Home Page arrow Chess Life Online arrow 2007 arrow February arrow Problem Anatomy
Problem Anatomy Print E-mail
By Steven Dowd   
February 5, 2007

by Dr. Steven Dowd

My editor, Jennifer Shahade was confused by some concepts from my last column, such as "set play" and "try play." We decided that I should break down some of those difficult concepts in a special bonus "solutions" column. I like this idea, and whenever possible, and especially when the full understanding of a problem requires more explanation, we will do this!

Some people want just to solve the problem. When I first started solving, I felt the same way. What’s the answer? was my main concern. But sometimes, as Trinity noted in the first Matrix movie, “It’s the question that drives us, Neo.” Brute force in calculation is fine, but you don't appreciate the problem that way. Knowing the whys and wherefores behind problems will.

Anatomy of a Problem with Good Actual Play, Poor Try Play

Original for Chess Life Online February 2007
Steven B. Dowd

White to Play and Mate in Three

Weak tries
Idea: 2.Qf4+ 2…K×e6 3.Qf6# 2.Bb4 [3.Qf6#] 2…K×e6 3.Qf6#
but 1…Kd6!
Why Bc6 is a weak try:
because the black king easily approaches the B on c6 to refute the try.

Idea: 2.e8=Q+2…Kd6 3.Bf4#
1…Kd6 2.e8=Q [3.Bf4#] 2…Kc7 3.Bf4# or 3.Qc5#
1…Ke6 2.e8=Q+ 2…Kd6 3.Bf4#
but 1…K×e4!

Why 1.e7 is a weak try:
There is some small justification for the idea, although it is fairly easy to see that the black king has plenty of room to escape to d3, c2, etc. Also, promotion to queen, unless it shows some special aesthetic effect, is rarely considered since the queen is such a powerful piece.

Idea: 2. Bc6 [3.Qf6#] 2…K×e6 3.Qf6#] 1…K×e6 2.Bc6 [3.Qf6#] 2…Ke5 3.Qf6‡
but 1…K×e4!

Why 1.Bb4 is a weak try:
Because the very piece that must be moved on the second move for the sequence to work is easily captured on the next move. A flight square is removed from the king.

Actual Play (Solution)

Position after solution, Qf8!

1.Qf8! (ZZ)
Kd4 2.Bc3+ (2... Kxe4 3.Qf3#) Kxc3 3.Qb4#
1... Kxe4 2.Qd6 Kf5 3.Qd5#
1...Kxe6 2.Bc6 Ke5 3.Qf6#

ZZ= zugzwang

The solution is interesting-The one drawback is the give-and-take key, meaning that one flight square is given back to the king, and one taken away. It is preferable to give flight squares to the king with the key, or not remove any. One could also argue that it would be interesting to increase the thematic play; for example, I would love to start with the white queen on f3, when the second mate would illustrate a thematic/schematic device known as a switchback.

The sweeping queen key is attractive to my eye – this is a matter of taste, but many solvers like such keys. The two bishops are sacrificed, with model mates. We have four diverse queen mates; I find 3. Qd5# the most appealing, with its two quiet moves, to compliment the sacrificial lines.

Position after Qd5# variation

This type of problem was very popular in the 1920s, especially amongst Czech and other composers, and is called a bohemian. Its main value is in the construction – it has an exceptionally fussy matrix in which small changes (take the a6 and a7 pawns off the board and watch the solutions change….) lead to big differences.

As opposed to the interesting actual play, the mates the tries provide are not interesting, nor do they show any thematic/schematic cohesion. If I were submitting this for publication, I would not indicate that try play was a part of the problem.

So I am saying my problem is simply bad? No, its fun to solve, the mates are interesting, but that’s it. That is all it offers. When I compose these sorts of problems, I often call them my “chess popcorn” – Chris Feather, the British helpmate king, calls these nicely schematic problems “matebuilders.” It’s not that hard to set these problems up today, especially with computers. One often finds them in small publications – I also call them my “newspaper problems” because I like to publish them in newspaper columns in various countries, where, again, you have folks who are mainly chessplayers who like to solve a problem now and again

Making the Try play interesting
Before we go over the problem's from last month, which will illuminate effective use of try play, let's look at a “simple” miniature to show good try play in “simple” form

Bob Lincoln is known around the world for his mate in two miniatures. They are filled with try play. Here is a recent one:

