Attacking with Tal – Part 1

April 27, 2018 Matthew Sadler 4 comments

Chess has taken a back seat for me after a recent move back to the UK from Holland. It’s been quite strange to fit back into British life after 18 years abroad, but 6 months on I can barely believe I ever lived anywhere else but here! As everything is settling down now, chess is starting to rear its head again and who else but Mikhail Tal to provide the ultimate impetus to start analysing again. I caught sight of this game in a book by Sakaev and Landa entitled “The Complete Manual of Positional Chess, Opening and Middlegame” and was astounded by Tal’s execution of the attack. As I tend to do, I tried to find more annotations of this game and found some in Tal’s own “The Life and Games of Mikhail Tal” and in Tibor Karolyi’s recent 3-volume biography of Tal. However, as always there were still plenty of gaps to fill in which meant I spent a pleasant number of hours analysing the game. I hope you enjoy it as much as I do! As always, a playable version of this game is available at

I have also created a video blog on this game which is viewable on my YouTube Channel at


Tal,Mihail – Uhlmann,Wolfgang

Alekhine Memorial 1971

Tal recounts that he very much wanted to surprise Uhlmann in the opening, but that he wasn’t sure how. In the end his choice fell on the the lesser-known 5.Bb5 in the French Tarrasch.


1.e4 e6 2.d4 d5 3.Nd2 c5 4.Ngf3 Nc6 5.Bb5



As Karolyi points out, Uhlmann – a leading expert on the French – had faced this line the previous year against Parma and chosen 5…Bd6. Tal notes that “Uhlmann is one of those chess players who does not attempt to be too sophisticated, and as a rule plays the opening quickly. The fact that he spent more than 20 minutes over his fifth move showed that the psychological “mine” (5.Bb5) had worked.” On the other hand, the fact that Uhlmann chose such a risky, concrete continuation (opening the centre against Tal!) and then repeated it 9 times subsequently (including a follow-up game against Tal in Tallinn 1977) indicates to me that Uhlmann had probably looked at this line at home already. From a general point of view, it feels a little odd to repair the only defect in White’s position – the bishop on c1 blocked in by the knight on d2 – at such an early point in the game. However, Uhlmann’s play is based on the tactical exploitation of White’s bishop on b5. That’s the thing about advanced attacking pieces: they exert pressure on the opponent’s position but they also provide nearby targets for the opponent to exploit!


5…dxe4 6.Nxe4 Bd7


Uhlmann’s idea, and a very clever one. Black now threatens …Nxd4 winning a pawn due to the discovered attack on the bishop on b5. Tal continues with 1–move development, ensuring Black has no respite by attacking the Black queen.






was Tal’s choice 6 years later.




(7…cxd4 8.Bg5 Nf6 9.Nxd4 Be7 was Uhlmann’s choice in 1992, although this looks worse to me than 7…Nxd4. 10.Bxf6 gxf6 11.Nxc6 bxc6 12.Bc4 Qc7 13.Qh5 Qe5 14.Qxe5 fxe5 15.Rad1 0–0–0 16.f4 (The obvious 16.Nd6+ Bxd6 17.Rxd6 Kc7 18.Rd3 looks like a little edge for White at least.) 16…exf4 17.Rxf4 f5 18.Nd6+ Bxd6 19.Rxd6 Kc7 20.Rd3 Bc8 21.Ra3 Rd1+ 22.Rf1 ½–½ Fritsche,L -Uhlmann,W Germany 1991)




(8.Bxd7+ Qxd7 9.c3 is Sakaev’s and Landa’s suggestion with compensation with the pawn which hasn’t been tried in practice yet.)


8…f6 9.Nxd4 cxd4 10.Bh4


Typical Tal, preventing Black’s development tactically.





Preparing development with …Nh6–f7


(10…Nh6 11.Bxd7+ Qxd7 12.Bxf6 Qc6 13.Qh5+ Kd7 14.Rad1 gxf6 15.Rxd4+ Ke7 16.Rfd1 Nf5 17.R4d3 is enormously strong. g4, Rc3 and Rd7+ are all in the air.)




