What I’ve seen of the reception of this one suggests some explanation is necessary.
The basic idea is sound enough – all around the concept of ‘filling cells’ and its connection with both completing a crossword and imprisonment. The first thing was to check for the existence of suitable prisons to be added to the grid. DARTMOOR splits nicely into word chunks, while SING SING has plenty of common strings of letters. Anything else familiar and of eight letters? Well, there’s ALCATRAZ but that wouldn’t…except isn’t there a word ATRAZINE? Looks like ALCATRAZ was named specifically to facilitate TALC/ATRAZINE.
So the plan was to have a carte blanche puzzle with blank cells in its initial appearance which are then filled with well-known collections of cells. It seemed a little too easy to have the grid supplied – it would take very little time to pin down the gaps and find the additions. I was also slightly wary of having the first word TINE (or indeed T____ (4)) when it would slip in to TALC’s slot so readily. It still does, of course, but the carte blanche presentation expands the problem and minimises this component.
Being a carte blanche, of course, means word lengths aren’t given. With the exception of the top, middle and bottom rows, doing so would have provided the grid pattern just as neatly as if we had put the bars in.
The preamble was something like the fourth attempt. (Preambles are sometimes harder to set than puzzles.) The key issue was how to indicate that certain clue answers started one cell away from the first square of the associated entry, being augmented by the added prisons. (Others, of course, finish one cell early, but that felt, from personal experience, less significant.
It’s always a good idea to have some concept of how solving might be approached. Here the idea was that the Y shared by TYPIFIED and YEAH would anchor the NW corner. Or perhaps the P of POLARIS – either the E or the O would then tell you something about EOLIAN. This does seem to have been the way many of the successful solvers got into the puzzle.
One change that might have helped more, in retrospect, would have been having ALCATRAZ and SING SING on the second and penultimate rows respectively. That way there would be a clear top row to provide greater anchorage.
There was concern about LENSINGS, which is overtly in Collins, but seems, in any case, a credible gerund formation. I can see a work-around – SAND BEDS, ?AND and LEAS – but the presumably more acceptable LEASINGS and LEAS seemed worryingly close.
Overall the puzzle seems to have been considered a good idea not sufficiently well executed, with the use of the top and bottom rows making it harder to get footholds in the grid. Even so, I don’t mind keeping you in suspense wondering whether the next puzzle is going to be a ripsnorter. (For what it’s worth, I think probably not.)
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.