Last weekend I read with interest an interview with David Astle, one of the 7 setters who between them produce the daily cryptic and quick crosswords for The SMH and The Age. It’s almost a given that in the course of a reported conversation with a cryptic crossword setter that one of two things will be discussed: “fairness” in clues and crosswords, or difficulty (in terms of allowing solvers to eventually complete a crossword). The first part, “fairness”, I’ve talked about a number of times: a non-existent factor as far as a setter is concerned (since no-one aims to create unfair clues and crosswords), and a variable one as far as a solver is concerned (since “fairness” isn’t measurable and exists in the eye of the beholder). The second part, difficulty, I’ve also talked about, but not in terms of the solver-focused crosswording philosophy of a setter. As I was pondering this the interviewer posed this question: “…was his (David Astle’s) aim to defeat people? Not so, it seems. It’s the art of losing gracefully.”, an often-quoted phrase used by setters when talking about presenting a challenge to solvers. I’ve used this language myself in the past and recently – “I’m happy to lose the battle with solvers, in fact, I must lose to win.”. But what does it actually mean, how does it shape a compiler’s mindset, and how does it manifest itself in cryptic clues and crosswords? I imagine solvers nod thankfully when setters say such things, boosting their belief that they can actually solve everything given the right circumstances as setters have indicated that they have left the door open for them. Sadly I don’t think most setters do anything specifically to ensure they “lose gracefully”, they simply create clues they believe can be solved and leave it at that. That is, they equate “losing gracefully” with “creating a solvable crossword”, which, in my mind, doesn’t go far enough. Why? All cryptic crossword clues that are part of a crossword are solvable. There’s no such thing as an unsolvable cryptic crossword clue: even the most complex wordplay construction coupled with an obscure definition for an archaic word is solvable by someone; even a clue with impenetrable wordplay containing mistakes can be solved as long as the definition has a remote connection with the answer. Solvability alone can’t be used as a gauge for fairness or difficulty, and besides, who sets clues with the aim of completely stumping solvers? So, scrap solvability as a factor in losing gracefully, and what’s left is the setter’s attitude towards clueing, that is, actively considering how it would be for a solver when tackling clues. Let’s look at aspects of setters that I believe do demonstrate that they are aiming to lose gracefully: they don’t overly use obscure or foreign words and meanings; they don’t expect solvers to spend a lot of time researching on the net; they don’t use wordplays that can only be resolved in hindsight; they don’t use narrow themes that restrict the solving pool; they create crosswords commensurate with the time allocated to solve (that is, a daily cryptic should take no longer than a day to solve); they provide two distinct ways in each clue to obtain an answer; they consider the ability of all solvers, allowing more than just the elite to enjoy some entertainment; they consider all clues from a solver’s perspective. There are plenty more, but you get the drift. There are a number of difficult cryptics out there and I would argue, that if they are consistently difficult for some of the reasons above, the setters don’t really understand the concept of losing gracefully. Surely a solver of average ability, who can solve most cryptics available, is already beaten if they don’t attempt a crossword due to a setter’s constant desire to bamboozle in the extreme?