Thumbnails

Genius idea for a Sunday puzzle, a true WITT (Wish I'd Thought of That). The theme answers relate to American currency, so the one-dollar bill is represented by DENZEL WASHINGTON crossing 1 SEC, the ten-dollar bill by LINDA HAMILTON crossing HANGS 10, etc.

Incredibly difficult construction, perhaps one of the hardest I've run across in five years of analyzing the NYT xw. Not only does John incorporate 1.) a dollar sign in black squares (in the center of the grid), but 2.) a two-part revealer and 3.) SEVEN crossing theme answer pairs. The fact that he uses several three-letter answers inside the dollar sign necessitates some wide-open sections in all regions of the grid (due to the 140 maximum word count restriction). Amazing that John was able to fill it at all!

As a solver, I was frustrated to not accurately finish the east section, not having heard of HAIRCUT 100 (I searched for their "top hits" and hadn't heard either of those songs) and being only vaguely familiar with MHO (stupid MS in electromechanical engineering, did you teach me nothing?). Every puzzle must trade off thematic density and quality of fill, and I wonder what the result would have been if one of the bills, the 100, had been removed. That would still allow for thematic consistency (incorporating all American bills up to the 50) while giving more flexibility to fill those tough, wide-open areas.

Overall though, I loved the fresh concept and the fantastic "aha moment" it provided.

– puzzle by John Farmer

Excellent Monday puzzle from Jim, who amazingly has only been constructing for about a year. Incredible progress from when he started (we collaborated on a LAT puzzle). Solid theme, fill accessible to a wide range of solvers, interesting cluing. Monday puzzles are so difficult to make, because it's nearly impossible to satisfy the experienced solver (who's likely to be bored) as well as the novice (who could be stumped by anything remotely esoteric or crosswordese-flavored). A very nice balance today.

Note the NW and SE corners. A 5x5 section of white space is usually difficult to fill cleanly, and Jim does an admirable job. It's not ideal to have a partial at 1-across because it doesn't make a very good first impression, but it's certainly acceptable. Typically a maximum of two partials are allowed in a puzzle, because partials (as well as abbreviations, pluralized names, etc.) aren't usually as elegant as real words. Aside from that small hiccup, Jim's fill is clean, even incorporating a V in the SE without making it feel forced.

A comment about "cheater squares", black squares which do not affect a puzzle's word count. Will doesn't like cheaters since their inclusion is a mark of inelegance, and I generally agree. But I think a pair of cheaters in the very NW and SE (making APAIN into PAIN, ALIST into LIST, EDENS into EDEN and TRESS into TRES) would be a net gain. Your mileage may vary.

Finally, I didn't notice that OAKLAND AS and HAPPY DAYS are part of the theme until Jim sent me his write-up. I was all set to make this my POW! until I realized I had completely missed two theme answers (I had originally written that they provided very nice long fill). It's certainly a construction feat to interlock theme answers (HAPPY DAYS intersecting MAYONNAISE, e.g.) but I would have liked all the theme answers to "pop" more strongly, perhaps by starring their clues or by running them all horizontally. Even having five themers (all oriented horizontally) instead of six could be a good solution, since I liked this puzzle very well even when I thought it only had four themers.

– puzzle by Jim Peredo

A nice twist on the puzzle type where all the theme answers are definitions of a single word. Dan's additional element of using homonyms of the single word (WII, OUI, etc.) helps keep the idea fresh. And I laughed when reaching "Whee!" A major goal of crosswords is to entertain, and seeing that brought a smile to my face.

This theme type isn't seen much these days, since the answers tend to sound like made-up phrases. PERSONAL PRONOUN and CRY OF DELIGHT are both snappy answers, in-the-language, but FRENCH FOR YES will cause some solvers to grumble, saying that it's not as satisfying to solve such an answer. All puzzle themes must evolve or die (the simple "word that follows the first half of X, Y, Z-across" is largely dead, for example) so I think in the future, this theme will need to include all in-the-language phrases as theme answers or have some other advancement in order to be successful.

In construction, it's generally best to spread out theme answers as much as possible, since this provides for flexibility in filling. Typically rows 3 and 13 contain the first and last theme answers for this reason. Ten and 11-letter answers in these positions often make for an easier to fill grid, while eight and nine-letter answers add a layer of difficulty. Today's is prime example of this: note how the 9-letter theme answers in rows 3 and 13 create wide-open 5x5 white spaces in the NE and SW corners. Anything 5x5 or bigger can be very difficult to fill cleanly, and having something like STERS is not an optimal result. There are ways around this, breaking up CAME ONTO and OLD FILES for example, but that takes out some good long fill. A difficult trade-off.

– puzzle by Dan Schoenholz

Entertaining offering from Joel, putting a spin on the "two letters spoken aloud" type theme. It provided a nice aha moment after I finally stopped trying to shove ARTE JOHNSON into place and thought about it.

Six theme answers in the horizontal direction is a major challenge, and Joel executes it well. It's possible to put six themers into rows 3, 5, 7, 9, 11, and 13, but that often makes black square positioning very difficult. Joel's solution of overlapping themers at rows 3/4 and 12/13 works well, with only the ??NF?? crossing restricting flexibility. As a result, the fill in the north and south is smooth.

Sparkly fill is a mark of a great constructor, and as Will mentions, Joel excels in this area. As for the shorter, supporting fill, OKEMO isn't something I knew, or found terribly interesting after looking up, but what other options are there for a O???O pattern? Partials (ODE TO), prefixes (OSTEO), or even more esoteric places (ORONO) are worse. OUTDO perhaps, but alternating pattern of vowels and consonants typically makes for better flexibility in fill.

