in a human game. }) 26. Rg1 { The attack on g7 reaches its peak and Black is helpless aginst the incoming onslaught. } { [%cal Rg1g8] } 26... Ra7 27. Nd4 { Bringing the knight closer to the action and increasing the pressure. Now the threat is Nxf5 and Nxd5, among others. } { [%cal Rd4f5,Rc3d5] } (27. Bxa6 { Is pointed out by the computer as strong. I guess the point is that after } 27... Bxa6 28. Nd4 { e6 is under threat and after } 28... Bc8 29. Ncb5 { White wins the exchange back but - more importantly - eliminates one of the defenders of the g7 point. }) 27... Ng4 { Trying to block the g-file desperately. } (27... axb5 28. Ncxb5 { Once again wins, as in the previous note. }) 28. fxg4 Be5? { The final mistake! } (28... Bf4+ { Would have offered some chances to resist after } 29. Kb1 Bxh6 30. g5 Bxg5 31. Rxg5 { When White has an extra piece, but Black has two pawns. This was definitely much better for Black than what happened in the game. }) 29. Nc6?! { Sufficient, but not the most precise. } { [%cal Rc6a7,Rc6e5] } (29. Nf3 { Was much stronger. The point is to have some crazy Nh4 idea in the case of } 29... axb5 30. Nh4 { and apparently this is winning for White. Crazy how much tactics the computers see in a position like this and how often paradoxical ideas are the best. } { [%cal Rh4g6] }) 29... Bxc3? { Capitulation! } (29... axb5 { Was the only way to stay in the game. White's problem is now that capturng on b5 is not possible and also that } 30. Nxa7 Qxa7 { Doesn't work as the bishop on e5 is defending the g7 pawn. } { [%cal Ge5g7] }) 30. Be3 { A deadly bishop retreat, clearing the h-file and creating a nasty threat of Rh1 and Qxg8. } { [%cal Re3a7,Rh8g8] } (30. bxc3 { was also sufficient as after } 30... axb5 31. gxf5 { The pressure on g7 is too much for Black to handle: } 31... Qxc6 32. Rgxg7) 30... d4 { Nothing else really works. This actually forces White to find the only winning way! } 31. Rgh1 { With the nasty threat of Qxg8 and Rh8 mate. } { [%cal Gh8g8] } 31... Rd7 { Clearing the f7 square for the king! } 32. Bg5? { Amazingly, this turns out to be a mistake that allows Black to escape! } (32. gxf5 { Seems to be the only winning move. The idea is to take on e6 next and prevent the king from escaping once again, while also trying to open the a2-g8 diagonal. } 32... dxe3 (32... exf5 { Also loses to } 33. Qxg8+ Kxg8 34. Bc4+ { [%cal Rc4g8] }) 33. fxe6 { Now the deadly threat of Qxg8 is still there and Black is just lost. }) 32... axb5 33. R1h6 { The culmination of this entire fascinating game! } 33... d3? { The final mistake, allowing a devastating attacking finish. } (33... Qxc6 { Black had two ways of staying in the game. The first was the "simple" recapture of the knight.White has nothing better than } 34. Rf6+ gxf6 35. Bh6+ Rg7 36. Bxg7+ Ke7 { When there is no immediate mate and White doesn't have more than a dynamic equality! }) (33... Bxb2+ { Was also possible, with very similar ideas after } 34. Kxb2 (34. Kb1 Qxc6 { is similar. }) 34... Qxc6 35. Rf6+ gxf6 36. Bh6+ Rg7 37. Bxg7+ Ke7 { when the engine gives the eternal "0. 00" evaluation. }) 34. bxc3 { I guess this recapture escaped Koblents' attention. White eliminates the defender of the g7 pawn and now intends to win with the Rf6 checking mechanism. } 34... d2+ 35. Kd1 Qxc6 36. Rf6+ Rf7 (36... gxf6 37. Bh6+ { Wins this time, as after } 37... Rg7 38. Bxg7+ Ke7 { there is } 39. Bxf6+ Kd6 40. Be5+ { This check wasn't possible with the bishop on c3, but now the rook on a7 will eventually drop and White will obtain a decisive material advantage. }) 37. Qxg7+ { And here Koblents decided to resign due to } 37... Ke8 38. Qxg8+ Kd7 39. Rhxf7+ { And Black is getting mated shortly. A fascinating, mind-boggling and typical Tal encounter, where he very successfully took his opponent into the "deep dark forest" and outwitted him in a sharp tactical melee. Games like these probably contributed to Botvinnik calling Tal's style "gambling", but even with the modern engines it is necessary to dive really deep to understand all the intricacies and nuances of different tactical lines. --- I can only imagine how difficult it was to face this style over the board - especially in the Botvinnik-Smyslov time where logic, harmony and positional player trumped this sacrificial, combinational style. } 1-0