Or, to ask the question a bit less opinionatedly, "what new ideas have been introduced in chess problems over the last 20 years?" I imagine many problems have been composed in that time, but it seems like all of the stipulations (model mates, selfmates, etc.), all of the structural elements (pinnings and unpinnings, interference, and so on) and all of the tasks (Phoenix tasks, etc) have been long since sussed out. The only relatively recent development that I'm aware of are various 'Babson cycles' that have been composed by Peter Hoffmann (see Wikipedia's page on the Babson task for details). Are there any other relatively recent developments of note in chess problems?
An unsorted list of arguments for the vitality of chess problems (probably my answer is a little bit biased because I am an enthusiastic chess composer ...)
Computers: A lot of new ideas have been developed during the last decades. One reason for it is the rise of computer programs which help to check the correctness of a chess problem. This simplified the chess composition dramatically, especially selfmates and helpmates with very complex and long solutions are now possible...
Helpmates: by the way help mates. This is an area I know a little bit and during the last 20 years there was phantastic progress. Here some non representative samples http://kobulchess.com/en/problems/originals2013/321-chess-problem-helpmate-viktoras-paliulionis.html or http://www.yacpdb.org/#320933 or http://pdb.dieschwalbe.de/search.jsp. Study the solutions and look for other problems of these authors (and other like Caillaud, Mazlar ...) to find other good help mates ... also there are still many open themes which are not realised. such as the '100$ theme' , the 'Oudot theme' or a helpmate with more than 27 moves...
Specialisation: Chess compositions branch in many different ways. Some "theoretical" knowledge on certain themes is restricted only to a very small group of composers. So for example if you ask for the development of some themes for three movers (mate in three) the current state of the art is probably known only to a very small group of composers (probably less than 100 people world wide). I am definitely sure they know exactly how their field developed during the last years. And definitely they will argue for a good progress in their field of work. But from "outside" the progress seems very modest (of course I don't want to blame the three mover community but this is an area I know almost nothing :-)... There is a similar situation in mathematics or other scientific areas where the current state of development is often only understood by a small group of experts...
Fairy chess: If you think of chess problems in the broader sense of generalized chess the we come to fairy chess. This is the genre where the chess rules and the stipulations are extended far beyond the orthodox setup. For example new varieties of chess (varieties of circe chess, ghost chess ,... et.al.) and many many different pieces have been developed during the last decades and all of them help to represent new ideas in chess problems (or can you think of triple check or ...). For much more on fairy chess look at juliasfairies.com
selfmate, retro problems ...: There are a lot of other areas where chess problems develop. For example look at the (german) retro blog http://www.thbrand.de/retroblog/ where you can learn a lot about retro chess (if you are aware of german)...
I hope all this helps to believe in the vital functions. For much more infos check the websites I mentioned above
Oliver Sick listed many good examples. Here are some of my thoughts:
- Andrew Buchanan's "Dead Reckoning" (a consequence of Article 5.2b of the Laws of Chess)
- Parry series mates
- Goals other than (stale)mate, e.g. check, capture, or White to move a unit to a specified square.
Further developments in genres that are more than 20 years old, e.g.:
- proof games
The idea of the task is more than a century old, but recently new sorts of task have been invented and worked on, e.g. construct the longest sound problem of a specific genre and with a specified number of units (or with specified force) in the diagram.
There's more information about various chess problem genres at
I can only speak of the direct twomover genre, where I am...less nonexpert. The twomover has been declared dead more often than all members of the X-Men combined, and yet it is still alive. True, the classics all have been found. Probably. But there always have been new ideas. (Suggestion: look up who got a grandmaster title in the field, check his problems in the original sources and what the audience commented.)
And Fairy Chess is unlimited by definition (and practically depends only which novelties catch on with composers and public).
I'm not aware of any recent examples beyond the "cyclic Babson" type of problem mentioned in the Wikipedia article you posted. I would agree that virtually all meaningful tactical motifs (pins, deflection, desperado, different types of mates, etc.) have been classified, as well.
In my opinion, problems like the Phoenix example are remarkable because they defy our natural impulse to evaluate a position against an instinctive set of strategic patterns and rules, and instead bluntly illustrate the fact that chess is fundamentally a tactical game.
The Babson Task class of problems seem to have particular aesthetic value because they exhibit "other interesting properties" (I'll call them OIP) - in this case, the symmetrical beauty of White's promotion to any of the four possible pieces that Black promotes to (or some asymmetric mapping in the cyclic case) - as part of the forced sequence.
I would argue that the OIP within such chess problems augment the aesthetic value of the composition; if a problem has no or some incomplete set of OIP in its solution, then its aesthetic value is less than a problem which contains a more complete or harmonious set of OIP. Thus, the example of the Wolfgang Pauly composition mentioned in the Babson Task Wikipedia article could be said to have less aesthetic value than a true Babson problem because of the fact that the Bishop under-promotion does not force a win.
As an example of why the chess problem might not be a 'dead' form of art, I might conjure up a hypothetical class of problem with an intriguing set of OIP, or extend the depth of an existing scheme, for which an example might at least conceivably exist. I'm no good at this, but as a poor example, let's take a variation of the Allumwandlung pattern: perhaps a set of initial conditions exist such that White can forcibly win by promoting an a-pawn to a Rook, or a b-pawn to a Knight, or a c-pawn to a Bishop, or a d-pawn to a Queen. The OIP of such a solution would have obvious relevance for the fact that these files correspond to the starting position of Black's pieces. Voila - define a set of OIP, in this case a more restricted type of Allumwandlung (the easy part), and if one qualifying initial condition can be found to satisfy these boundary conditions (the hard part), a new class of problems is born.
On the other hand, perhaps we have already classified all plausible and meaningful OIP.
While I find it difficult to believe that the chess problem as an art form is dead, I do think that there are limits to to the number of classes of chess problem that would be of interest to human beings, due to our relatively limited ability to evaluate concrete lines beyond a certain depth.