For a puzzle that's this thematically dense, there will almost always be one or two fill patterns that cause difficulty. Joel does very well to make them nearly invisible. A great puzzle is often defined not by what the solver sees, but what he/she doesn't see.

– puzzle by Joel Fagliano

★ Amazing work from Damon. This is only the sixth Shortz-era NYT "Schrödinger" puzzle; see the others here. These types, having two equally correct answers, are so tough to construct without making the clue too broad ("Five-letter word", e.g.), favor one answer over the other, or sound forced. We've had some good puzzles this week (and still more to come), but this one is fantastic.

From a construction standpoint, this grid may not look like much. Four pairs of intersecting theme answers and four 4-letter words, what's the big deal, right? Consider this: because eight of Damon's theme answers are short (three, four, or five letters), that means he had to incorporate longer fill in order to obey the 78 word limit. His grid has six long fill slots, and he populated them with OSCAR NODS, NIGHT OWL, TIGHT WAD, ART TATUM (one of my favorite pianists), STRATEGO, ODD OR EVEN, all snazzy entries. Six out of six! Talk about sticking the landing.

To be fair, as with any puzzle there are slight compromises. Some solvers might find the UNEVIE/ACELA crossing unfair, and DORR isn't as nice a theme answer as the much more natural PORT, CORP, and MORT. But these are minor issues, ones easily overshadowed by the puzzle's overall excellence.

Finally, David Steinberg uncovered perhaps the first Schrödinger puzzle in the NYT. Amazing what David has turned up in his pre-Shortzian efforts. Please save a job for me for when you take over the world, David. Er, Mr. Steinberg.

– puzzle by Damon Gulczynski

A sustained professional duo, working together to produce exponentially more than the sum of its parts, is a rare and beautiful thing. Kerry Walsh and Misty May-Treanor's non-verbal communication on the sand volleyball court. Jeff Meckstroth and Eric Rodwell trading killer leads at the bridge table. Danny Ocean and Rusty Ryan working a heist-of-the-century angle. All of them are on notice, because Wilberson is on the rise.

It's infrequent that two experts can work so well together, producing quality of this level over an extended stretch of time (this is their fifth NYT collaboration and they have more together in the LAT) without a clash of egos. The grid is well-constructed, with little dreck, and contains so many of the types of phrases that either shine or teach a solver something new. I had to wrestle with the unfamiliar TREEGUARDS but enjoyed reading up on something I walk past all the time (the iron fencing around a tree planted on a city sidewalk).

The SW corner was a little difficult for me due to TREEGUARDS and CARO, but ultimately fair. It's amazing that this is the only hiccup considering the difficulty of this particular construction. Often times the marquee answers are pushed to the perimeter, which helps to isolate them and make construction easier (generally more segmented = easier construction). Doug and Brad extend WARSAW PACT, SAGE GREEN, STEADICAM, RETIREMENT, WEEDEATER, and WORRISOME into the center of the grid, raising the degree of interconnect and thus the level of difficulty. Not all those answers are sparkly, but brilliant clues like "Avocado relative" (not the fruit but the color) add to the solving experience.

– puzzle by Brad Wilber and Doug Peterson

Julian ups the ante in this puzzle, not just using triple-stacked 11's, but adding an 8-letter word (19A, 44A) into those massive white spaces. Look at the amazing result in the top: not only are there three fantastic 11's, but even the crossing entries are interesting. It's difficult to get anything of note out of three and four-letter words, but UNIX, JUDO, and Riddick BOWE provide spice.

The lower stack also contains three great 11's, plus CRONYISM is the cherry on top. Some solvers will point out that YAGO is a heavy price to pay, but certainly all the crossings are fair and inferable. Overall, beautiful work from Julian.

Will brings up an interesting point re: either knowing or not knowing the BURJ KHALIFA. It serves as a perfect example of an entry that's fair even if it's unfamiliar, especially given that a Saturday puzzle is intended to be by far the hardest of the week. The world's tallest building, towering out of the Dubai desert, is something educated people ought to know; certainly a piece of knowledge a late-week solver ought to have in his/her repertoire. Useful to throw in at cocktail parties too. Jill and I were there in Feb and found it stunning, especially as we looked at the pictures of what was there 30 years ago (hundreds of square miles of sand).

– puzzle by Julian Lim

Fantastic to see a brother and sister team up! I'm trying to get my nephew (age nine) to co-construct with me, but he stubbornly and heinously insists that LEGOs are more fun than crosswords. I will break him sooner or later.

Five grid-spanners (21-letter lengths) plus two 18's make for a tough construction. All those long across entries necessitate a spate of down entries which must cross two or more theme entries. Pete and Sue use a quasi-stairstep pattern (note how most of the black squares flow NW to SE) which is frequently employed to reduce these constraints and segment the grid into sections. Working with smaller subsections is much easier than trying to fill an entire grid as a whole.

Areas that are more cordoned from the rest of the grid tend to be easiest to fill. Let's look at the ROLEX/BRIM section at the west. There are five parallel constraints (five down words that must cross both of the theme answers), but the fact that this area doesn't have to connect to anything to the left makes it easier. Pete and Sue did an admirable job with a tough set of constraints, only OONA being slightly esoteric (I'm still not sure what her "Game of Thrones" role is, but then again, I still get Renley Baratheon confused with a dozen other people).

Now consider the adjacent POOH/SOMA section. This part must connect to the left (into the ROLEX/BRIM area), but also to the SE (the BITO/TCBY area), which already has its own constraints. The result isn't as nice, with a IMRE crossing SOMA. That's two subsections covered, and the constructors have roughly a dozen more to boot, each with its set of constraints. Is it any wonder that Sundays are so much more difficult to construct than weekday puzzles?