R Lincoln
Kudesnik 2006

White to Play and Mate in Two

I am going to list a variety of attempts to solve this problem, before going any further, determine which ones are tries and which ones are simply attempts and why. One is the solution, and you should be able to show why it works:

a. 1. Qh1
b. 1. Qg2
c. 1. Kc3
d. 1. Qd6
e. 1. Qc6
f. 1. b6
g. 1. Qb7

First of all 1. Qh1, 1. Qg2, and 1. Kc3 are mere attempts since more than one black defense can defeat white’s attempt. With 1. Qh1?, white is trying to mate the black king with 2. Qd1. But this fails to three black defenses, 1… ab5; 1. … Ka5 and 1. … a2. 1. Qg2 wants to mate with 2. Qc2, but this fails to both 1. … axb+ and 1. … Ka5. 1. Kc3? has 2. Qc8 or 2. Qc4/b3# in mind, but fails to 1… Ka5 or 1. … a2.

Bob builds his problems with all this and more in mind, which is why they are deceptively simple. Now 1. Qd6?, 1. b6? and 1. Qb7? are all actual tries; the first failing due to 1. … axb, the second due to 1. … a2 and the third to 1. … a5.

Note that the tries are not only defeated by only one move, as they should be, but they also show mating ideas similar to the actual solution 1. Qc6! and are very unlike the first three attempts. Take another look and see if you can see this. It’s not hard to find 1. Qc6 here; computers do it in milliseconds. But if you let the task become mechanistic, you don’t appreciate the problem for its full value. There are three basic mates, after the key, whereas the attempts showed rather crude, usually single line mating attempts: 1.Qc6 axb5+ 2.Qxb5# ;1...a5 2.b6#; 1...a2/Ka5 2.Qxa6#.

Set Play Basics

Many people say set play is the play seen if the first side could pass; in directmates, it is simplest to think that it is what would happen if it were black to move. Thus these are the mates white has set or as you might also say, as “ready to go.” So take a look at the following mutate, one of the first ways in which set-play was manipulated. In a mutate, there are a number of set-play mates, but after the key the mates change.

In mutates, White has no useful waiting move,so she must change the mates. I always thought of mutates as problems in which White was in a state of near zugzwang in which only one move worked. The move White makes typically places Black in a total zugzwang from which there is no escape.

Mutates are still seen today, usually with great depth and sophistication. John Rice gives the following as from a modern practitioner of the mutate:

RT Lewis
The Problemist 1984

White to Play and Mate in Two

Make a list of the set mates you see. I, like John, count eight. Here are two examples: 1. … Bxc5 2. Qxa8#; 1. … d3 2. Be6#

You should have seen that if the S on f8 goes anywhere, white mates with 2. Qe6#. That one is thusly notated: 1. Sd8~ 2. Qe6#; the others: 1. … Sxe4 2. Be6; 1. …e5 2. Qd6#; 1. .. Bb8/b6/c5 2. Oa8# 1. … Bc6 2. Qxa2.

Now the solution 1. Qxb5! (ZZ); 2. … S8~ 2. Qxd7; 1. … Sxe4 Rxd4# 1. … d3 2. Sc3# 1. … e6 2. Sxf6#. 1. … Bb8 2. c6; 1. … Bb6 2. cxb6# 1. .. Bc6 2. Qc4# 1. … Bb7 2. Qxb7#

Position after solution, 1.Qxb5

Now look at what he has done in brief that is impressive: there was one mate for 1. … Bb6; 1. … Bxc5, and 1. … Bb6 – a rather crude capture mate. Now there are three different mates for all of these moves – bravo!

Solutions (from January's Check-Up)


1.Sec5! Rxa4+ 2. Ba2# 1...Bxg4 2.Ba2#; 1...0–0–0 2.Sb6#; 1...Bh7 2.Qd7# Bonus: The set is 1. … Rxa4 2. Qxa4#, which changes, as we see in the first solution to 2. Ba2#. I only included this in the bonus because, although pickers of nits might say “there is also a set 1. … Bxe6 2. Bxe6#, and it isn’t changed.” They would be correct and this is the very reason I did not include it – it really has no bearing on the problem.


a. 1. Bf5! Which threatens 2. Rb8 mate; black can castle out of that but then falls prey to 2. Rg3#. The try in (a) is 1. Rb8, which fails to 0-0! With the h5 pawn removed, black can’t castle as his last move must have been with the king or rook, so the try is 1. Bf5 to threaten Rb8 mate, but after 1 … Rh3! that doesn’t work. So 1. Rb8! which threatens 2. Bh3#.

