David Phillips author page

I have a confession: constructing a 72-word themeless grid scares me. When I personally begin solving a themeless puzzle that goes up to the max word count, there's a tiny, overly entitled person in my head that mutters under his breath, "This puzzle better be fresh, lively, and contain zero crosswordese or I'ma gonna throw my hat in-a disgust." (For the purposes of this blurb, I wanted Mario to serve as my privileged persona, because Mario's funny when he's haughty.)

If, as a constructor, I went by Mario's judgment, I'd say this puzzle doesn't quite hit the mark due to a lack of super-fresh entries (perhaps with HAWKMAN and APP STORE serving as exceptions) and the existence of crosswordese/undesirable entries (including, but not limited to, URE, ESTE, CRU, and SRA).

If I were to take a more holistic approach and tell Mario to find his princess in another castle, I'd say the puzzle does just fine. The puzzle doesn't seem to contain a preponderance of entries in any single subject; plus, many of the clues/entries seem to hit my ideal combination of trivia, wordplay, and fun--with some thought-provoking material thrown in for good measure (see 36D; I'm glad Will/Joel/the editing team kept this clue).

Overall, I kind of like this little guy (the puzzle, not snobbish Mario) but would listen to the other little guy a bit more if I were to tackle another 72-word themeless for the Times.

With this themeless, I will admit that overambition might have gotten the better of me—but not by much IMO!

In an effort to go big, I set out to fill a dreaded quad stack and its symmetrical neighbor. If I remember correctly (it's been two years since construction, so give my cortex a break), I think I chose to put these quad stacks in the NE/SW, because, in my constructing experience, I've had the most difficulty filling wide open areas in those sections. In my mind, if I were to force many of the constraints to those sections and come out alive without too many blah entries, I might give myself a fighting chance to cross the finish line and perhaps add some nice entries in the less constrained NW/SE sections.

The central stairstep arrangement of black squares—going diagonally up from left to right—also helped by producing what I call a double coincidence of wants. In this case, those black squares make it so that, for example, the end letter of 6D is also an end letter for 32A, the first letter 37D is also a first letter for 37A, etc.

As for actually filling the grid, I started in the NE by placing I THINK I CAN in the second row, usually the most constrained row in a top stack. The result was bupkis. I then tried it in the third row and got a mostly good fill. EARNS A NAME sounded awkward when compared to the more common "makes a name (for oneself)," but I compromised for the stack. As you can tell from MADERA, CO-SET, and SAES, the other quad stack proved more difficult to fill cleanly. As I predicted before filling, the NW/SE wasn't overly difficult to fill.

At first, the clues looked unfamiliar. "Did Will and the team really rewrite that much?" I thought. In actuality, most of the clues were in my original submission ... except that a lot of the clues in the print version were my second or third picks. "Did I really write a fragrance clue for ARAMIS?! [looks at original submission] Huh! ... apparently, I did."

Overall, I'd rate this puzzle a B+. The grid design proved a bit too ambitious for my skill level at the time. However, I do think I'm getting very close to hitting my personal sweet spot.

This is an old-ish one...I want to say it's been between one and two years since acceptance.

At any rate, this was an experiment in big, bulky corners. As with my last themeless puzzle, I think many of the entries came out as more neutral than exciting. My whopping 2+ years of "constructing experience" tells me that the 7's combined with the open space (read as: lack of entry choice flexibility) are mostly to blame. If I had started with some fun colloquial phrases for the 15's instead of you-know-it-or-you-don't proper nouns, maybe they could have added some zest. (Fifteenth identical) note to self: use less proper nouns and more hip, in-the-language words or phrases. Bonus points for portmanteau words!

Criticisms aside, I did like a lot of my and Will's/Joel's clues on this one. My favorite Dave originals are 50A and 52A (What?! Adam Smith had a sardonic wit? Who knew?). My favorite moderately edited clue is 22A (original clue = [Like tritium]; the alliteration seems to make it pop to my ear). However, [Web feed?], a Will/Joel original, wins best clue of the puzzle, hands down. Honorable mentions go to 19D, 28D, and 35D.

From a constructing standpoint, I like this puzzle quite a bit. Even with its chunky corners and low word count, the puzzle (by my count) only has 3 objectively "bad" entries: ANTA (a relatively obscure org. at best), AHS (plural sounds/utterances = blech), and TIRO (second banana to Cicero = not famous enough IMO). I'm sure the more pessimistic among us might argue for COHERER's or TYNE's inclusion in that list, but those entries strike me as more neutral.