– puzzle by Pete Muller and Sue Keefer

★ We have some very good puzzles coming up this week but I give the POW! to Gary for his strong early-week work. There's at least one other puzzle which I thought deserved a POW!, but to make a Monday puzzle stand out deserves recognition.

Monday puzzles are often derided by expert solvers as boring or tedious, so having a theme which is different (or does something to entertain) is important in order to satisfy a large range of solvers. In today's puzzle, I had to look back at the theme answers when I was done to figure out how they all tied together, and the fact that Gary hid the theme words in plain view (see highlighted words, which describe a touchdown play) brought a smile to my face.

Additionally, Monday puzzles are perhaps the most difficult to make, because the constructor cannot rely on using ORTs or spreading OLEO in order to complete a tough section. A perfect Monday puzzle should be super smooth, enough for the novice solver to not get too discouraged (by thinking they must learn a foreign language including ORTS and OLEO) and also have a reasonable chance of finishing. Will and I may disagree on the last point, but I maintain that creating early-week puzzles which new solvers can achieve the "I finished the NYT crossword!" high is important for the future of crosswords.

So let's look at Gary's fill. Not only does he keep the crosswordese to a minimum (AMAH being the misdemeanor), but he gives us a ton of 7-letter fill including RAKES IN, DIETERS, and HELLUVA. I'm with Will on ROULADE, it's something interesting to learn (and tasty!). AGORA I'm less positive on, but since it's an important feature of Greek history and we still see its influence in the word AGORAPHOBIA, it's legit.

Gary obviously put a lot of time and care into this puzzle. A great start to the week.

– puzzle by Gary Cee

Fun change of pace from C.C. today. She has an interesting story, growing up in China and only starting crosswords back in 2008. Her blog, L.A. Times Crossword Corner, covers the LA Times puzzle on a daily basis. I started doing crosswords back in 2008 as well, but can't imagine doing them in my non-primary language. Very impressive!

Anytime a theme uses shorter theme entries (seven letters or less), it increases the difficulty level for the construction. This is because the word count maximum of 78 stays the same, so the rest of the surrounding fill must be longer than usual. Note the stacks of 7s in the four corners, the 7s in the middle, and ARTICHOKE in the middle. It makes for a quasi-themeless type of filling challenge, incorporating many more seven letter words than usual. On the whole the fill is pretty good, although DCCC and QID show how difficult it can be to get those wide-open corners to work.

Finally, it would have been nice for the theme answers to stand out a bit more, perhaps by isolating them into five rows? That would have made for a extremely challenging construction (and may have called for different themers), but would have made it easier to see afterward which were the words of Chinese origin. As it is, SILK and CHOW blend into the puzzle a little too well (I've highlighted them in the grid so they stand out better). Typically Will likes the theme answers to "pop" and for good reason, but sometimes it just isn't possible. Today's puzzle reasonably trades off "theme pop" for the fact that I found it interesting to learn where these common words came from. Nice work from C.C.

– puzzle by Zhouqin Burnikel

I had to stop myself from squealing when I first met superstar constructor Patrick Blindauer at the American Crossword Puzzle Tournament two years ago. We had a funny exchange of glances when his Puzzle 5 (usually the hardest of the tourney) was announced and people groaned, anticipating the combination of pleasure and pain. He's known for his devilishly hard masterpieces.

Patrick brings us a word ladder puzzle with a twist. Nice to see the progression go from a logical start to conclusion, WARM BLOODED to COLD HEARTED. Also nice that he found five snazzy phrases in which to incorporate the ladder. Slight deduction for CARD being a part of another word when WARM, WARD, CORD, and COLD are separate, because it's more elegant if all the theme entries work exactly the same. Consistency is one of the criteria many people use in judging a set of theme answers.

Look at the very high theme density: 11/13/15/13/11. It's hard enough to incorporate five shorter theme answers, and Patrick raises the difficulty level by choosing long entries. Note how he chose a 15-letter entry across the middle. Although it might seem like a central 9 or 11 would be easier, that would actually make the construction more difficult because the center row would need to have black squares at its ends. That arrangement breaks up the grid in a way which is hard to deal with in terms of overall black square placement. Try it if you're curious.

The high theme density requires there to be many vertical answers which cross two themers. The result is that virtually every subsection is constrained multiple times, making the filling process challenging. Most everywhere the fill is pretty good, although Robb NEN feels like an ORT (leftover bit). All in all a nice job considering the tough constraints.

– puzzle by Patrick Blindauer

I laughed when I hit ITS SUPERMAN. Rare that I get a chuckle out of a Thursday crossword, so kudos to Ian for the clever idea.

Not only does Ian squeeze in six theme entries, a feat that should carry a "Kids, do not try this at home!" warning, but look at the parallel 9's at 2D/3D and 33D/34D. Ian could have easily put black squares in the middle of 2D and 34D (to increase his word count to the max of 78) and had a much easier time filling the grid, but he chooses to treat us with additional long fill. It's typically difficult to incorporate two adjacent long fillers because there's already the fact that they cross a fixed theme entry (ABOUT THAT and JUMP SUITS crossing ITS A BIRD), but Ian executes it well. So many levels of difficulty in this construction and lots of payoff. I had a hard time with UDALL but for a Thursday it's fair.

Almost every crossword contains trade-offs requiring value judgments. I love that Ian recognizes (and agonizes over) the fact that JEFF KENT will be somewhat controversial. Self-awareness is a mark of a great constructor. I tried to think of what other entries might have worked better, but as Ian already pointed out, only KENT STATE came to mind (and CLARK KENT of course). Due to crossword symmetry requirements, KENT STATE (9 letters) would have to be paired with ITS A PLANE (9 letters), which is possible but would throw the order of themers out of whack. So I think using JEFF KENT is a reasonable trade-off to make this concept work.

