How to Avoid Blunders, Part 2

  • GM DanielNaroditsky
  • | Apr 28, 2014

Last week, I discussed a three-step method which you can (and should) apply to every move you are planning to make in a complex position to diminish the risk of blundering. As promised, today, we will actually test this method out and determine if, at least to an extent, it is possible to override innate human imperfection.

I will begin by sharing one of my own painful experiences. In December 2011, I played in the strong Groningen Open in Holland (the Netherlands). By round five, I had three points and was well on my way to a GM norm. I vividly remember spending more than three hours preparing for my next opponent, Dutch GM Sipke Ernst. When my opponent tentatively deviated from my preparation, I was convinced that he had overlooked a potentially decisive sequence. 

The position is head-explodingly complex. White is a pawn down, but his pieces are tremendously active and he potentially threatens to root out the g7 bishop with Nf6+. Furthermore, the c7 pawn is currently attacked, and I was not particularly thrilled with the position after 20...Nce6 21.Nxe6 Nxe6 22.Nf6+ Bxf6 23.Rxf6. However, I quickly found a way to highlight the only flaw in White's position: the weak e3 square. After the forced sequence 20...c6 21.Nf6+ Bxf6 22.Rxf6 Qe3+ (diagram), White's position seems to collapse. 

White's main problem is that 23.Kh2 loses a rook to ...Qe5+, while 23.Kh1 allows 23...Nxh3 and now 24.Rf3 falls prey to 24...Qg1 mate! But let us conduct a quick blunder check in this position. Both 23.Kh2 and 23.Kh1 are clearly losing, but White has a third move - 23.Bf2 - after which things might not be so obvious.

Sipke Ernst | Image ©

After the forced 23...Nxh3+ 24.Bxh3 Qxh3, White's pieces are still very active and Black must ensure that there is no coincidental tactic. Of course, there is no checkmate or double attack in sight, so all that remains is to check for pins. With an exposed king on g8, the chances of a pin are quite high, and with a queen so close to the g-file, they are even greater. After the simple 25.Rf3!, Black has no choice but to take on g4 (25...Qxg4+), and 26.Rg3 pins the queen. 

I did conduct this check - but it was only after I played ...Qe3+! Of course, with 23...Nxh3+ impossible, Black must retreat and is completely lost. My opponent's technique was not impeccable, but it did the trick:

Notice the two factors that should have set off tactical alarms: an unsupported queen deep in the heart of my opponent's camp, and a severely weakened king. In conclusion, you should keep in mind that to conduct a successful DAP check, you must frequently involve your tactical intuition. Always be on the lookout for pieces that might end up in pinnable locations.

Now, try your luck at the following exercise. There are several equally good moves that White can make; I can only set one of them as the "correct" solution, so you should focus on avoiding Janssen's blunder rather than solving the problem on your first try! 

Black's tactic is difficult to see, not so much because it is somehow unique or tough to calculate, but because the position appeared so quiet - so devoid of tactical possibilities - that White did not think twice about his choices.

Now, let us examine a more complex scenario in which we will have to go through the three-step method in its entirety in order to preclude a disastrous oversight.

The three-step method might seem cumbersome, and therefore it is extremely tempting to trust your gut when calculating a potentially decisive move. Of course, you must draw the line between caution and paranoia, but by forcing yourself to systematically check your calculations, you will approach the position with a fresh pair of eyes and (hopefully) root out miscalculations. 

When it comes to tactics, there is no alternative to practice. Therefore, I would encourage you to tackle the following exercise. You will only have to make one correct move, but the challenge will consist in applying the three-step method and discovering the devilish pitfall. Good luck!

In general, I am not a fan of catchy titles that promise a lot and deliver very little, but I hope that this article does not fall into that category! As I mentioned at the outset, there is no way to eliminate blunders altogether, but if you remain vigilant throughout the game and keep your self-confidence in check, you might very well find that winning a won position is not so difficult after all! 



