Adventure in composition

In this post we try to show with examples the adventure in the composition of a chess problem.
The new and very productive composer Diyan Kostadinov, coming from Bulgaria, proposed during the World Congress in Crete a composition contest with a new fairy piece, the KoBul King. (Ko Kostadinov, Bul Bulgaria).
The KoBul King acquires the abilities of moving and capturing of the last captured friendly piece. (If Black captures a white Rook, then the white King continues to be a royal piece but moves and captures exactly like a rook). The white King returns to its normal situation when Black captures a white Pawn.
We have now a piece that is controlled by the opponent, thus nice problems with tactical play can be composed.

The crucial fact today is that there exists no software to check compositions with the new piece, so we have an opportunity to work as we used to work many years ago : only with our minds.

We decide to compose a helpmate in two moves.
Idea : In a helpmate problem, black is moving first.
B1 : [The black KoBul King bKK will go to a square].
W1 : [The white KoBul King wKK will capture a Bishop and will become instantly KKB (KoBul King Bishop)].
B2 : [The KKB will move someplace else...].
W2 : [...where will be threaten by the wB (but the bKKB will not be able to capture the wB, because wKK will be transformed instantly to wKKB and will capture the bKKB) thus will be mated].
There will be also a second variation with Rooks.

Let us see the Plan A.

	(Problem 471 Plan A) Emmanuel Manolas, original (please do not copy), 07/10/2010, h#2 2111 KoBul Kings (4+4)
[8/8/2p5/8/2P2k2/3B4/2bKR3/4r3]

Desired solutions :
1.KKg4 KKxc2(bKK=KKB) 2.KKBd7 Bf5#
1.KKg5 KKxe1(bKK=KKR) 2.KKRc5 Re5#
Seems to be good. The white pieces do not stay idle in the variation where they have not the first role. In the first the wR holds e8, in the second the wB holds c4.
Unfortunately there is a cook :
1.KKf3 KKxc2(bKK=KKB) 2.KKh1 Be4#
Let us move all the pieces one place left to help bKKB escape. See Plan B.

	(Plan B) h#2 2111 KoBul Kings (4+4)
[8/8/1p6/8/1P2k3/2B5/1bKR4/3r4]

The desired variations, slightly modified, still work.
Unfortunately the cook remains because wRd2 holds h2 (But how did that escape from our keen eye?) :
1.KKe3 KKxb2(bKK=KKB) 2.KKBg1 Bd4#
Let us move the position one file left, hoping that bKKB will manage to escape. See Plan C.

	(Plan C) h#2 2111 KoBul Kings (4+4)
[8/8/p7/8/P2k4/1B6/bKR5/2r5]

Besides the desired solutions, a new cook has appeared :
1.KKe4 KKxa2(bKK=KKB) 2.KKa8 Bd5#
So, if we can not correct the problem moving the pieces to the left, let us try to move the initial position A one file to the right. See Plan D.

	(Plan D) h#2 2111 KoBul Kings (4+4)
[8/8/3p4/8/3P2k1/4B3/3bKR2/5r2]

Unfortunately this placement has got six solutions as normal helpmate, totally undesirable for our goal, to compose a KoBul Kings helpmate.
1.Kh4 Kf3 2.Ra1/Rb1/Rc1/Rd1/Re1/Rg1 Rh2#
Will we be forced to add a white pawn to the initial position, to stop bKK from going to f3? Let us see Plan E.

	(Plan E) h#2 2111 KoBul Kings (5+4)
[8/8/2p5/8/2P2k2/3B4/2bKR1P1/4r3]

It does seem safe, doesn't it?
Would you say that the following is a cook?
1.Rxe2(wKK=KKR)+ KKRxe2(bKK=KKR) 2.KKRf1 KKRxc2(bKK=KKB)#
NO, it is not! Black continues with 3.bKKxg2(wKK=KK)! without fear of threat.
It seems safe, but the idea of putting a white pawn was not good, because we should try first to find a better arrangement of pieces adding black pawns.
Let us see Plan F. It is Plan B with a bP on h5 and the bK on h4. We expect it to have the two desired solutions starting with keys 1.Kg3 and 1.Kg5.

	(Plan F) h#2 2111 KoBul Kings (4+5)
[8/8/1p6/7p/1P5k/2B5/1bKR4/3r4]

We the amount of experience we have by now, it is easy to see the cook (the cooks) in this edition.

1.Ra1/Rf1/Rg1/Rh1 KKxb2(bKK=KKB) 2.KKBe1 Rf2#

What would you say about putting a bP on f2? See Plan G.

	(Plan G) h#2 2111 KoBul Kings (4+6)
[8/8/1p6/7p/1P5k/2B5/1bKR1p2/3r4]

Is seems to be ok, but it is not economical. And that white pawn on b4 could be removed. But wait a moment! The black pawn on f2 can stop the move Kh4-e1, but it can also be promoted. And here is the new cook :

1.f1=S Kxb2(bKK=KKB) 2.Sg3 Bf6#

Let us turn the chessboard 180 degrees to avoid promotions, and also replace the white pawn. See Plan H.

	(Plan H) h#2 2111 KoBul Kings (3+6)
[3r4/1p1RKb2/4Bp2/2k5/8/5p2/8/8]

We expect to have only the desired solutions with keys 1.Kb6 and 1.Kb4.
For a moment we thought we saw a new cook :
1.b6 Rxd8(bKK=KKR) 2.KKRc6 Rc8#
Luckily there is 3.KKRxe6(wKK=KKB)+! and the bK is completely safe.

But we cannot escape so easily! See this cook :
1.b6 Bc4 2.KKc6 Rc7#

Last try : We move bPb7 to c6. See Plan Ι.

	(Problem 471 Plan Ι) Emmanuel Manolas, original (please do not copy), 07/10/2010, h#2 2111 KoBul Kings (3+6)
[3r4/3RKb2/2p1Bp2/2k5/8/5p2/8/8]

Solutions :
1.KKb5 KKxf7(bKK=KKB) 2.KKBe2 Bc4#
1.KKb4 KKxd8(bKK=KKR) 2.KKRf4 Rd4#

(Five from nine pieces are on white squares, so the problem will be discernible when printed. We do not flip it left/right).

Mr Christian Poisson has promised that in the next edition of his software WinChloe, it will include the condition KoBul Kings (Rois KoBul?). We will check then to see if this problem we composed today is cooked or not.
With computers we lost a great part of the adventure of composition, but cooked problems have ceased to be published any more.

I hope that I have relayed to you a part of the fascinating process of chess problem composition. The friends composers Themis Argyrakopoulos (from island Ios) and Kostas Prentos (from Thessaloniki) had crucial role in cook-hunting.

Now the problem will be sent to a composition contest to be judged, and maybe receive a distinction.