...which brings me to the aspect I dislike most about this puzzle: all of the neutral answers. As a solver, I'm not certain that I would like struggling with a Saturday NYT only to uncover swaths of ho-hum. Does the challenge of the wide-open grid make up for this? Will the few interesting long entries (as always, your mileage may vary, but I'm including STREET SMART [probably my fave clue in the whole puzzle...an-n-nd...my original clue stayed...hooray!], SHOPLIFT, SCABBARD, SHALL WE, and MEGATRON in this category) and some tough, (hopefully) entertaining clues be enough to provide a satisfying solve? I have a feeling the answer is "for some solvers, this challenge will do just fine; others (speed solvers maybe...?) will likely be bored."

At any rate, the grid pattern/filling challenges/puzzle solving experience will be new, which is something I strive for. Like Jeff, I also admire Patrick Berry's/others constructors' continued efforts to produce puzzles that do not fit the typical themeless cookie-cutter mold, and, thus, I make it a priority to attempt the same.

Hope you enjoyed the puzzle and notes. I'm anxious to gauge solvers' reactions to this one. Happy Guy Fawkes Day!

This puzzle started as an earnest attempt to integrate central 14s (which, based on feedback, seemed to go over well in my themeless debut) and some corner stacks. Having gained the wisdom granted by a two-year publication lag, I still like the plural JEDI MIND TRICKS over the singular and UNIVERSAL DONOR with its only slightly edited clue. (Will/Joel added the "always.")

However, I now think a few of the corners creak a tad too much, especially the SE and NE. FLASH MOB (which surprisingly has not yet debuted in the Times), (full name) ELIHU YALE, I LOST IT, the scrabbly LL COOL J, NEO-DADA, STREAKING, THE HULK, BUM A RIDE and the "Hey-I've-seen-that-thing-before-but-couldn't-tell-you-what-it's-called" BALDRIC still strike me as interesting long entries, but the remaining seem too "so what?"/neutral. Plus, having to rely on DYER (blech!), ETAT, OEIL, DAHS, and CRUE to hold the puzzle together isn't ideal in a themeless. Oh well! Live, learn, and improve.

• Favorite (practically unedited) clues: [Coin collectors?] for SOFAS and [Auto-reply message?] for ECHO.
• Favorite Will/Joel clues: [1,000 or 1,000,000] for CUBE, [Soaring expense?] for AIRFARE, [One who's green after seeing red] for THE HULK, and [Feeling toward a supervillain] for ODIUM.
• Clues I'm most bummed to see edited out: [Planned engagement party?] for FLASH MOB and [Force-ful persuasion?] for JEDI MIND TRICKS.

Also, for those interested in my non-crossword persona, I successfully passed my second actuarial exam on June 17th and am currently searching for an entry-level actuary position or internship. If there is anybody out there willing to help a young'un jumpstart his career, I would be greatly appreciative and definitely buy you some cake... if I could somehow devise a way to deliver cake via email. Find a way, technology!

Or, if you're more interested in my crossword persona, feel free to email. Assuming my schedule is not too overloaded, I welcome anyone willing to put up...er, I mean...collaborate with me.

DAVID S.:

Dave and I met at the now-defunct Silicon Valley Puzzle Fest back in 2012, which feels like just a few months ago to me! Time flies when you're busy with crosswords, I guess. Anyway, Andrea Carla Michaels and I were giving a presentation about crossword construction, and Dave (who at that point had just started constructing) was one of the more enthusiastic-looking audience members. He approached us after the talk and showed us some of the puzzles he had been working on, which I remember being very impressed with. In fact, our first collaborative effort, which appeared in the Los Angeles Times back in 2013, spun off of one of the puzzles Dave showed me that day!

Dave and I have remained in touch ever since. I was thrilled to see him make his New York Times debut with an ingenious Paint It Black puzzle in 2014! He followed that up with one great themeless after another (and a handful of fun themed puzzles to boot!). Dave was also one of the most accurate Pre-Shortzian Puzzle Project proofreaders—he used to make Excel files with all the mistakes he would find, which were always both helpful and fun to read.

I don't remember much about the construction process of this particular puzzle (my 50th published New York Times crossword, built back in late 2014), but I'm pretty sure we started with my bottom stack in one of Dave's insanely wide-open grids! We went through a ton of possibilities for the top before finally settling on the one you see. Most of the clues are Dave's, though "Piano-playing Cat" for STEVENS was definitely mine!

These days, I haven't had as much time for crossword construction, but Dave and I still keep in touch . . . even though he went to Berkeley and I ended up at Stanford! We hope you enjoy our puzzle.