  • 21 months ago


    @unusualkid: Well, there is no way to stop mate. If 49. is anything then 49...g5#. The only thing to delay it (and it's not a solution) is to play 49.g4 to which black answers with 49...Qxf4 when 50...g5# is unstoppable mate.

    Hope this helps :)

  • 22 months ago


    how does kh6 make gulko feel sick? I don't get it

  • 2 years ago


    • In your game against Sipke Ernst, the "blunder" is actually a six-move-long difficult-to-calculate sequence in an extremely complex position, leading to material equality and a slightly worse endgame for black.  I'm not sure that qualifies as a "blunder" to anyone who isn't a GM.  =P

    • Haha :)
  • 2 years ago


    in the last diagram(Van Wely game) instead of 32...Rf7,32...Qe2+ seems to be a lot easier.

    32...Qe2+ 33.Rf2 Qxf2+! 34.Kxf2 Rxf6+ 35.Qxf6 Bxf6 or

    32...Qe2+ 33.Kg1 Qxf1+! 34.Kxf1 Rxf6+ and in either case black will emerge a piece up.anyway i wanna thank Mr.Naroditsky for such helpful and informative articles.

  • 2 years ago


    What about Rg4 in the last puzzle?

  • 2 years ago


    Awesome article indeed!! 

    Thank you very much Daniel. To be honest I have liked all of your articles, they not only gets you in the topic but it motivates you, not to mention they are always very interesting!! 

  • 2 years ago



    Indeed. Also, I think it take a great amount of courage to post one's lost games.

  • 2 years ago


     Why not play, in the first game, after 23. Bf2, Nh3; 24. Bxh3, Qxh3; 25. Rf3, Bxg4!  26. Rg3? or maybe another move, but I don't see a move for white without losing material and the attack on the white king goes on.

  • 2 years ago


    Gulko must have felt sick after Kh6. Great article! Thanks

  • 2 years ago


    I agree with the comments below - this is very enyojable and accessable to read, indeed a great article!
    @Chuckieman: One can write columns in advance ;-) 

  • 2 years ago


    Daniel, the question I have for you is...

    How are you able to find time to write your Friday column when you're busy playing at the US Championships?!?

    Shouldn't you be spending your time preparing for Kamsky and Lenderman instead of writing this article for

  • 2 years ago


    I think that GM Naroditsky deserves a lot of praise for how his column has evolved. While the pure chess content has always been high, their were a number of complaints about the writing style, some justified, others not so much. It is clear, in reading the last two articles that the author has responded to these issues. The last two or three articles have had much simpler sentences and noticably fewer "ten cent" words.

    Regardless of your stance on GM Naroditsky's initial style, I think that it was a very humble and nice gesture on his part to have modified his style to meet the needs/wishes of the readership.

  • 2 years ago


    These two articles were great, and I enjoyed reading them.

    I want to thank the author for using simple English, which makes it very easy to read for international readers. :-)

  • 2 years ago


    Great article Daniel. It takes real modesty to give examples of blunders that you made from your own games.

  • 2 years ago


    thanks i always blunder

  • 2 years ago


    I mean Qf8.

  • 2 years ago


    'In Gulko's game black could take e5 pawn : 48.Nd7 Bxe5 49.Nxe5 Qe8#, and if 49.Qc8  then 49...Bf6+. How he missed this one??'

    I think the point is that white gets to mate 1 move sooner by Qc8, then Qh8.

  • 2 years ago


    In Gulko's game black could take e5 pawn : 48.Nd7 Bxe5 49.Nxe5 Qe8#, and if 49.Qc8  then 49...Bf6+. How he missed this one??

  • 2 years ago


    In your game against Sipke Ernst, the "blunder" is actually a six-move-long difficult-to-calculate sequence in an extremely complex position, leading to material equality and a slightly worse endgame for black.  I'm not sure that qualifies as a "blunder" to anyone who isn't a GM.  =P

  • 2 years ago


    In the first game, 23- Nxh3 24- Bxh3 Qxh3 and the white pawn on g4 is hanging... Is there anything I am missing?

Back to Top

Post your reply: