On Constructing Hnefatafl Puzzles

Monday, 13th January 2014

One of the characteristics of chess culture is the chess problem. You've probably seen them in newspapers, maybe even tried to solve one or two. If you've not come across them, they consist of a diagram of a position in a chess game, and an objective, such as "white to move and mate in two".

These have a very long history. In his book "A History of Chess", H. J. R. Murray devotes a great deal of space to them, with many 1000-year-old examples from Arabic literature, dealing with the mediaeval chess game of shatranj. More recently I've started to follow a Facebook page that gives regular problems for xiang qi, or Chinese chess.

Chess problems come in many forms, but the most accessible ones I've seen are of the "white to move and win" variety. The solution usually involves forced moves: white can only come to a predictable win by forcing black into a particular move on each turn, any other black move meaning an even earlier victory to white.

The problems are made interesting by a variety of devices. Usually, black has a counter-threat against white, so that if white fails to force a particular move on black, then black will immediately checkmate white. The problems often involve white sacrificing pieces to achieve the position that allows victory, making the winning moves look disastrous at first glance.

To date, hnefatafl doesn't have an equivalent culture of puzzles. Tim Millar posted a few to Aage Nielsen's forum, which can be seen at http://bit.ly/1all0Yv. But these are isolated affairs, and so far have lacked the rigour of chess problems - i.e. one tightly-defined solution.

I recently spent a morning looking into how hnefatafl puzzles could be contructed. And I've found it extremely difficult! So far, I've not managed to create a single robust problem that doesn't have an obvious solution. The closest I came was mid 2013, when I created the forced moves example for the "Short-term Tactics" page (see the link in the "See also" section).

In the composition of these problems, hnefatafl presents some difficulties that are not present in chess. First and foremost is the lack of standardisation. A problem will need to be accompanied by the game rules, or a reference to them, if players are to know how to solve it.

The simplicity of hnefatafl is a nuisance here too. Since all the pieces have identical moves, it becomes difficult to construct positions with the subtleties of chess. The long range of all the pieces increases the strategic possibilities, but this works against the construction of problems, which thrive on reducing opportunities to a few key options. The use of straight-line movement also means that moves stand out much more clearly on a hnefatafl board.

"Piece interference", often a key characteristic of chess problems, isn't possible in hnefatafl; in a chess problem, an ill-advised move of a bishop may block a threat by its rook colleague; in hnefatafl the moving piece would itself reinforce the threat.

Another difficulty results from the single king of hnefatafl. In a chess problem where white has to mate black, it may be that white's king is under attack. White has to be mindful of two situations on the board, and failure to force a move by black results in white being checkmated. Such a dual focus is impossible in a hnefatafl game, where all attention is focussed on the sole king.

Most of these and similar difficulties reduce to the observation that forced moves in a hnefatafl puzzle become more obvious than in chess, and therefore make it harder to compose a robust problem that isn't trivial to solve. Tim Millar's attempts, mentioned above, have resulted in problems with more diverse solutions than originally intended.

On the other hand, hnefatafl can provide some interesting possibilities. In chess, the colour to move and win is arbitrary; the problem would be no different if the colours were reversed. In hnefatafl, a problem where the king is to move and win is very different to one where the attackers are tasked with victory. The long move of the king means that problems, like real games, could take advantage of an unexpected switch from one area of the board to another.

Despite my morning's lack of progress, I'll be spending more time looking into hnefatafl problems. I think a good start would be to try and devise more and more varied forced-move situations, in a quest for some where the forced move isn't immediately obvious. If I'm successful, then you can expect series of hnefatafl puzzles to become a part of the site and its social media pages.