We are better equipped to “play bad chess well” than to “play good chess”. I offer you an experiment to demonstrate (one that will improve your chess, so it’s a win-win).
The best method for practicing chess is not playing chess. Playing chess is kiddie playground time, but to practice with a view toward improvement, play through a master game (ideally with a master’s explanatory annotation) while covering the winner’s moves.
When it’s “your” move, make the move you’d play in a serious game, then uncover the master’s actual move. If you guess correctly, pat yourself on the back, make the opponent’s move, guess again. When you guess incorrectly, ask yourself why the master’s move differs from your guess. With practice and patience, it truly helps; guessing Capablanca games is how I learned to play passably well.
Perform this guessing exercise with two games: Capablanca-Masyutin, Kiev 1914; and Capablanca-Tartakover, New York 1924.
These games were played under very different conditions. Black played the Dutch Defense in both, but during a simultaenous exhibition, Capablanca had fun with the “unsound” and “incorrect” 2. e4. Against Tartakover at a top-flight tournament, Capablanca played “good chess”.
For the Masyutin game, start guessing Capablanca’s moves with move 4. Work through that miniature, then guess Capablanca-Tartakover beginning at move 9. I predict you’ll make a higher percentage of correct guesses on the Masyutin game and its straightforward attack than on the Tartakover game with its delicate middle- and endgame.
An aspect of “playing bad chess well” is mindreading. If you sense your opponent is driving toward a cliff, don’t stop him.
How did you do at the guessing? My teacher sets a target of 80% correct guesses, but that is really, really hard (I’ve managed that just a few times ever), and there’s a very narrow window for improvement. Say you worked on a 32-move game, and you skipped the first 12 moves of the opening, to make 20 guesses. You can get 50% right by making recaptures, getting out of check, and moving through forced sequences; leaving about 10 moves to determine whether you hit his target or not. To reach 80%, you’d have to guess 6 out of those 10.
For the Masyutin game, my teacher and I ask students to guess 16 moves in a 19-move game (which is more than usual, because it’s a game from which to learn the right style in the opening).
When I was a kid, I missed a couple between moves 4 and 10, no one gets 12. Bxh7!!, and I threw the wrong knight check at move 14. 12 of 16 correct guesses is 75% (while 13 of 16 hits the mark at 81%; like I said, a narrow window). But 75% correct in the Masyutin game was much better than I did at the Tartakover game, and that includes after working through the Tartakover game once or twice before!
I’ve had students tell me over the years that they guessed those two games equally well; I think they were 1) kidding themselves, or 2) racing through both of them. Purdy said the best procedure is to set up a board in front of you, and actually make the move on the board before you uncover the actual move, because if you don’t, it’s easy to talk yourself into thinking you made the right guess. (These days, most students — and I, for that matter — play through games digitally, so we don’t have the option of making a move, then uncovering the right move. It’s easy to kid yourself that way.)