(11.Qh5+ g6 12.Qe2 is also worth a try, again preventing Black’s development with tactics: 12…Nh6 13.Nxf6+ Bxf6 14.Qxe6+ Be7 (14…Kf8 15.Qxf6+ Qxf6 16.Bxf6 wins a pawn) 15.Bxe7 Qxe7 16.Bxd7+)


11…Bxb5 12.Qh5+ Kf8 13.Qxb5 Qd5 14.Qd3 dxc3 15.Qxd5 exd5 16.Nxc3 Rd8 17.Rfd1 d4 18.Nb5

and Tal somehow managed to squeeze out a win from this! Black has equalised however. Tal-Uhlmann,W Tallinn 1977




Again demonstrating the drawback to advanced attacking pieces. This disruptive check costs White time, forcing the knight to retreat to c3 to protect the bishop on b5.





A crucial moment in the game. This position is a nice conflict between easy, 1–move development (Tal’s development) and development based around specific tactical factors in the position (Uhlmann’s). In general, we can see nothing wrong with White’s position: the knights are out, the bishops are out, his king is ready to castle, the centre is reasonably open and White has the d4–d5 break at his disposal if Black doesn’t react quickly. Black on the other hand has moved his queen early in the game, only developed his queenside pieces and his king is stuck in the centre. Listening to this litany of woes, you may feel that it’s time to bring out the wagging finger, shake your head sadly and say “Grandmaster how could you?”

There is another side to the story however. Black’s queen is a very powerful piece: it pins the knight to c3, attacks the bishop on b5 and don’t forget that bishop on g5 on the same rank (the queen on a5 won’t!) That bishop on g5 is protected by the knight on f3, but that knight can easily be drawn away due to Black’s pressure on d4. Add the opposition of the bishop on d7 and the bishop on b5 into the mix and you start to see some areas that might be of immediate concern to White. In fact, these tactical factors give Black a number of reasonable possibilities. One of the keys to this is that although Black’s kingside is not yet developed, there are no obstacles to the easy development of these pieces: both the bishop on f8 and knight on g8 have multiple good squares available. If White doesn’t find a way to get at Black quickly, it won’t be long before Black has caught up in development.




Risky, especially against Tal, but not a bad move. Karolyi suggests 8…h6 and I was particularly keen on 8…a6. Sakaev and Landa give an interesting variation which was the starting point of my investigations (as I didn’t quite understand it!)


8…a6 9.Bxc6 Bxc6 10.d5



(10.0–0 is really bad! 10…cxd4 11.Qxd4 Bxf3 wins a piece! The triumph of specific piece placement over general development. That useful-looking bishop on g5 turns out to be loose! 12.Rad1 Bxd1 13.Rxd1 Qc7;

10.Ne5 Bxg2 11.Rg1 Bd5 is my engine’s line with advantage to Black. It’s worth a try in my opinion, but I think that Black should cope.)


10…exd5 is the only line given by Sakaev, which seems illogical to me. 10…Bxd5 feels like a much safer option and was indeed one of Uhlmann’s choices.


(10…Bxd5 11.0–0 Bc6 Feels like the best idea to me: 11…Bxf3 gives White free and easy development, and at the very least, it doesn’t challenge White to find any difficult moves.



(11…Bxf3 12.Qxf3 Qc7 13.Rad1 Be7 14.Bf4 Qc6 15.Ne4 (15.Qxc6+ bxc6 16.Na4 Nf6 17.c4 is another idea) 15…Nf6 16.Nd6+ Bxd6 17.Qxc6+ bxc6 18.Bxd6 (18.Rxd6 Ke7 I didn’t appreciate this idea 19.Rxc6 Nd5 20.Bg3 Rhc8 looks pretty equal) 18…Nd7 19.Rd3 0–0–0 20.Ra3 looks pleasant for White)


12.Ne5 Qc7 The best idea for a number of reasons, one of which is simply that Black maintains real winning chances only as long as his c-pawns are not doubled. 13.Re1 Nf6


(13…Be7 14.Bf4 followed by Qg4 / Qh5 is surprisingly unpleasant)


14.Nxc6 Qxc6 15.Bxf6 gxf6 16.Qh5 and now 16…Rg8


(16…Be7 17.Re3 h6 18.Rae1 Kf8 19.Nd5 Rd8 20.Nf4 e5 21.Ne2 Rd2 22.Ng3 turned out a little fraught for Black, but Uhlmann won through in the end 22…Rxc2 23.Nf5 Qd7 24.Rg3 Qd2 25.Rf1 Qe2 26.Qxe2 Rxe2 27.Rd3 Ke8 28.Rfd1 Re4 29.f3 Rd4 30.Nxd4 cxd4 31.a4 a5 32.g4 Kd7 33.Rc1 b6 34.Rc2 Bc5 35.Kg2 Ke6 36.Rd1 f5 37.h3 h5 38.Re2 Rg8 39.Kf1 hxg4 40.hxg4 fxg4 41.f4 f6 42.fxe5 fxe5 43.Rde1 Rg5 44.Kg2 d3 0–1 (44) Nicevski,R (2395)-Uhlmann,W (2555) Skopje 1976)