Someone invent a new phone and call it the KALEL PDA, will you?

– puzzle by Ian Livengood

I e-mailed with Patrick this week and wanted to share words from the master: "I find it easier to comment on the construction of themed puzzles, since those tend to stick in my mind a little better. Themeless construction is almost like meditation — you go into a trance, then come out again when the grid's full."

I'll have what he's having.

A 66-word themeless may not seem all that different from a 70-worder, but there's an inverse exponential relationship at work. Consider the gigantic open space in the middle of the puzzle. Typically a partial, pluralized name, or other bit of crosswordese is required to make such a section work, but Patrick laughs at that notion. Every one of the entries is lively, and there's not a stinker amongst them. What a feat to stack five long phrases without needing a single compromise in the downs.

As with every PB construction, I studied this puzzle for a few hours, trying to unravel the mystery of what makes his themeless skills so strong. One thing I've picked out is that Patrick tends to prioritize in-the-language phrases which incorporate fewer "Scrabbly" letters (JQXZ) and use more of the "Wheel of Fortune" free letters (RSTLNE). This is in no way a knock on him — it's an observation of his greatness. Figuring out a way to make construction smoother while still maintaining liveliness in longer entries is one of the factors that distinguishes PB as the cream of the crop.

Fifteen minutes of bliss. A close second for this week's POW!

– puzzle by Patrick Berry

Joe DiPietro is one of the NYT's most prolific crossword constructors, only 10 people having more Sunday-size puzzles than him (in the Shortz era). Amazing to see his creative output, including such mind-bending beauties as I SURRENDER! and BYWORDS.

Today he brings us a more standard type of theme, wordplay changing HER sounds to the actual word HER. Note that he uses nine theme entries, higher than typical (seven or eight). He also interlocks some of the themers (CHECK HER BOARD and SHUT HER BUG; COUNT HER PARTS and LET HER BOMB), always an impressive feat. Adds a touch of elegance when a constructor is able to do that.

Thank you to Joe and Will for the way they clued LICK HER BOTTLE. That could have gone in a very different direction. Ahem.

Sundays are so difficult to fill because the relatively low word count maximum requires at least a few wide-open spaces. This means that a constructor usually can't become good at Sunday-sized puzzles until they gain proficiency at themelesses, and indeed, these are two of Joe's fortes. Today's NW and SE corners both have triple-stacked 7's intersecting sets of long answers, making them quite difficult to fill. Each corner contains a lesser-known proper name (OLEANNA and LOESSER) but Joe takes care to make each of the crossing answers fair. RADIX might be tough for some, but it's a valid mathematical term.

– puzzle by Joe DiPietro

Nice start to the week; Ed gives us a lively set of theme answers that haven't been seen much before. It took me a while to remember the ad slogan as "M'm M'm Good", not "Mmmm Good". Funny how much difference a single space makes.

I tried to think of potential theme answers that had two sets of double-m's, and only came up with a few. I find JIMMY KIMMEL hilarious, but including him in the same thought as MM MM GOOD...I'm glad Ed didn't go down this route. Mr. Kimmel, if you're reading, I'll take my 15% cut in unmarked bills for the skit idea.

The theme density today is high, with five themers including a 10/15/10 in the middle. Ed does a nice job of spacing them out to give maximum flexibility, but there are so many crossings between MUMMYS TOMB, MILLION MOM MARCH, and MAMMY YOKUM that the center section becomes tough to fill. O??U doesn't have many options (OAHU only), and surprisingly, neither does M??M (MAAM, MAIM, MARM). That duo constrains the fill heavily, causing the ?AA? pattern at 45A (BAAL). So many constraints in that region.

Finally, a discussion about the difficulty of using eight-letter revealers. Because a revealer is often best placed at the bottom of a grid (so it wraps up the theme at the very end), notice how by necessity the SW and NE corners are six white spaces across. Bigger sections are harder to fill, and a 6x5 swath of wide-open space can be especially challenging. Ed does a nice job in the SW with the beautiful LET ME SEE and just one partial (A TIE) as a blemish, but the NE suffers a little with the "roll-your-own" RETAB next to the poetic ENORM.

Constructing for a Monday is very hard. Most of the aforementioned entries would be totally acceptable on a Thursday, especially if necessary to carry off an ambitious theme, but for a Monday, they offer a very tough challenge for beginning solvers.

– puzzle by Ed Sessa

Pete is one of the most published constructors in the Shortz era, and has the distinction of being one of fewer than 20 people who have doubly "hit for the cycle" (having published a puzzle for every day of the week not just once, but twice). It takes a wide range of constructing skills to achieve this feat — he's in rarefied air.

Straightforward theme, the letters DATE being in found in sequence within the four movies. Consistency is good, with each of the movies being well-known and all of them having the D-A-T-E in sequential order. It would have been really neat if the word DATE (without breaks) had been hidden in the theme movies, but that's likely too much to ask for. After 30 minutes all I could come up with was THE MANCHURIAN CANDIDATE, which is way too long. Even more awesome would be some visual of a nervous young TEEN trying to YAWN his ARM around his DATE. Definitely too much to ask for.

Note on PEWIT, which I have a feeling will draw complaints from some solvers. At first I had a negative reaction to it, but I then found it to be an interesting answer to look up. I think I would have seen PEWIT as an asset to the grid if it had been clued something like "Bird whose name imitates its cry". Funny sound, indeed. What can I say, I'm easily amused.