Well, it would seem I have the honor/curse of batting cleanup for St. Patrick this week. (Hey, Kameron, if you're out there reading this, I'll take one of those "I Survived ..." t-shirts [referencing his 11/14/2015 note], assuming I don't drown in negative solver feedback over the weekend.)

Accepted in October 2013, this puzzle was my third accepted NYT themeless out of seven attempts. I started with the central 16, THE BIG BANG THEORY, and then I placed black squares below the two G's hoping that I'd find at least one spicy -ING word/phrase. After those initial block placements, I went for as much open space as I could muster without going over the 72 maximum word count. (For some reason, I remember still wanting to clear the word count despite the extra column.)

In retrospect, I like the relatively wide-open space, a few neat entries (e.g. NO HARM DONE, EDO PERIOD, LIVERMORE [clued in reference to the new element], BROMANCE, TROLLING, HOUND DOG, GLEE CLUB, and some others), and the nerdy vibe overall. However, there's also a bit more glue and/or "killer" entries, notably APT TO, NYAH, BATOR, SES, SML, and BASINETS, than I'd allow in my current submissions. Looking back, these trade-offs are a bit hard to stomach, but I hope they don't detract too much from the puzzle's positive aspects.

As for cluing, Will and Joel changed a lot for the better IMO. Top honors go to 20A, 21D, and 31D with honorary chuckle points for 15A and 4D.

If you couldn't already tell, I'm kind of a comic book nerd. (More specifically, I'm a graphic novel fan; I prefer the freedom of reading a complete story told in one volume better than having to read the same story via monthly comics. My impatience knows no bounds!) So, unsurprisingly, I started this puzzle with GENERAL ZOD--somewhat fresh from his appearance in 2013's "Man of Steel," and soon to make a cameo in the upcoming flick "Batman V. Superman: Dawn of Justice" (...but you didn't hear that from me).

My initial interest in ZOD was simply for the geek factor, but, from a cruciverbal standpoint, I also valued the entry for its (mostly) alternating vowel/consonant pattern, common letters, and the single Z. True to character, ZOD seemed to fit well among the wide-open spaces.

Trying to improve upon my past themelesses and those from my predecessors, I aimed for: (1) few three-letter entries, (2) long entries all around, and (3) a central spanner in addition to that sprawling center section. I'm also a sucker for connectivity, which might explain my puzzles' relative ease.

At any rate, I like how this puzzle turned out and hope solvers cotton to it as well. =)

P.S. Did anyone else notice the eerie number of dichotomies in this puzzle? GENERAL ZOD crossing SUN TANNING (Kryptonians love themselves a yellow sun), BEARCATS and TOY POODLES, KCAL and CALTECH (with two different meanings for CAL), and CALTECH bordering DEVRY, among others. All connections are purely coincidental, but you're free to think otherwise. =)

This puzzle was directly inspired by the Andrew Ries Fireball puzzle "Straight and Narrow." Even though the puzzle was very well-executed as conceived, the meta aspect left me and other solvers a bit cold. IMO, Amy Reynaldo sums up the puzzle perfectly in her concluding paragraph: "I got this meta quickly, but it was extremely tough overall: just five solvers got the correct answer, a number so low that, for what I believe is a first for Fireball, editor Peter Gordon gave out a hint and an extra 24 hours after the contest had closed." For reference, I've seen Peter Gordon record upwards of a few hundred correct responses depending on the difficulty of the meta; so, Fireball's solving base is by no means small.

This means that at least 95% of Fireball solvers--which include many of the top-ranked crossword solvers among their ranks--may not have understood or even located the theme! This fact dismayed me--and not just because I was one of the Fightin' ">95%." To me, it signaled lost potential, which, as a constructor, frustrated me to no end. Frustration led to curiosity, and curiosity led to SIDE/BARS.

The biggest constructing difficulty for this one definitely had to be fixing parallel theme entries in the outermost columns. Many "bars," when placed next each other, didn't produce nice crossings. For the most part, I targeted "bar" pairs that had alternated consonant/vowel patterns; and, for the few spots that didn't yield favorable bigrams (e.g. NC at the start of a word or TM at the end of a word), that was where black squares and a few key three-letter entries came in handy.

As for cluing, my first impression is that Will/Joel eased up on the difficulty to help solvers better glean the theme...? At any rate, though the final edit has a few deceptive clues, there are no explicit pun ("?") clues. As sad as I am to see [Iconic line of computers?] for TOOL[BAR] get removed, solve-ability and the thematic "aha" are more important.

