Thursday, January 7, 2010

NPR Puzzle 1/3/10 What's Correct, What's Not Correct, and Who's All-Correct

Here's this week's NPR Puzzle:

Write down the digits from 2 to 7, in order. Add two mathematical symbols to get an expression equaling 2010. What symbols are these?
And here are the two solutions we discussed on Sunday:


2345 × 6 ÷ 7 = 2010  (CORRECT)

2/3 × 45 × 67 = 2010  (INCORRECT because it uses three operators)

I do enjoy the maths puzzles.  I'm "allowed" to solve the four numbers-based puzzles published annually in The Listener, the puzzle Ross used to edit.  This is completely safe because it's relatively easy to double-check.  With the crosswords (the other 48 puzzles each year), Ross and I solve them together, but then he re-solves them as he fills in the grid to be mailed to the official checker.  (The answers aren't published for three weeks so that people have enough time to get their entries in.)  There's a ferocious honor code about not collaborating on the current puzzle because it really matters to people that the end-of-year statistics are accurate.

Here's what happens:  The "official" scorer keeps track of every single entry (up to 1,000 for a maths puzzle; in the mid-three figures for the crosswords), checks it for accuracy, notes all mistakes, and keeps some sort of master list.  All solvers for the year are then tabulated as to number of errors, with the "all-corrects" being ranked by the length of time they have been all-correct.

These days, that number is pretty large (in the 20s, last I checked) although one does move up the list as people make errors (or miss a puzzle) and thus drop off the list while new people are all-correct for the calendar year.  Back when Ross was all-correct for the longest time, it was much harder to do because there was less electronic help.  These days, with Chambers (the official dictionary used by setters and solvers of The Listener) available in a searchable format, it's easier to look up answers by searching for the definition.

And, as Dan commented on Sunday, Ross's own software makes quick work of a lot of cryptic wordplay.  So -- with all these resources, how is it anyone ever makes a mistake?  Well, these puzzles are hard.  How hard?  Here's Brendan Emmett Quigley, one of this country premier crossword compilers, discussing a recent Listener he encountered in the London Times:
The Listener crossword was set by Jago. It was a 12x variety that ran in the Times of London on December 26, and well, let's just say I couldn't get past the instructions. The clues, well, they relied so heavily on surface meaning (it was a story), that I couldn't make anything out. A small sampling of three consecutive clues clues: {Mum raised and fed pup on Xmas leftovers (5)}, {It's a sore point, considering what's in most food. (7)} {Both together for Xmas? (3)}. Yeah, right.
Yup, that's what we're talking about.  Thanks, Brendan for making the point better than I could.  (For the rest of BEQ's discussion of puzzles in the English papers, go here.)

There's also a quaint and charming tradition: if the same people are all-correct and in the same order as last year, the solver at the top of the list cedes the "Solver Silver Salver" (the prize for best solver) to the next solver on the list who hasn't yet been awarded the salver.  Ross figures we have a shot at the Solver Silver Salver in maybe 2012.  And I get to make the speech!  Which would be embarrassing as I -- just like Mr. Quigley -- would not be able to solve a Listener all by myself.  Except the maths puzzles.

Thank goodness for maths puzzles.

2 comments:

Tinyc Tim said...

I heard about this one first from my friend Howard, later was reminded of it by my friend Andie and two days ago was asked by my brother Chip if I had solved it. Each encounter generated brief attempts to solve it. I often find problems like this go thru stages before they get resolved. Yesterday I worked on it in earnest. I tend to go the computer program route on such things. Here's my solution:

http://primepuzzle.com/tc/npr.html
 

Magdalen said...

Interesting. Enjoyed seeing all the two-operator solutions!