Nice NE and SW corners, containing such goodies as PAVLOVA and MOLIERE and NINE IRON. The NE suffers a bit from the crosswordese AWN, but all the crossings are fair. The SW is really good; a clean and fun triple-stack of 7's including the topical LEONARD. On that note, I enjoy it when a constructor personalizes his/her grid, and Pete's comment on LEONARD is a highlight for me. I like how Ben Tausig sends blurbs about his American Values Crossword constructors — wouldn't it be cool if the NYT were to even occasionally publish these little stories alongside the grid?

– puzzle by Peter A. Collins

Nice change of pace to see a rebus puzzle on a Wed. Paul did something unusual, using the rebus to incorporate not just one but two entries which are longer than the usual 15-letter weekday limit (SHORT ATTENTION SPAN and ONCE BITTEN TWICE SHY). Neat how constructors are continually thinking of ways to bend the rules.

I also appreciate how he eschewed entries like PUP TENT, ones where TENT is a separate word. Consistency results in elegance, and the fact that Paul either breaks TENT across two words (guTEN Tag) or hides it in in one word (peniTENT) is pretty good. Having strictly one or the other would have been fantastic, but that's a lot to ask.

One aspect that threw me off was the fact that there were several answers longer than the rebus entries: VENGEANCE, TAKE A HINT, ABOLISHES, and TEMPLETON. All nice to excellent answers, but typically the rebus entries are the longest ones of the grid so that they stand out. The other side of the coin is that throwing off the solver can be seen as a positive, in that the puzzle gains an additional level of challenge.

The grid construction is a toughie, especially given those big corners in the NW and SE. Typically these L-shaped corners are only seen in themeless puzzles because of the difficulty in filling them. Throw in the fact that each corner is further constrained by intersecting theme entries and the result is a real challenge. This arrangement of black squares does make filling the NE and SW corners easier, but a price is paid, in that the NW and SE corners are not quite as clean as I like to see.

How cool would it have been if the five tents actually formed the shape of a tent, in the same style as this puzzle by Patrick Merrell? It's awfully hard to draw a tent shape with five squares though, so perhaps it would come out looking too much like an upside-down V or a teepee. Especially for me and my 1st-grade level drawing skills.

– puzzle by Paul Hunsberger

A nice twist on the "add-a-letter" type theme with a cool revealer, C AND Y COATED, meaning the theme answers have C and Y on their ends. It's difficult to come up with a new spin on a traditional crossword themes, so kudos to Michael.

Check out the NW and SE corners, big white L-shapes often difficult to fill, similar to yesterday's grid. The big difference is that yesterday's had theme answers running both horizontally and vertically into those L-corners, while today's only has horizontal themers intersecting them. This freedom allows Michael to toss in such goodies as TWEEDLE, CASABA, and LACEUPS. Sure, OGDENS and CII are blemishes, and the MONTELL/ARNE crossing is going to cause some consternation, but almost every tough-to-fill space is going to have some trade-offs. The clue for OGDENS is nearly crazy enough to redeem a strange plural. Ogdens' Nut Gone Flake? Really? Okay, I'm smiling now at its sheer insanity.

Will's point about the themers is well taken. It's elegant if the themers are consistent, either all being natural or all being wacky. CART FAIRY and COLD MASTERY have that smile-inducing je ne sais quoi, but CHOSE DOWNY is much more straightforward. Luckily, the base phrase, HOSE DOWN, evoked images in my head of Animal House-like shenanigans.

Finally, nice job on the sheer quantity of snazzy fill. Incorporating ALLOW ME, MARTYRS, APOGEES, LACOSTE, EAT CROW, SKI CAP, EROTICA, and BACK END without major compromises enhances the solving experience.

– puzzle by Michael Blake

Debut for Mangesh! If you haven't read his 2012 ACPT letter you're missing out. Great story about the power of crosswords not only to delight but to educate. I had the pleasure to meet him at that ACPT and enjoyed swapping stories.

And how can you go wrong working with the venerable Doug Peterson, an expert constructor and a hilarious guy? I'm envious of the fun times they must have had. Doug and Mangesh incorporate double-stacked 15's, using a black square to break up rows 1 and 15. That pair of squares may seem insignificant, but this arrangement is so much more flexible to fill than a triple-stack of 15's. The top stack is very nice, both phrases being in the language and catchy. There are a couple of blemishes in OBS, EDD, and TER but overall it's pretty good. PENNON was a bit of a head-scratcher, but I enjoyed looking that up and reminiscing about The Tudors and Game of Thrones.

One issue with stacks of 15's is that they don't leave much room for other feature entries in the grid. Note how aside from the 15's, there are slots for only six 8's. It's relatively hard to find fresh 7-letter (or less) answers, so there's not a lot of room aside from the 15's to incorporate more marquee answers. TO DIE FOR is, well, to die for, but there aren't many other flashy entries.

To Will's point about reuse of 15's in stacks, some of us constructors have a running joke about SCARLET TANAGERS and A LOT ON ONES PLATE, which have been used many times in themeless stacks. Anytime a crossword has a 15-letter answer with any sort of bird clue, I automatically fill in SCARLET TANAGERS. And if I don't want to do something, I'll say "Sorry, I have a lot on one's plate."

Finally, a highlight of the puzzle is the clue for NINJAS. What a fun piece of information! Watch as I activate my ninja powers and turn invisib

– puzzle by Mangesh Ghogre and Doug Peterson

In addition to the usual criteria for judging a themeless crossword (lively entries, lack of crosswordese/partials/abbrs., fun cluing), Will hits upon an topic we don't often have a chance to discuss: grid aesthetics. By itself, I found the grid stunning, not just because I shivered at how difficult it would be to fill cleanly, but for its sheer visual artistry. Jim put together a gallery of Grid Art that caught his eye, and this one's also a beauty.

