Strip or massage?


by Steffen Slumstrup Nielsen (“The Macedonian Problemist” no.67 2022)

When Carlsen, Firouzja and the rest of the speed brains get together at super tournaments, I like following the live streams to be entertained and inspired. I am often struck by a beautiful move, a series of moves or an unusual tactical device from top games and indeed from games at all levels. Over-the-board games are one of four inspirations for me as a study composer. The others are: 1) previous studies/problems; 2) lines suggested by my computer; and 3) toying around with the pieces until something interesting occurs.

Surprisingly often I watch sequences of moves in games that do not remind me of anything from my previous chess memory. When this happens I hurry to the board – well, to the computer – hoping to put the idea from the game into a good form that is suitable for an (endgame) study.

Strip – the well proven method

The traditional way to deal with a promising game fragment is to strip it. By this I mean removing all the units that are not necessary for showing the idea. It is important, of course, to do this and to show the idea as economically as possible. Over-the-board players are not limited by such refined worries in their games, and it is often possible to remove a large number of pawns and even pieces from the game position without destroying the idea. Experienced study composers can do this fairly quickly, improving both the economy and weeding out possible side solutions. In the process the position might be shifted or even rotated, but the main aim is to lower the number of pieces involved. I don’t consider this process to be particularly creative, but rather somewhat mechanical.

When you use the strip method, most of your creative powers are used in the second part of creating the study. Once you have stripped the board of unnecessary units, you still need to create an introduction leading up to the idea borrowed from the game. Creating a good introduction is not a top quality of mine. Apart from the creative aspect of working backwards from a given position and creating play with pieces that are not on the board anymore, generating introductions is also exceptionally hard work. Duals are everywhere and often can only be dealt with by adding extra material, captures and even piece exchanges. Keeping the introduction clean and interesting is a real craft. The masters manage to add extra ideas to the introduction, but more often than not, composers are happy to make a workable introduction at all, and even that is hard work.

In my view, the move played by Dubov is perfectly suitable for stripping. In fact, it can be shown with only six pieces, removing no less than 12 pieces from the game position. The following is the final product by Martin Minski and me.

Massage – altering the idea

I suggest that the alternative to the established strip method should be called massage. It is best explained with an example.

Instead of stripping, there is another route to travel, namely to stand back and try to decipher what Tal’s 21.h4 is really all about, looking at the position abstractly rather than concretely. My interpretation is that the move is fundamentally about “leaving pieces en prise and playing elsewhere”. This is no profound idea in itself, but imagine that Miller had the chance to react in a similar way, doubling the idea. Imagine he had some way to cover h2 so that the back rank would still be weak and the queen on e5 would still be immune. And imagine that Tal had a counter-counter, still leaving the queens to stare at each other in a perfect act of suspense.

I am far too unstructured to be able to retrace the steps that followed as I tried to create a study from Tal’s beautiful move. But basically I tried to turn the game brilliancy into a situation where both kings were potentially under threat of mate. Weak back ranks are perfectly suitable for creating and maintaining this kind of tension (and better suited than the threat on c7 in the game). After some months of work – where I probably ate and slept as well – I had the following study on my screen.

Notice that the final position still involves 14 pieces, which is roughly the same as after the attempt at strip shown above. The economy suffers.

My study came about by massaging the original game fragment. Massaging involves a more comprehensive change to the original position. One may argue that it is a change to the content/idea whereas stripping is a mere change of the form. The change when massaging can take many shapes: doubling the idea, adding black counterplay, diagonal/orthogonal change etc. In the present case, the idea of leaving the queen en prise was expanded by a number of moves. In the process an extra idea of leaving other (in fact all) pieces en prise was added as well. It is not uncommon that massaging an idea from a game ends up in an entirely different idea that has almost no connection to the original game fragment.

The escape of the bishops was very appealing and I felt it was suitable for a study. Again, stripping was possible, but I felt that what I really wanted was to prolong the escape of the bishops to more than the three moves shown by my opponent in the game.

Some massaging went on and relatively early in the process I stumbled on an ambitious idea. What if the bishops would be able to eternally escape? Or alternatively: what if a rook was able to eternally harass the bishops?

The idea that such a mechanism could even exist was probably very optimistic. In fact, this is often the worst part of composing with the massage method. Some abstract idea is conceived, but there is no guarantee that it is achievable.

This time, I was lucky. An extra rook was required and in fact also the active involvement of the king. Within a day I managed to compose a workable study.

The next version is more pleasing.

The final game example is quite extraordinary, because I believe a fundamentally new idea was found at the board. It is an idea that is perfectly suitable for studies, but which apparently had never been done before Valentina Gunina sat down to her first-round game in the Grand Swiss Tournament in Riga in 2021.

Gunina’s idea is easily defined:

  1. Piece A and piece B can both be sacrificed on square X.
  2. Piece A is sacrificed on square X. The sacrifice is declined.
  3. Piece A moves away.
  4. Piece B is sacrificed on square X.
  5. The sacrifice by piece A in some way logically prepares for the second sacrifice.

Having identified this concept, it would be a shame to use the strip method to find a study letztform of Gunina’s combination. In other words, the scheme used by Gunina (Black’s queen on b6 protecting square h6, but being interfered with on e6) is only one possible rendering of the theme and therefore massage is a more suitable method.

As it happened, the scheme from the game proved to be so excellent that it worked for an expansion even without the usual shifting around and remolding that is so typical of massage. In fact, the game may have been sent from heaven.

The obvious follow-up question is whether the Gunina-theme can be shown in another manner? Can there, for instance, be four sacrifices on the same square each preparing the following? Or might the piece from the first sacrifice return to the key square on the third sacrifice? I have no answers for such questions yet, but I know that only the massage method will be able to provide them.

Below I attempt to sum up the different characteristics of the strip and massage methods. Of course, there is no clear dividing line between the two methods and the composer may end up using a mixture of both.

Characteristics of strip:

  • Structured, mechanical process
  • Goal-oriented
  • Perfecting the form
  • High emphasis on economy, searching for a “letztform”
  • Limited change in material used
  • Emphasis on building a fitting introduction
  • Few pieces on the board
  • The study features tactics and manoeuvres typical of the endgame
  • Pointed finish (climax)
  • Guaranteed end product once you have the idea.

Characteristics of massage:

  • Chaotic, organic process
  • Process-oriented (curiosity driven)
  • Perfecting/Expanding the idea
  • Low(er) emphasis on economy
  • Large change in material used
  • Emphasis on building both introduction, main play and afterplay (nachspiel)
  • Many pieces on the board
  • The study features tactics and sequences typical of the middlegame
  • Often the high point (climax) of the study comes early and the finish may be unsatisfactory.
  • No end product guaranteed.

Without a doubt there is still plenty of fascinating novelty to be found in miniatures and other studies with relevance to endgame theory. The endgame remains the main scene of studies. Nevertheless, the middlegame study is definitely here to stay. In the recent Timman 70 JT the average piece count of the six prize-winning studies was 13!

Not surprisingly, the middlegame study finds a perfect inspiration in games. In my view, composers who seek their inspiration from games will be best served by using the massage method. The process is much more fun (though at times frustrating) and I am convinced the method is ultimately better suited for expanding the horizons of chess and producing better, more ambitious studies.

This article is a version of a short lecture originally given at the 2021 WCCC in Rhodes.