New York Times, Monday, May 29, 2017

SETH: Jeff and I have collaborated since 2014, and it's great fun. For this one, we wrote around 150 emails, some of which read like an absurdist script...

Seth: PITINO and RAISA seem fair, but the AKIMBO, SENECA, GRIT, and MATING goodness of Version C gets my vote.... Maybe, for a themer... BEST DAY EVER?

Jeff: Hmm, maybe a reach. I'm not a SpongeBob watcher though.

Our original plan was to make the units of time narrow down to an AHA MOMENT. Couldn't quite make that work.

Hope your solving time isn't a dog's age on this … and thanks, Will and Joel!

JEFF: Ah, Seth, that sneaky devil. He keeps changing the order of our names just before submitting the puzzle to Will! He deserves first authorship on this one.

Seth is a brainstorming monster. I love getting his emails, usually loaded with 10+ ideas to consider. They don't always contain a usable seed idea, but more often than not, something sparks another thought, and it morphs into something fun.

For this one, it took us a while to zero in on the idea of PRIME TIME being "synonym for good" + "time duration," but I thought it would be great … if we could 1.) get all strong phrases, 2.) have crossword symmetry, and 3.) present them in a logical sequence.

Some constructors would still do the puzzle without the third constraint, but I felt like it'd be too inelegant for my taste without that. It meant we had to do a ton of research. Luckily, another of Seth's strengths is his tenacity in research. Soon, he had a gigantic list of possibilities for us to sift through.

RED LETTER DAY made the construction challenging though — not only did it require a 16-wide grid, but it sort of cut the puzzle in half. Took a lot of trial and error, testing and ditching layout after layout. Finally, I felt like the skeleton you see was promising — enough room for some long bonuses, while not looking problematic in any one area — and Seth took over from there.

We always have a lot of fun going back and forth, comparing possibilities for various regions (although I can't remember how SpongeBob was related now). In the SE for example, I badly wanted to work in DOUBLE O (as in double-O 7), but DOODLES made that area much smoother. Ah well, the solver comes first.

16-wide grids can be surprisingly difficult. Just one extra column creates such challenges. One of the bigger issues is that I hate going above 78 words (the usual max), even though one could argue that it ought to be 81 words for a wider-than-normal grid.

Why? As a solver, I'm used to 15x15, and if my solve feels slow, that makes me like the puzzle a little less. So as a constructor, I try to keep a 16x15 grid feeling like it's a 15x15. That usually means using longer words than normal on the whole, all the while keeping things smooth. Not easy.

In this grid, I felt like we needed at least two pairs of long bonus answers. It's easy to work in the first pair of long downs (BEDHEAD / DOODLES), but it gets exponentially harder after that. Putting in RUB NOSES and IN THE WAY meant using fairly big corners, piling on the challenge.

Finding more long slots was tricky — long acrosses like INFIDELS and MAIL IT IN are usually tricky to incorporate, given that it means stacking them with theme answers, but some testing made this arrangement feel like it'd be doable.

Finally, we debated long and hard over ETERNE — was it worth the price of admission for AMIRITE, RUB NOSES, MUFFLER? Ultimately, it seemed favorable, especially since we didn't have that much other crossword glue in the puzzle. The eternal constructor's dilemma of snazzy vs. smooth rarely has a clear answer. Hope you agree with our decision.