Tim not just flirts with the lowest word count record but does so with a relatively clean grid. And compare it to the other 11 58-word puzzles, noting how much less segmentation it has. Low word-count puzzles often depend upon splitting the grid into four quadrants so each can be worked on individually, but this one has a massive middle white space with high connectivity to four swaths in the corners. Daunting.

Usually the knock on low-word count puzzles is that they're inelegant and/or no fun because they depend on too many "roll-your-own" words (carrying an artificial RE- prefix or -ER/-ERS suffix). There are some of these here (I see you, RESEES) but surprisingly few. I'm mixed on TOOLER — in my first career as a mechanical engineer, I worked closely with moldmakers who we occasionally called "toolers", but that seems pretty esoteric.

Finally, check out the good stuff Tim throws in: SOLO HOMER, LETTERMEN, SMART ALEC, KNEE PATCH, all intersecting in the center. To those who might argue that this doesn't have any super-snazzy answers (or call it a "stunt puzzle"), at the very least this puzzle is a great change of pace to the usual 68 to 72-word themeless puzzles. A close second for the POW!

– puzzle by Tim Croce

★ JEFF: Hello from down the road (Mike and I both live in Seattle)! Brilliant meta with a great theme reveal, F-E-E-L T-H-E L-O-V-E formed in Braille rectangles, using O's as the Braille dots. How did you come up with the idea?

MIKE: Pretty much the same way I come up with every idea: in a highly disjointed manner. I was just thinking about how puzzle hunt people are probably the second largest group of consumers of Braille other than people with real sight issues and those who cater to them. Then I thought that a puzzle could use O's to represent Braille dots in a boxed array, and then it was one more leap to "O = love" (as in tennis).

JEFF: It must have been a bear to construct, given that there are no other O's in the puzzle. Elegant. What was the toughest section to fill, given that heavy constraint?

MIKE: So, you might think it was the lower left which had TOUCH TONE PHONE and CONTACT POISONS, which are loaded with O's in specific spots. But that actually just meant those were ordinary theme entries. The upper right was a monster to construct. The two theme entries had no O's, but were stacked with non-theme entries with O's in very specific spots. That section got rewritten multiple times. I had other entries in PRESS SECRETARY's spot: MASCARA BRUSHES (wrong format), FINGER PAINTING (not the best verb meaning "touch"). Nothing really worked till I came up with PRESS SECRETARY. Then it all eventually came together. Eventually.

JEFF: Funny, I was just about to say that TOUCH TONE PHONE must have been tough to incorporate, given that the O's had to be locked into certain places. For me, that was the best theme answer; perfect for the meta. Did it take a lot of brainstorming to come up with FEEL THE LOVE as the meta-answer? What were some of the other metas you considered?

MIKE: I can't find a version of the puzzle without FEEL THE LOVE. I did spend a long time wondering what the puzzle would say, but I didn't start laying down any tracks until I knew how the train worked. One other thing though: I have an entire version of this puzzle with no theme entries whatsoever, just the Braille boxes and the central instruction. I think there are people out there who would prefer the themeless version, since they don't appreciate a grid that's highly constrained. I didn't like it, though. You need a reason for all that activity, in my opinion, and the theme entries provided it.

JEFF: Totally agreed, especially for the NYT audience. It would have been cool without the theme answers, but things like PRESS SECRETARY and TOUCH TONE PHONE add another layer of elegance. That raises a question I wondered about while solving — at Lone Shark Games, you're typically creating puzzle hunts for an uber-puzzle-geek, someone who's entrenched in puzzles and metas. Did you want to make this contest harder? Was there some back and forth with Will on the question of difficulty?

MIKE: Right, I'm associated with super-hard stuff as a rule. I guess that informed why I've not had standard crosswords in the Times before. I'm always trying to make things with lots of layers. The Maze of Games is really just a giant interconnected web; the Puzzlecraft book is an attempt to tell a narrative about puzzle construction. So of course I came at this with an "all the things!" approach. Will pretty much let me roam unchecked, which made some people very happy and some people not so happy. I guess my primary goal was to make Will happy, which he was.

JEFF: Would you have preferred to make it harder? Did you feel like you had to keep yourself in check for the audience? Or is this about the level of difficulty that you typically put into a metapuzzle? And on a separate note, can you give us an update on The Maze of Games (drooling)?

MIKE: Nah, it definitely didn't want to be harder. I mean, there was a version with *no* boxes. But that was inscrutable. This hit what I wanted. Now, please stop drooling on my internet. The Maze of Games is almost done in layout; I'm down in LA right now going through approvals. It looks just gorgeous; Pete Venters and Elisa Teague did some stunning work. We are about to try to beat every blemish out of the main section of the book in a major round of playtesting. Then we will turn our attention to gussying up the Conundrucopia section, which has some major luminaries like yourself in it.

JEFF: (wiping drool) Lucky for us, there are six other Intertubes we can use to communicate. Anything else you'd like solvers or constructors to know about this puzzle? And what's next for Mike Selinker? Will we see more NYT xws or is this a one-time thing?

MIKE: I have to thank Will and Deb Amlen and the crew for being willing to take a chance on something this new and bizarre. They were braced for a hurricane of feedback based on this puzzle, and they definitely got it. But I think they thought it was worth it. I hope I can return the favor when Will and company see how awesome their puzzles look in The Maze of Games, alongside Patrick Berry, Scott Kim, and so many other puzzle heroes of mine. As for me, I'm going to continue to treat the New York Times as the greatest puzzle canvas in history. You don't sully it with lesser work. If and only if I have something that's worth painting on that canvas, I'll submit it. I can't guarantee it, but I hope that's someday soon.

JEFF: Amen to that, brother. When I first started constructing, I threw pretty much any crazy idea I had at Will, and he gave me great feedback (and very politely asked me to send only three at a time). Took me a while to realize that not just anything can make its way into the NYT. Hopefully other constructors can learn from your well-considered philosophy and only submit their very best stuff. Good luck to you and I'll look forward to receiving my copy of the Maze of Games!

MIKE: Thanks, Jeff. This has been a fun and thoughtful dialogue, and I hope people have enjoyed it. I'll go out on this email I just got from a fan named Lewis: "Letterboxes was the most moving puzzle I have ever done. By decoding the Braille, I felt like I had more empathy, just a bit, for what it must be like learning Braille and first using it; and by learning something about its history, I appreciated something of the care and even love that went into its development. This link between the process of experiencing a work of art and the message of the art itself is something one sees mainly in great poetry." I don't know about the poetry part, but I think the empathy part is bang-on. We are lucky people to be able to experience life to the fullest, to spend time imagining ourselves with the difficulties that others experience. If this puzzle had a little touch of that for anyone else, then I'm so glad I constructed it.

JEFF: I'll wrap up on my side by saying what a cool e-mail from Lewis. Neat how much puzzles can enhance people's lives! And that I might have a slight man crush on you. Not that there's anything wrong with that.

MIKE: (ahem)

(awkward silence)

– puzzle by Mike Selinker

WITT (Wish I'd Thought of That) concept for a Monday puzzle; a "how are these disparate things linked" theme. Especially clever that three of the four theme answers are colloquial: roll your eyes, roll your r's, roll the dice. Balls are rolled, yes, but that one isn't quite as nice as the others. I wanted to give this one the POW! based on the idea alone, and might have done so if the fourth entry was something like TAX CREDITS.

A discussion on non-theme fill. It's usually a giant asset for a puzzle to have long fill, assuming it doesn't force ugliness. Snazzy 7+ letter entries can really spice up a puzzle, turning a good one into a great one. Susan has some really nice stuff today, tossing in 9-letter MENS ROOMS, SORCERERS, and VICE VERSA (in addition to five theme answers!). MENS ROOMS might not pass the breakfast test for some, but I think it's a legit answer. I choose to visualize a fancy one where I awkwardly try to figure out if I'm supposed to tip the guy who hands me a towel I didn't want.

However, I probably won't be the only one to spend a minute trying to figure out how ROOMS and VERSA can be rolled, as per the theme. The revealer does specify which are the theme answers, but the clue is long enough that I didn't want to take the time to read the given numbers. I've highlighted the theme answers to make them stand out, but ideally I like theme answers to pop on layout alone. That's typically why most long fill (especially that of 9+ letters) is placed in the vertical direction, not the horizontal.

And in this case, an additional benefit of not including long across fill is that the first and last themers could have been placed in rows 3 and 13, spreading everything out. I'm not positive, but this most likely would have improved the THOS/SOG section.

But that's all nit-picking, my constructor's brain doing its usual thing. Very nice work overall; a rewarding change of pace to get a Monday puzzle with some cleverness.

ADDED NOTE: Susie wrote me after reading my notes to say that her original grid in fact contained TAX CREDITS! We shared a chuckle.

– puzzle by Susan Gelfand

Enjoyable to hear the interplay between constructor and editor. Nice puzzle today, made nicer with Will's touch of adding the circled E and T. If only the puzzle could add a sensory element to those circles, making them taste like REESES PIECES. Seriously, we have smartphones and life-saving medical devices but no one's invented "Solve N Taste" yet? Constructors: contact me for Solve N Taste® licensing rights.

An unusual layout today. I always appreciate seeing something different out of a construction, if nothing else to break convention and open one's mind to what else is possible. And this grid certainly allows Kevin to work in a lot of good stuff about a memorable movie. A drawback is that the short thematic material tends to get lost in the shuffle (I've highlighted all the theme material in the grid below to help it stand out). For some, that may not be an issue at all, but in my opinion it causes a slight loss of elegance (n.b.: some might see this layout as having increased elegance, due to the high degree of theme interlock). Additionally, the cross-referencing (OUTER/SPACE and REESES/PIECES) often causes complaints from solvers, forcing them to jump around as they solve.

Amazing that Kevin was able to jam in some long fill too, even given all the thematic constraints. DOORBELL and THE MRS are especially nice. EAR DOCTOR is a real profession (otologist), yes, but as a phrase, isn't not as snazzy as the other long fill. And given that the north and south sections suffer a bit with STER, AGIO, A GIRL, I might have preferred less long fill if it meant getting rid of some of that. But it's hard to say if breaking up some long answers would even be possible with this layout (without making the puzzle feel too segmented).

Overall, neat theme idea with a lot of good stuff packed in.

– puzzle by Kevin Christian

Nice example of the "both halves of the theme entries can follow a certain word" theme type. Judge Vic and Bonnie are seasoned constructors and pull off this ambitious grid. Not only do they utilize six long theme answers but they intersect them. And then to top it off, they leave wide-open corners in the NW and SE when they could have broken them up (going from 76 words to 78 words). Bravo!

This is a difficult genre in which to construct, because it's often tough to find snappy two-word theme answers where both halves fit the "word that can follow X word" criteria. HOT WATER (hotline, water line) is fantastic, as is HARD TIME and AIR SUPPLY — three beautiful entries I would gladly put in one of my own puzzles. DATE BREAD and BOTTOM LAND seem to be "things", all right (at least according to the Google), but neither feels like a great answer to me. BUS SERVICE is certainly familiar, but not nearly as snazzy as the first three.

I appreciate how the theme entries stand out just due to puzzle layout, especially relevant because I felt yesterday's puzzle suffered a bit from this issue. I've highlighted the theme entries as a point of reference but note how it's not really necessary, as there are no answers longer than the themers.

