While solving puzzles on ChessTempo, I came across this puzzle:

[FEN "8/8/1p6/p1p3p1/P1P1B1K1/1P2kP2/5n2/8 w - - 9 1"]

1. Kf5

I played the intended correct answer,

[FEN "8/8/1p6/p1p3p1/P1P1B1K1/1P2kP2/5n2/8 w - - 9 1"]

1. Kf5 g4 2. fxg4 Nxe4

My question now, is this final position drawn, thus invalidating the puzzle?

I tried some lines but they all seemed to draw. Example:

[FEN "8/8/1p6/p1p3p1/P1P1B1K1/1P2kP2/5n2/8 w - - 9 1"]

1. Kf5 g4 2. fxg4 Nxe4 3. Ke5 Kd3 4. Kd5 Nf6+
5. Kc6 Kc2 6. Kxb6 Kxb3 7. g5 Ne4 8. Kxa5 Kxc4 {Tablebase draw} 1/2-1/2
  • @user929304 Was inactive for a period of time
    – ericw31415
    Mar 1 '18 at 12:00
  • ah okay :(, hope all is well.
    – user929304
    Mar 1 '18 at 12:29

Yes this seems to be a drawn endgame, most variations actually end up in a queen vs queen+knight situation which is generally a drawn endgame (one example below). But that doesn't mean the puzzle is flawed because as far as I know chesstempo puzzles are just about finding the best moves a priori, thus not always about a winning line. Good tactics trainers (such as chesstempo, lichess...) usually only say which side is to move, and the objective of the puzzle is actually part of the puzzle itself (that is, whether one is to play to mate, play to win material, or play to draw etc.)

In this particular position black has to come up with the best moves in order to hold the game it seems.

 [title "One possible continuation, 1/2-1/2"]
 [fen "8/8/1p6/p1p3p1/P1P1B1K1/1P2kP2/5n2/8 w - - 0 1"]

 1. Kf5 g4 2. fxg4 Nxe4 3. Ke5 Kd3 4. Kd5 Ng5 5. Kc6 Kc3 6. Kxb6 Kb4 7. Kc6 Nf7 8. Kb6 Ng5 9. Kc6 Kxb3 10. Kxc5 Kxa4 11. Kd6 Kb4 12. c5 Ne4+ 13. Ke5 Nxc5 14. g5 a4 15. g6 a3 16. g7 a2 17. g8=Q a1=Q+ 18. Kf5 Qf1+ 19. Kg6 Qg2+ 20. Kh7
  • The puzzle is flawed then, because any other move would also create a draw.
    – ericw31415
    Feb 10 '18 at 2:36
  • 1. Kf5 Nxe4 2. fxe4 g4 3. e5 g3 4. e6 g2 5. e7 g1=Q 6. e8=Q+ 1/2-1/2
    – ericw31415
    Feb 10 '18 at 3:16
  • 1
    @ericw31415 just because there's another drawing line it doesn't mean it's a flawed puzzle. It's always about the most accurate (optimal) variation. I'm sure you've seen lots of checkmate puzzles where for example there are a ton of other checkmates but that are longer (or any other winning variation) than the accepted/best variation, and they don't mean the puzzle is flawed.
    – Ellie
    Feb 10 '18 at 19:04
  • But say you put a drawn position into a tablebase. It'll just give you a random drawing move. You can't "draw faster".
    – ericw31415
    Feb 10 '18 at 20:00
  • There are no 10-piece endgame tablebases to my knowledge, so I don't what you've looked up. The starting position of the puzzle is what matters.
    – Ellie
    Feb 10 '18 at 21:09

I agree that this game seems to be a draw. And this fact DOES make the puzzle flawed, according to the rules of ChessTempo.

The rules of problems depend on the website. On some websites, the problem simply requires to find the best move. In those cases, the best move can be a winning move, or it can be a move that draws or equalizes an apparently lost position.

But on ChessTempo, the rules are different, the player is required to find a move that wins, and this is defined as giving a won endgame, or an advantage of at least two pawns.

So, if this position is really drawn (as it seems), then this puzzle is wrong, and should be reported to the administrators of the site.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.