17.Ne4 Rg6 was my analysis before I realised that Uhlmann had played this already! This also looks good for Black)


Back to 10…exd5


11.0–0 f6 Sakaev’s line




(11…d4 was another alternative I couldn’t refute easily.




(12.Ne5 Qc7 (12…dxc3 13.Re1 Be7 (13…Ne7 14.Nc4) 14.Nxc6 bxc6 15.Qd6 Qd8 16.Bxe7 Nxe7 17.Rxe7+ Qxe7 18.Qxc6+ wins) 13.Qh5 was the idea, to attack f7 before Black can play …Be7 and …0–0–0 (13.Re1 –> 12.Re1) 13…g6 (13…Bd6) 14.Qh3 dxc3 15.Rae1 was the idea, but it doesn’t look too terminal for Black 15…Be7 16.Qxc3 f6 Couldn’t find a way through, and nor can my engine.)


12…Be7 13.Ne4




(13.Ne5 Qc7 Looked fine for Black to me (surprisingly) 14.Qg4 (14.Qh5 g6; 14.Ne4 Qxe5 15.Nf6+ gxf6 16.Rxe5 fxe5) 14…g6 is not clear, but I don’t think Black is doing too badly. 13.Ne4 looks stronger.)




The move Black wants to play if he can.




(14.Bxe7 Nxe7 15.Neg5 Very strong I thought but my engine finds extra resources 15…Ng6 16.Nxf7 Rhf8 17.Nxd8 (17.N3e5 Nxe5 18.Nxe5 may be an edge for White) 17…Bxf3 18.gxf3 Qxd8 gives Black some compensation for the exchange)


14…Bxe4 15.Bxe7 Nxe7 16.Nxf7 Bd5 17.Rxe7 Bxf7 18.Rxf7


is definitely an edge for White)


12.Nxd5 fxg5 13.Re1+ Ne7 Sakaev’s line



(13…Be7 was not obvious to me either. I came quite a way, but I needed my engine to help me through the final defences!


14.Nxe7 Necessary to stop Black from castling his king away to safety


a) 14.Ne5 Qxe1+ 15.Qxe1 Bxd5 16.Rd1 Nf6 looks fine for Black;


b) 14.Nxg5 0–0–0 15.Qg4+ (15.Rxe7 Bxd5) 15…Kb8 (15…Bd7 16.Qf4 (16.Ne6 Bf6 I couldn’t find anything for White here. Nor could my engine 17.b4 cxb4 18.Qc4+ Bc6) 16…Bxg5 17.Qxg5 Nf6 18.b4 Qa3 19.Nb6+ Kb8 20.Qxc5 Bc6 21.Qe5+ Ka7 22.Qc5 Kb8 is my engine’s draw by repetition.) 16.Nxe7 Nxe7 17.Rxe7 Rhe8 My engine thinks Black will have sufficient play which seems fair enough;


14…Nxe7 15.Nxg5



(15.Qd6 Qd8 (15…0–0 16.Rxe7 looked like a slight edge to me) 16.Qxc5 Bxf3 17.gxf3 still felt dangerous for Black, but my engine is confident!)




Abandoning queenside castling, but intending to bring the Black queen into the game, possibly via d2. The rook is also available to defend the knight on e7 from d7. (15…Qd8 16.Qg4 Qd6 17.Rad1 Qf6 18.Re6 is very strong)




a) 16.Qe2 0–0 17.Qxe7 (17.Qe6+ Kh8 18.Nf7+ Rxf7 19.Qxf7 Ng6 is good for Black) 17…Rde8 wins. A nasty trap!;


b) 16.Qg4 with the threat of Rxe7+ and Qe6–f7+ mate 16…Rd7 is not easy to break. I tried (16…Bd5 17.c4 forces Black to put the bishop to g8 before the rook on h8 has been activated. 17…Bg8 18.Ne4; 16…Rd4 17.Qc8+ Qd8 18.Rxe7+ Kxe7 19.Qe6+ is a typical tactic; 16…Rd6 17.Qf4 is the engine solution 17…Qc7 18.Rad1 Rxd1 19.Qf7+ Kd8 20.Rxd1+ Kc8 21.Ne6) 17.Re3 (17.Re2 0–0) 


b1) 17…0–0 18.Qe6+ Kh8 19.Rh3 The reason for Re3 instead of Rh3.;


b2) 17…Rf8 18.Rae1 Bd5 I found this tough to break. In the end my line is my engine’s top line. 19.Nxh7 Rf7 20.c4 Qd2 21.f3 (21.R3e2 Qd3 22.cxd5 Qxh7 23.d6 Rxd6 24.Qc8+ Rd8 25.Qxc5 g6 looked fine for Black which my engine confirms.) 21…Qd4 22.cxd5 Qxg4 23.fxg4 is much better for White.;


b3) 17…Qd2 18.Rae1 0–0 19.R3e2 Qd6 20.Re6 Qd2 is my engine’s draw by repetition.;


b4) 17…h5


My engine’s top choice again, given as advantage for Black


18.Qe2 c4 19.Re5 Qb6 20.Re1 0–0 My engines felt Black would escape. This doesn’t matter though as White has a subtle and strong idea:;


16…g6 17.Qe2 This takes away the option of …0–0 due to the mate on h7.



(17.Qh6 Qd2 18.Qg7 Qxg5 19.Qxh8+ Kd7 20.Rad1+ Bd5 wins for Black; 17.Qh4 Kd7)




(17…0–0 18.Qxe7 Rde8 19.Qxh7# is the difference)


18.Rad1 Qd8 (18…Qc7 19.Ne6) 19.Qe5 Rg8 20.Rxd7 Kxd7 (20…Qxd7 21.Nxh7) 21.Ne6



is extremely painful for Black:


21…Qa5 22.b4


(22.Nxc5+ Kc8 23.Qe6+ Kb8 24.Qxe7 is the easy human way to advantage) 22…Qb6 23.Nf8+ Kc8 24.Qe6+ Kb8 25.Nd7+ is the fantastic engine solution.)


Back to 13…Ne7



14.Nxe7 Bxe7 15.Nxg5 c4


is the line Sakaev gives, and indeed the engines are all in agreement: Black holds!




(16.Qg4 Mentioned by Sakaev as worthy of investigation.)


16…g6 17.Qg4 Qf5 18.Qd4 Qxg5 19.Qxh8+ Kf7 20.Qxh7+ Kf8 21.Qh8+ with a draw by perpetual]


In the game, Uhlmann played 8…cxd4


9.Nxd4 Bb4


Uhlmann isn’t holding back in this game, consistently choosing the riskiest lines.


9…Be7 Tal’s thought during the game, was tried twice by Uhlmann in subsequent years and looks very safe indeed. It’s nice that Black can develop by gaining a tempo against a developed White piece! 10.Qd2 (10.Bxe7 Ngxe7 11.0–0 0–0 12.Nb3 Qc7 13.Re1 Rfd8 14.Qh5 Nb4 15.Rac1 Bxb5 16.Qxb5 Nbd5 was even in: Rooze -Uhlmann,W Bad Zwischenahn 2008) 10…Nf6 11.0–0–0 “with at any rate a sharp struggle” – Tal 11…Rd8 12.Nb3 Qc7 13.Bxf6 Bxf6 14.Ne4 Be7 15.Nd6+ Bxd6 16.Qxd6 Qxd6 17.Rxd6 Ke7 was dead equal in: Horvath,T -Uhlmann,W Austria 2005


10.0–0 Bxc3 11.bxc3 Qxc3



This is the decisive mistake, which loses to an astonishing tactical resource. I spent some time looking at


11…h6 12.Qh5 interested me a lot


(12.Bh4 Qxc3 must be an improved version of the game. You’ll see why it’s important to knock the bishop off the c1–h6 diagonal! 13.Nf5 exf5 14.Re1+ Be6 15.Rxe6+ could be the only negative point for Black in the inclusion of …h6 and Bh4 but… 15…fxe6 16.Qh5+ Kf8 wins)


12…a6 Again harrying White’s advanced attacking pieces and trying to make use of the queen’s attack on the g5 bishop along the 5th rank.


a) 12…Nxd4 13.Bxd7+ Kxd7 (13…Kf8 14.Be7+) 14.Qxf7+ The key tactical point of 12.Qh5. 14…Kc6 15.cxd4 Qxg5 (15…hxg5 16.Qxe6+ Kc7 17.Qf7+ Kc6 18.Rfe1 Rh6 which I assessed as winning for White. My engine gives: 19.Re5 which certainly does the trick.) 16.Rae1 My line with which my engine agrees;


b) 12…Qxc3 looks like the other crucial try but 12…a6 looks safer!.;




(13.Nf5 exf5 14.Rfe1+ Kf8 unfortunately looks completely safe for Black: too many bishops are hanging.)




(13…Bxc6 14.Nxe6 looked like way too much! 14…g6 15.Ng7+ Kf8 16.Be7+)


14.f4 Is murky and worth a punt! However, my feeling is that Black should be fine. 14…Nf6


(14…Ne7 15.f5 Nxf5 16.Rae1 0–0 17.Bxh6 gxh6 18.Nxf5 exf5 19.Qxh6 Qc5+ 20.Kh1 Rae8 21.Qg5+ was my engine’s draw by repetition)


15.Qh4 Ne4


(15…c5 16.Bxf6 gxf6 17.Qxf6 Rg8 18.Nf5 exf5 19.Rae1+ Kf8 20.Qxh6+ Rg7 21.Rf3)


16.Be7 Qxc3


(16…g5 17.fxg5)


17.Rad1 was the line I was looking at. It looks pretty fraught: I’m not at all sure what’s going on here!;



And finally, 11…a6 was Tal’s suggestion, but seems pretty good for White. 12.Bxc6 Bxc6 13.Nxc6 (13.Qg4 also looks awkward) 13…Qxg5 14.Qd6 Ne7 15.Rfd1 Tal 15…Nxc6 16.Qd7+ Kf8 17.Qxb7 was Tal’s line]





Back at the board, Tal has just uncorked this!


12…exf5 13.Re1+ Be6 14.Qd6 a6



You might think that Black could survive this as there is no easy way to break through Black’s position and all of Black’s weak dark squares seem to be covered by undeveloped Black pieces! The knight on g8 covers e7 and the rook on a8 covers d8. However, there is one undeveloped piece that is contributing nothing, and Tal exploits precisely that one!





15…Qxc2 16.Bb4



Wow!! What a switchback! f8 is the one dark square close to the Black king that cannot be defended and Tal finds his way there!




16…Nf6 17.Qe7#


17.Qf8+ Kd7 18.Red1+ Kc7 19.Qxa8


And Uhlmann resigned. He cannot develop his kingside and 19…Nxb4 20.Qd8+ Kc6 21.Qd6# is mate




4 Comments on “Attacking with Tal – Part 1

  1. Great analysis, Matthew! Thanks!
    Incidently, it casts doubts on the quality of TCMOPC of Sakaev & Landa. As a matter of fact I just bought those 2 books, but I’m starting to have mixed feelings about the quality. Making so many mistakes in the very first game analysed in their Volume I, is not very convincing.

    1. Hi Job, thanks for your mail! Yes, I have mixed feelings too about these books. I like the format idea of small, separate lessons very much as you can play through each lesson reasonably quickly and feel you’re making progress. However, the material is a bit uneven. I liked the second book more than the first book, so you may want to start with that one! Best Wishes, Matthew.

  2. Hi Matthew, Thanks a lot for your reply! I’ll follow your advice and start with Vol. 2! Moreover, I plan to go over each lesson reasonably quickly and try to just grasp the main ideas instead of going into too much detail.
    Let me take this opportunity to thank you for your many “Sadler on Books” columns in the wonderful magazine of New in Chess! I’ve read everyone of them and found them very instructive and useful. They were surely helpful in choosing what chess books to buy, over the years. After reading your latest “Sadler on books” column, I have put Alekhine’s Odessa Secrets on my wishlist. And also the Rozentalis book from Thinkers Press, a relatively young Chess Publishing Company that deserves our sympathy, to my opinion. Keep up the good work! Best regards, Job de lange, the Netherlands

    1. Hi Job, thanks for your comments! I really enjoy my job of reading chess books, so glad the reviews have been helpful! “Alekhine’s Odessa Secrets” is a great read, and indeed I like the Thinkers Press books a lot too. That’s the advantage of having strong players in the editorial staff I think: they commission books that they would like to read themselves! Best Wishes, Matthew

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.