Here is my only other composition, done when I was 12. White to move, black to win, or black to move, white to win.
Bill Smythe
Yeah for some reason in my head I had white at the top. Good thing there were no pawns to move. 
One person I showed this puzzle to asked me, “Which way is white going?”. I was tempted to answer “right to left”.
Bill Smythe
Although it’s obvious to most chess players, it might be a valid question. I noticed on youtube, “Kingcrusher” channel, he does defense videos with black at the bottom of the board. Although the board does have algebraic notation along the side and bottom of the chess board with the lower left corner being h8. I understand he’s trying to rotate the board in a way that “helps” the viewer, but in my opinion, it’s better to go with the industry standard and always have the bottom with white, with a1 being in the lower left corner.
Anybody know the difference between CT-Art and the Chess Tactics 1/2/3 from Chess King?
Bill’s answer is illuminating. In this position, with no pawns on the board, it doesn’t matter which way the board is set up - (other than the actual designation of the squares) it doesn’t matter if white would be at the top, bottom, right or left.
My point was, that if there are no pawns on the board, it doesn’t make the slightest difference which way white is going. It could even be sideways.
Actually, though, it might make a difference in a few positions. If there are a white king and rook on d8 and a8, the board might be upside down, and castling might be possible. Or, if at h5 and h8, sideways, with the same possibility.
Bill Smythe
Good point. +1
I clearly will have a long road to follow. I went through some random tactical problems from Chesspuzzles.com and failed miserable. 
No idea what level the problems were, but I’ll have to start with the basics and go from there. I think maybe one of the tactics series from Chessking will be a good starting point. I emailed them to clarify which tactics course I should get, since there is at least 4 different ones.
-Not to mention opening and endgame stuff I’ll get to eventually. But right now it’s the middle game tactics that seem to be most problematic for me in club play. Most of my games are decided in the middle game where tactics rule. I have enough opening knowledge to steer most games into some sort of decent middle game.
You must have showed this to me at Jules Stein’s over 35 years ago.
Now let’s see a proof game There are plenty of pieces that were captured, but the kings have to be walled in first, and the captures have to occur on specific squares.
Once upon a time I did come up with a proof game, as I was curious as to whether it was possible. Each side must have made 5 captures with pawns, with 4 of them (for white) being on dark squares. This should be easy, as each side has had 7 of its pieces (not even counting pawns) captured. It might be necessary (or at least helpful) for black’s light-square bishop to have been captured at b3. As for the missing g- and h-pawns, 1.h4 h5 2.g4 g5 3.gxh5 gxh4 4.Rxh4 Rxh5 takes care of them nicely.
One feels a moral necessity that the moves of any proof game be as symmetric as the position, i.e. whenever white plays Uwn-xm, black responds with Vyk-zj, where U=V (same piece), w=y (same departure file), x=z (same arrival file), n+k=9 (same departure rank), and m+j=9 (same arrival rank).
Let’s have a contest. Who can come up with the shortest symmetric proof game?
Bill Smythe
OK, I have entered my own contest. I have come up with a symmetric (i.e. copy-cat) game which leads to the diagrammed position.
The rules of the contest are:
- (1) The moves must be legal and must lead to the position above.
- (2) The game must be a copy-cat game. Each black move must be a mirror-image copy of the corresponding white move. For example, if at any point white plays n.Ng4-h6, black must respond n…Ng5-h3.
And, just for fun, let’s throw in another restriction:
- (3) There must not be any pawn promotions.
As you start to work on this, you will quickly realize that the copy-cat requirement severely limits the possibilities.
Contest winner is the person who comes up with the shortest game that meets the above criteria.
My entry (which I shall not reveal just yet) is a 26-move game. Can anybody do better?
By the way, I also see that nobody has as yet submitted a solution to the original problem, which was (in the diagrammed position) white to move, black to win, or black to move, white to win. (Of course, with a copy-cat game leading to the diagram, it would have to be white to move.) I’m sure, though, that Bill Brock knows the solution, because decades ago either he figured it out, or (less likely) I showed him.
Bill Smythe
Once P-K4 has been played, take it … P=K8/N and N-B7#.
Yes, I know I wrote the letter, but with the symmetry I thought I would throwback to DN.
OK, now we have a solution to the original “white to move, black to win” puzzle, but what about the other contest, to come up with a copy-cat proof game with no promotions, in 26 moves or less?
Is anybody ready for a hint or two?
Bill Smythe
OK, I guess everybody needs some hints.
-
First, the kings must have arrived at a3 (a6) via b3 (b6), since if they had arrived via b4 (b5), both kings would have been in check, illegal.
-
Second, in a copy-cat game, no rook can cross the center line. For example, if white plays n.Rh2-h5, black cannot respond n…Rh7-h4 without running into the other rook. Queens, bishops, and knights do not have this problem.
-
Since all five of the necessary pawn captures must involve pieces that have crossed the center line, it follows that none of these pieces can be rooks. Therefore, the five pieces captured on these squares must be the five remaining pieces: two knights, two bishops, and queen.
-
Only one of the capture squares is a light square for white, and only one is a dark square for black. Therefore, black’s light-square bishop and white’s dark-square bishop must have been captured at b3 and b6, respectively.
That oughta make it rather easy.
Bill Smythe
Since coming up with that game involves retrograde analysis, I see that Raymond Smullyan, who wrote at least two books with retrograde problems, passed away recently.
Where is the White King?
Umm, where is the white king for what?
I think the point is to ascertain the unique square that the White King must occupy to make the position legal.
Hmm, interesting. The white king is on c3, having just captured there from b3 after black had played …b4xc3 en passant double check.
Bill Smythe
Correct, jwiewel got if first in a PM. With WKb3, WPc2, and BPb4; Black just played 1. … Bd5+ from anywhere else on the h1-a8 diagonal, maybe captured something at d5. 2. c4 bxc3 ep+ 3. Kxc3. If there was a BP at c3, the last move would have been 3. Ka3.