Checks on the wK are always of interest in a two mover. Recently I have been exploring an idea that changes a set capture reply to a check by exploiting a self-pin. Specifically, there is a battery aimed at the bK with the front piece pinned. In the set play the capture of the front piece checks the wK and is countered by a capture by the rear piece of the battery. The key will make some type of error that abandons this reply but now White exploits the fact that Black self-pins when it checks. This idea can lead to some fantastic keys. Please also see my arcitle in the September 2016 Problemist Supplement. We begin with a classic by Mansfield.

Mansfield executes the idea to prefection in this beautiful Meridith. Notice the set check 1...Rxc7+ 2.Rxc7. However, this cannot be preserved because the wB must withdraw to mount a threat and in doing so will give a flight square at d4. This check forces the wB to the square h6 so it can interpose the check. The reason that the set mate is abandoned is the flight on d4 after the wB moves away. 1.Bh6! (>2.Bg7) 1...Rxc7+ 2.Bg7 (threat) 1...Bf2 2.Qd2 1...Kd4 2.Qe3 1...Rb6 2.Sb5 By the way, I think he should have gone with a more ecomical version: -wPa2,wPb3,bPb4 +wSa5,wPa3. He probably had them there to avoid duals, but that might be the only time I ever improve a Mansfield problem! |

Here is my version of Mansfield's problem in which I change the axis from diagonal to orthogonal. Set play: 1...Qxf3+ 2.Bxf3. This check forces the wR to d6 to self-pin and block the oncoming check. There is not much play but the economy is nice and the rendering of the idea is clear. 1.Bd6! (>2.Rd5) 1...Qxf3+ 2.Rd5 (threat) 1...Sc6 2.Rf5 |

Here is a favorite problem of mine. In the above problems the changed check is met by the threat. This problem shows that a threat is not necessary to pull off the idea! In fact this is the only such problem that I know of that does not have a threat. 1.Re3 (-) 1...Kd8 2.Bb6 1...Bb7/Ba6 2.Qxd7 1...Qg7+ 2.Rc3 Beautiful execution in a near minature! |

We now turn to the idea in which the key piece is a knight. Here is an example that I concocted. It is the only example I know of that is a minature. The problem contains a double threat but each of the threats is forced. 1...Qxe5+ 2.Rxe5 1. Se4! (>2.Sf6/2.Sg3) 1...Qxe5+ 2.Sf6 1...Qf1 2.Sg3 |

Here is quite possibly my favorite composion that I have made. True, it is not the most complex problem, nor does it display a modern theme. But it is so elegant. No pawns, the bK is mirrored (nothing occupying his field) and there are some nice model mates in the solution. 1...Qxd7+ 2.Rxd7 1.Se6! (>2.Sc7) 1...Qxe6 2.Bc6 1...Qf4 2.Sxf4 1...Be5 2.Se3 1...Bd4 2.Qxd4 Even though it is met by the threat, the mate following the check by the bQ is worth another look. That's right, we have a pin-mirror-model mate that is quite possible the most beautiful mate contained in any of my chess problems. |

In the previous problem the key always ungaurded a square that was needed for the set reply to the check. We now look at problems where the key makes a different error. The follow problem was used in the 2004-05 British Solving Championships. Notice the half-pin along the 6 rank. The set check exploits this: 1...Bxe4+ 2.Bxe4. The rook must retreat to make the threat but it can only retreat so far 1.Rh1? 1...Bxe4+! The key is 1.Rh2! threatening 2.Rc2 (to block the oncoming check). The bPf7 plays a star role by opening the bQ's line but interfering with the bBs: 1...f6 2.Sxd4 and the amazing mate 1...f5 2.Re6. Let's analyze that last variation. By moving to f5 the bP opens the bQ's line thereby defeating the threat, but interferes with the bBg6 unpinning the wRe4 which then pins the bPd6 and shuts off the bQ to mate. Bravo! |

Here is one of my own half-pin abandonments. This time there is extra set play to make the half-pin even more enticing. 1...fxe4/Qxe4+ 2.Rxe4 1...f4 2.Shf6 The R+S batter is awefully strong and maybe the key should be something that liberates it. Moving the wK is an option. 1.Kd2? (>2.Sef6) 1...Qa2+ 1.Kd1? (>2.Sef6) 1...Qb3+ 1.Kf1? (>2.Sef6) 1...Qc4+ Instead what we have is 1.Qb5! (>2.Qe2). And now those same moves that defeated our threat in the tries allow it as a variation: the Dombrovski paradox. Critics might say that the checking refutations are too strong. |

Finally we end with some of my problems that have been published but whose solutions are not yet available. Happy solving!

Here is my only problem published in the prestigious Die Schwalbe. Notice the set checks - one of them will be changed. Also this problem accomplishes something call regisive refutations. There are 5 "tries" that are defeated by more than one refutation and these refutations decrease with the different tries. |