I’ve been working on this week’s Listener puzzle — a number puzzle this week, called “Number or Nummer” — for only about twelve minutes, and I’ve already narrowed down the possibilities to just 393,120.

Later: Is too much to ask in a number puzzle to have math that looks like math? With minus signs (HTML code: −) or at least en dashes (–) instead of hyphens, and exponents that are actually raised (<sup>2</sup>) and not just in a different font?

Plus: In the clue for 13 Across, which is in the form X − Y / Z, could you maybe have added parentheses so we could be 100% sure whether you mean X − (Y / Z) or (X − Y) / Z? I hope I find out I’m wrong, but I have a growing suspicion that you mean the latter.

Still later: The order of operations in 13 Across is correct.

I finished this week’s Listener puzzle, “Toga” by Nod, this morning. I really thought I was going to finish last night — around 9:00 I had only solved about half the clues, but that turned out to be enough for me to figure out the “advice” spelled out by the extra letters, how to apply it, and what to do with the result. By figuring out in advance what was supposed to be the final step in the puzzle, I obtained several letters in the grid that I hadn’t actually solved yet, and that quickly gave me a few more answers. I was quite sure that I’d have it all nailed before I went to bed. But I got sleepy and ended up crashing with one clue still unsolved, and two more for which I was pretty sure I had the right answer but hadn’t yet worked out how the wordplay was supposed to work.

I finally solved them this morning — I was right about the two answers and worked out the wordplay, and the third clue turned out to involve some proper nouns I didn’t know that aren’t in Chambers Dictionary; thank goodness for the Internet. All in all, an enjoyable and not overly brain-busting puzzle.

Just now finished this week’s Listener puzzle, “More Collusion” by The Magpie. I’m grumbling a bit because I think there’s some ambiguity about the extra letters to be deleted from some of the clues, and I had to backtrack from the finished grid to figure out the exact path I was probably supposed to have taken to get here. However uncertain I am about a couple things along the way, though, I don’t see any ambiguity about the finished grid, so I’ve got my entry ready, and a Friday night finish.

This weekend’s Listener puzzle, A Keep by Phi, took me till Saturday morning to finish. I found it a challenge the whole way — blank diagram, numerations omitted, misprints in some but not all of the definitions, good grief — but all in all a very satisfying puzzle. The completed grid is a handsome piece of construction.

The clues are mostly nicely done, though one of the definitions seems to me to be oblique to the point of not being a definition at all; the fact that, say, sentimentality can be a definition for cheese doesn’t mean that sentimentality can also be a definition for Camembert. But then, maybe I’m missing something about how the clue works.

The clues are in their normal order, from the first of the acrosses to the last of the downs. The introduction omits to say this, inadvertently I’m sure, but it’s crucial information for solving.

It’s 8:20 pm on Saturday and I just finished this weekend’s Listener puzzle, “Jumping to Conclusion” by Sabre, over dinner. It’s a stinker, and despite some nice clues and a surprise at the end, this may be my least favorite Listener puzzle in quite a while. A number of the words and parts of words have to be entered not in a straight line but in a series of chess knight’s moves, which opens up the number of possibilities so greatly that it’s kind of tedious to work out the right answer.

Also, there are ten clues that you need to solve with no help available from crossing letters at all, and an eleventh (34 Down) that you have to solve with the only possible help being the first letter (from solving 33 Across). These clues are entered entirely in knight’s moves, but you can’t work out more than a few of the knight’s moves until you have solved every one of those eleven clues. When I went to bed last night I had all but three clues solved, yet I still had 19 empty squares in the grid, 19 squares where there were too many possible paths for the words to take to be able to fix any more letters. I took another stab at the puzzle this afternoon and solved two more clues, but I still had 11 squares I couldn’t nail down. Eleven squares left indeterminate because of one unsolved nine-letter word! And what I had worked out didn’t give me so much as one single certain letter to help with solving the last clue! It looks as though, for the knight’s-move portion of the diagram, the constructor gave the absolute minimum number of clues you need to determine where all the letters go once you’ve solved all the clues. That seems kind of stingy, and it’s not hard to find several more words of five letters or longer that the constructor could also have given.

Even once I did have the last clue solved, working out the one possible way to enter the letters so that all the words could be made took me quite a while because of all the possible paths that had to be considered and narrowed down. There were a couple of nice surprises at the end as a reward for trudging through all of that, so I wouldn’t say it was a bad puzzle, but all in all not an interesting enough puzzle to reward the amount of slogging needed.

A Friday finish this week: I finished this week’s Listener puzzle, “The X-factor”, a little after 11 pm. It’s a numerical and an ingenious one with a lot of subtle deductions to be made along the way. Very tough at first — I think after my lunch hour I’d identified the value of one letter and filled one cell in the grid. But the pace picked up steadily as I chipped more and more of it away and I must have filled in the last quarter or so of the grid in under fifteen minutes.

10:53 pm and I just finished today’s Listener puzzle, “Cruciverbalism” by Poat. It looks incredibly tough at first, and I was wondering whether I’d be finishing this one by Monday at all. The puzzle is another diagramless sort, and I had to solve about half the clues cold, without any help from crossings, before I was able to start to fit even a few of the answers together; but the puzzle became quite a bit easier once I’d gotten the first few words into the grid, and then when I saw what the omitted letters were going to spell, the rest fell together pretty easily.

It’s about 3 pm on Saturday and I just finished this week’s Listener crossword, “All New”. A workout, but a really enjoyable one, and with some nice surprises cropping up at the end.