Minute A Day About Movies

MAD Puzzle No. 1
1/1/10

John’s First MAD Puz
Puzzle
Answer Grid

Comment
If you’re not a regular crossword person, this comment may seem rather inside baseball (not to mention, No. 1 would’ve been a tough puzzle to cut your teeth on).  Suffice to say, one desirable feature in crosswords is use of uncommon letters.  Think of Scrabble.  You get a higher score for QUIZ (22 points) than for TEST (4).  Scrabble scores are all that matters in Scrabble:  the highest score wins the game.  Scrabble scores for crosswords are relatively unimportant.  Other factors are better indicators of whether a puzzle is any good.  But Scrabble scores for crossword grids can be tracked.  So they are.

For more on Scrabble scores and the “sabermetrics” of crosswords, you can check out Jim Horne’s most excellent Xword Info site.  For New York Times puzzles during the “Will Shortz era,” the average Scrabble score is about 1.55 (that is, an answer letter is worth about 1.55 Scrabble points, on average).  Of the past 5,885 puzzles in the Times, exactly eleven have had a score of 2.00 or higher.

Before I get to my confession, let me say that  I never had a puzzle with a score near 2.00.  (My highest was 1.80, which isn’t bad.  It’s among the top hundred.)  Till Jim started tracking the stat, I hadn’t given much thought to “Scrabble scores” (though, like any constructor, I’d probably use a Q instead of an S if I had the choice).  For a time, puzzle stats were the kind of thing I’d look up after a puzzle was published in the Times, but recently Jim added a feature allowing constructors to easily run the numbers prior to publication.  That’s a dangerous tool for a bunch that shouldn’t be trusted!

With J’s and Z’s in five theme answers, this puzzle had potential for a better-than-average Scrabble score.  I was curious.  I uploaded the completed grid at Jim’s site and discovered I had crossed the 2.00 bar:  2.02, in fact.  Whaddya know! 

There was, however, a problem.  I had a word in the grid that was no good (SMEES, at 8-D).  There was no good way to justify leaving it in, so I reworked the NE corner to what it is now.  That change lowered the Scrabble score.  It went down to 1.99.  Ugh.  If the original score had been, say, 1.92, I wouldn’t have thought much about having a score above or below 2.00.  But after seeing that 2.02, I really wanted a 2.00 or better.  I was just a single Scrabble point away.  What could I do?

This:  I made a change to the SE corner, replacing ABIE and NYPD in the earlier grid with ABBY and NYSE for the final version.   The Downs went from BIP and BED to BBS and BYE.  More or less, an equal trade.  The new score:  2.01!

I made it.  So what, right?  I can’t say I feel great about it.  I am glad I didn’t end up with a just-miss, but the final Scrabble score in this case had virtually nothing to do with any solver’s enjoyment, so it’s pretty meaningless.  The changes I made didn’t do any damage (otherwise, I wouldn’t have made them), but the whole thing got me to thinking about the changing nature of crosswords.  There are some great tools available today that help constructors make better crosswords, including statistical analysis you can do now during the puzzle-making process.  That can be helpful–a 1.32 score may mean you need to spice up the fill, for instance.  But statistics often become not just a way of tracking performance but a goal unto themselves.

We’re a society obsessed with numbers.  Schools “teach to the test” and are focused disproportionately on student scores.  Students worry more about their G.P.A. than what they’ve learned.  Batters bunt for base hits on the last day of the season, doing anything they can to raise that .299 average.  Sometimes there’s more at stake than pride, and sometimes it’s understandable, but the numbers fixation does distort behavior and what we think is important.

Except for the chase for lowest number of black squares in a grid (18 is the current record), crosswords have been relatively free of the obsession with numbers.  But until recently, there weren’t many stats that were tracked.  Now there are a lot more (e.g., the Times record for most U’s in a Tuesday puzzle is 15, achieved twice).  Is that a good thing or not?  I think it’s great.  I think numbers are fun.  I get a kick out of looking at them.  And I try not to take them too seriously.

With more stats available now than ever before, I’d guess they’ll become more of a focus than in the past.  There may be new records people shoot for, which isn’t really a problem, to a point.  Numbers, though, whatever they’re worth, don’t make better crosswords.  For any good puzzle, the numbers are usually incidental, not the point, and shooting for records often means paying a price.  I’d even say some puzzles with high Scrabble scores–you know, the ones with lots of J’s and Z’s–seem to be trying too hard.

I know what you’re thinking:  I should talk.

What else is on your mind?  Any confessions of your own, crosswords or otherwise?