This sort of cyclic interchange is very popular amongst problemists, and is shown here in a simple, yet attractive form. For the bonus (100% credit), you had to show this cyclical interchange,. I mention the 100% credit because for the bonus, I will give partial credit for an answer that is close (terms differ for players) and shows thinking.

So, if you can’t identify the two important tries, all you will see is that this problem has two nice little solutions. And sometimes you don’t see it, and it comes out in the solutions list (as we are doing here), and you may go, “Hey, I solved that problem but didn’t see that interchange idea! Neat!” And there is nothing wrong with that. No one is so good they see all the nuances of a problem at first glance.


No less than three line pin mates are shown (bonus) after 1.Sd6! Ka3 2.Qxb3#; 1...Qa3 2.Qb5#; 1...a1Q 2.Qb4#; and then the more prosaic mate 1. …Qxc2 2.Sc5#.


1.Sxf4 Bxf4+ 2.Qxf4# 1...Se5 2.Sg2#; 1...Bxf5 2.Rxf3#; 1...Sg3 2.Sd5#

My personal two favorites here are the S mates; they deserve a looking at! A fantastic start to a long and continually rewarding composition career.

There is no real interesting set play here – moves of the Bg6 and Sd7 to e5 or the Bb8 to a7, but the tries are phenomenal 1. Rxf4? with the idea of a Rf3# - 1. … Rh8! (not at all easy to pick out) 1. Qa5?, hoping to mate on d2, 1 … Bxd3! And if the idea of moving the queen wasn’t seen by you as bad – 1. Qd5? Se5!. Trying 1. Sc1/e1? Shows why there is a S on h1 – 1 … Sg3!

Sometimes I think if problems were explained as “White has to mate in two or he loses, ” more chessplayers would see this interesting side to problems.


1.Bd5 Kh5 2. Nxf6#; 1...Kf3 2.Nc3#; 1...Kh3 2.Nf2#; 1...Kxf5 2.Nxd6# Bonus: 1. Sc3? would provide four bishop mates in a star pattern – for example, 1. … Kxf5 2. Bd3# 1. … Kf3 Bd5 # and so on, but 1. … d5! is the only move that refutes the try. Solver Barry Keith very astutely noted that 1. Bd5 prevents 1. .. d5, and again shows why seeing try-play and its refutation is nice:

If you see the try and how it fails: your mind should of course then think, Ok that was neat. Of course, sometimes solvers see only the try and quickly assume that is the solution – see below for a chessplaying analogy. But they need to look at their attempt, think for at least a minute, and look for refutations. Then, when you find the actual solution (and the logic process has guided you here), it become doubly rewarding.

When I was playing chess, we often laughed at the old chess story “The threat is stronger than the execution,” because that again showed a desire for wish chess – for your opponent to fall into your trap, rather than be defeated by strong play. I imagine solvers who stop after an obvious try are much like those players….Some solvers think of this as a form of trickery: nonsense! It is a way to add content, to the challenge, and to the beauty of a problem.

Gary Ware also sent me some neat stuff on pattern recognition that I will email to anyone who requests it. It’s helped him already in his understanding of problems!


1. Rxe5! 1...Qxd8 2.Rh5#; 1...d4 2.Rxe4#. The set mates were all battery mates 1. … Qxh6 2. Rh7# 1. … Qc6 2. Re6# and so on. The actual mates are all unpins.

Now that you understand set play, take out a board and play this all over again, looking at how each mate differs. Impressive, yes? This is what has led to today’s modern problem, which contains as much of this kind of content as possible.


1. Rxe5! (again!) with the idea of 2. Kf7.1. ... Rc1 2. Kf8 Rxc7 3. Re4#; 1. ... Rd1 2. Kg8 Rd7 3. Re4# 1. ... Re1 2. Kxg6 Rxe5 3. Rf7#.


1.Sc5! threatening Sa3+(#2) 1...a1Q 2.Se4 threatening Sc3#(#1) 2...Qa3 3.Sxd2+ Ka2 4.Rxa3# ,or if black chooses the other logical promotion, 1...a1B 2.Sa3+ Ka2 3.Sb5+ Kb1 4.Sc3#.

E-mail Dr.Dowd with questions or comments on this article at [email protected]. Or simply post a comment below.