Australian Crosswords

A few years ago I ran a number of community college courses on how to solve cryptic crosswords. Each course covered the whole term (2-hour sessions for 8 weeks) as I believe there’s a need to learn and do and repeat. It’s easy enough to explain the basic devices to people and show a few examples in a short space of time, (in fact, I believe many people would walk away from such a session like they could take on the world,) but that head knowledge is not enough to form a true foundation of cryptic learning. What’s needed is a steady diet of cryptic knowledge with the opportunity to go away and practise and come back with questions to fill in the gaps. Recognition of the various devices is key to tackling a crossword, but these really can only be learnt one at a time, and need to be understood before moving on. Unfortunately even the most basic published cryptic crossword has multiple devices so is at odds with the initial learning process, which is why an extended course is the best way to learn how to solve cryptics.
A quick presentation or afternoon overview is likely to leave potential solvers deflated rather than inspired as the first crossword they see afterwards will bring them straight back down to earth.

In my early Australian Financial Review (AFR) days, I was told to create a difficult weekly crossword for use in a competition. “Times style” was the term used – the idea was to create a real challenge and keep a lid on the number of competition entries. I used to get regular updates from my contact at the AFR about weekly entry numbers, and these were relatively consistent and quite low. I wondered how many people ALMOST finished the crossword, maybe being one or two answers short, and therefore didn’t send in an entry. As an experiment, I intentionally eased up for a couple of weeks to see what would happen. The result was amazing: a tenfold increase in the number of entries, showing clearly that people were solving the puzzle but normally found it too hard to complete.
Occasionally I list some stats from my website showing, for instance, number of unique visitors and how many are doing the Stickler Weekly on average. The actual stats provide daily totals and accumulate across the month, which gives me some insight into the habits of solvers as a group. I’ve noticed over the last three months that there are less solvers doing the puzzle in the week it’s released, and access numbers for previous weeks slow down but still grow more than they used to. This indicates to me that the puzzles are too hard for most, and people are either taking longer than a week to solve them, or prefer to have the solution at hand as they solve. This would certainly make sense for U3A groups who only meet once a week where the course leader would like to have the solution available before class.
I’ve decided to “ease up” for a while to see if the puzzle stats shift, as I prefer to have a puzzle that most can finish in a week than service an elite group of solvers. As usual, I welcome feedback on this as real people will tell me much more than a few numbers on a report.

Eagle-eyed Steve noticed that I annotated a solution from last week’s puzzle differently from how it’s usually done, and flagged it as a possible mistake. Annotating crossword solutions has no set formula – maybe there should be one. In fact, a number of years ago I asked the contributors to the rec.puzzles.crosswords newsgroup if there was a standard and if not, whether someone should establish one. The response was lukewarm, and most believed there wasn’t a real need. There are some interesting styles out there including the colourful Guardian one used by Alan Connor, where different elements of clues are colour-coded, and I guess it doesn’t really matter as long as the message gets across. Interestingly, Alan’s method changed recently. Read his blog entry here.
Back to my “mistake”. I decided when I first started doing these solutions that I wanted to reflect the result of any device used in a clue in the text of the solution. That is, whatever is referred to in the clue would be targeted using any annotation. For example, in 1-across, “…to the point, mostly…” resolves to most of TERSE, and is shown as (TERS)E, with the brackets highlighting the result. In 19-across, “…mistake initially” is shown as (M)ISTAKE and in 7-down “…short character…” ends up being (LETTE)R. I’m not sure if I’ve always been consistent with this annotation, but the key is to recognise that the bracketed letters relate to a used device.

I still think there’s a need for a standard of sorts in this area – what do you think?