Nice clean construction with very few ugly entries; the mark of pros at work. I had to search to find answers that I didn't like, which is always a great sign. RECOOK feels like a "roll-your-own" and TABOOED...at first I groaned a little, but after verifying that it actually can be used as a verb, I think I like it. Maybe if enough of us use it, we can make it "a thing", like eating a candy bar with a knife and fork. Or has that been tabooed?

– puzzle by Victor Fleming and Bonnie L. Gentry

★ Clever WITT (Wish I'd Thought of That) idea and clean execution; a winner of a puzzle. CLOCK is incorporated in the center of the puzzle, and the clock number (in the proper position) needs to be added for the clue to make sense (DOZEN becomes TWELVE DOZEN, for example). Perhaps a touch on the easy side for a Thursday puzzle, but what an enjoyable five minutes of solving. Kudos to Tom and Victor.

The difficulty of the construction might not jump out at you because of the excellent execution, but this perimeter theme arrangement is a bear. Most recently, the legendary Liz Gorski did it on a Sunday puzzle and commented on the challenge. Such degree of interlock in the corners places high constraints on the grid, making each corner an individual nightmare to fill.

But Tom and Victor have done it well, even incorporating such great long stuff as SIAMESE CAT, SHIPSHAPE, DEATH STAR, and GO TO SLEEP. The SW corner is especially smooth, I appreciate how much care they've put into it. If AGRI is your only blip (and it's an awfully minor one) I call that a giant success.

To be sure, there are signs of the construction challenge in the AMIGA/GALOP area and the obsolete GMAC, but those are very small prices to pay. And I would bet Tom and Victor tried many other entries in place of I AM A CAMERA before settling on it. Note the alternating vowel-consonant-vowel-consonant (repeat) pattern, which often makes construction easier, especially when surrounding fill like SIDED exhibits the same pattern. I AM A CAMERA not a first-rate answer, but it does its job. Such is the difficulty in incorporating long fill with this sort of perimeter themed puzzle.

There are more nice puzzles coming up this week, but this innovative and beautifully executed xw gets my POW!

– puzzle by Tom Pepper and Victor Barocas

What a cool idea to have the first across answer = GERMINATE and the last one = TERMINATE! Themeless puzzles are usually, well, themeless (that sounded smarter in my head), so I love seeing a change-up, especially when it's a neat mini-theme.

The skeleton of the puzzle (TERMINATE/GERMINATE/CHANGE ONE LETTER) is quite clever but I agree with Will that those entries alone are a bit thin to hold their own as a full-blown theme (here's an artist's rendition of what the original skeleton might have looked like; note the shifted pair of black squares in the NE and SW corners). I liked this puzzle as a mini-themed themeless, especially since there's enough good fill so that the puzzle can stand as a 72-word themeless in its own right. The bottom TAG TEAM/AREA CODES/ARMPIT section is snazzy in particular, especially given the clever clues for those entries. Excellent work in the two 15's, both in clues and answers.

Interesting interplay between Will and Pete today. Will and/or Frank changed a small corner of one of my early puzzles and it was jarring at first to see an unexpected change in the grid. I wondered, why not kick it back to me and ask me to jazz it up? It took me a while to realize that Will carries an enormous workload and has a proportional quantity of associated communication. Selecting, cleaning up, and finalizing 365+ puzzles a year (plus variety puzzles) is a monumental task, and going back and forth with hundreds of constructors for small changes would be a bear (he does kick grids back when he likes the concept but there are major issues). It would be great for us constructors to have a say in every step of the process, but that can't always happen.

Frank, by the way, is one of the nicest crossword people you'll meet, and very unassuming. Plus, he did a nice job on my previously mentioned puzzle, adding some needed spice to my SW corner, inserting such goodness as DR EVIL and WOOKIEES.

– puzzle by Peter A. Collins

I love to climb. Every week I do at least two good hard sessions at my local climbing gym, usually topping out at 5.11b-rated routes. I tend to avoid 5.11c or higher because I'll usually flail, exhaust myself, and fail, but one of my climbing partners usually goads me into getting onto a 5.11c or even 5.11d. "It'll be good for you," he says with an evil smile. And he's completely right. I barely ever make it fully up, I usually come down with heart pounding as if I were dying, and I always learn something, coming away better for the experience.

Let's evaluate the grid from a constructing perspective. There are only a couple dozen themeless puzzles with a lower count than today's 64-worder, and Joe does something most of these other ultra-low word count puzzles do: keeps the grid wide-open (mouse over some of those other puzzles to see how segmented they usually are). As a constructor, it's a daunting task to face those two gigantic corners in the NW and SE. Even a 6x5 corner of white space will cause problems, so just imagine what a challenge an 8x8 region represents. Kudos to Joe for having the guts to try it.

For the most part, he manages to get in reasonable entries without a ton of junk. Nothing besides LISTENS TO REASON and ORNAMENTAL TREES are a "wow!" type entry, but most do a workman-like job. In the NW, subpar answers like POTTIES, AT US, and REMISE hold the region together, which is what those big sections often require. In the SE, THERMOSET plays that "glue" role. That BARTOLI/THERMOSET crossing was a total guess for me (I could have sworn I worked with THERMASET materials back when I was a mechanical engineer) but after some thought, that's more my fault than the puzzle's. A recent Wimbledon winner should be fair game, so I'll file MARIAN BARTALI away for the future.

Aargh, it's MARION BARTOLI, consarnit!

It takes a great deal of practice and perseverance to be able to accurately solve something like this. Puzzles like these are useful in that they force many of us to stretch beyond our abilities and learn something, coming away better for the experience. I have a great deal of respect for Joe as a constructor and appreciate today's 5.11d+ workout.

– puzzle by Joe Krozel