Anyway, I hope this was a fun, crunchy, Thursday outing! Get it? Crunchy...NOUGAT...oh, never mind. =)

When constructing a themeless, one thing I like to do is design my black square arrangement around an interesting entry. If done correctly, I feel this technique can help to squeeze in a few unusual letters, avoid the temptation of a prefab grid, and perhaps even create a unique solve.

For this puzzle, TJ MAXX was the entry of interest. Once I had placed black squares around TJ MAXX (a few under the X's and one above the J), the resulting NW and SE sections seemed to want to form stacks. But, placing a black-square finger in columns 5/11 meant that what is now 27D would've been either a long entry starting with "J" bordering a stack (would this be fillable...?...I pondered) or a stumpy 3-/4-letter entry starting with "J" (if I added another black square). Since I didn't like either of those options and wanted to avoid stacked 11's in the NE/SW, I opted for stacked 9's all around (with extra 8's in the NW/SE). Lastly, I tried to place the remaining black squares in such a way to give solvers multiple opportunities to enter each corner.

Of the entries in my original submission, Will and Joel liked the NE/SE but disliked a buildup of proper nouns in the SW and the (non)phrase LEAVE LATE in the NW. After a little more back-and-forth, we settled on the fill you see. Stuff like NESSES and ANAS aren't that great, but they're relatively isolated and don't have too many bad crossings IMO.

As for clues, Will and Joel changed quite a few to bring this puzzle up to Saturday difficulty. [Auto racer Luyendyk] became [Part of "Die Fledermaus"], [Widow's peak, e.g.] became [Plugs can move it forward], and [Figures in red] became [Red menaces?], etc. Fortunately for my ego, Will/Joel kept one of my favorite original clues--[Still the most?] for STEADIEST--and even felt that a question mark was warranted. At any rate, I hope solvers appreciate the added "crunchiness" in their Saturday puzzle; make sure to thank your tireless editors! =)

That's it from me. I'll have Ms. Swift lead you to the Sunday puzzle. Happy holidays and happy solving!

Not surprisingly, I got the inspiration for this theme after noticing a few sandbags on the side of the road. My first snarky thought was something to the effect of "How much use do these things even get? This place [the Twentynine Palms USMC base] is on the edge of the Mojave ... how often is flooding a problem?" (It happens occasionally as I later learned.) After this, I switched to mocking the sandbags themselves. "And what good of a bag are you anyway?" I thought. "If you lose your sand, you just become a regular, old bag."

Snark soon gave way to curiosity. Were there other bags whose identity was solely dependent on what they held? From the outset, I prioritized including ICE and SAND, since the ice bag and sandbag best illustrated the theme concept I was going for, i.e. a bag specifically designed to hold only one item.

After ICE and SAND, theme answers were a bit harder to find. I initially thought AIR would be a good choice, but airbags typically get filled with pure nitrogen or argon gas and not the 78% nitrogen, 21% oxygen, etc. "air" to which we Terrans are accustomed. Also, technically speaking, even ice bags and sandbags contain air. There were also a good number of "___bag" entries that didn't work with my theme definition, e.g. shopping bag, windbag, handbag, and a few others too lewd for me to list.

I ended up choosing TEA and BOOK but also liked MAIL and MONEY. (I didn't like BODY or GARBAGE due to the unpleasant imagery.)

As for the fill, it would seem that I tried to squeeze in some long downs and/or a tinge of scrabbliness. This resulted in a bit more glue than I'd allow in my current submissions; nonetheless, I hope the longer answers and lower word count add a little variety to your Tuesday solve.

Enjoy the rest of this week's puzzles. I know I certainly will!

To be frank, I was fairly shocked when Will accepted this puzzle in May 2014. (I sent it off in January of the same year for those curious about the publication timeline.)

After Will had accepted three of my first seven themeless attempts — two of which [the MTV.../GRAND... and WIISPORTS themelesses] have already published — I tried experimenting with different ways to build themeless puzzles. In my last NYT themeless, I locked the black squares in place first and then scrambled to find nice fill for the center that would also enable clean corners; however, with this puzzle, I started with the NW stack and tried moving a few black squares around until some goodies (GIGAHERTZ, ERUDITE, K.C. JONES, RAM INTO [with the vague clue]) locked in place.

This technique produced a pretty nice, albeit proper, NW section but eventually led to a more difficult time in the South — STRASSE, ATTIRER, IN SCALE, NTS, and AD-AWARE standing out as particularly irksome entries.

As always, Will and Joel did an awesome job spiffing up the clues. My favorites among their edits are [Goes one step too far?] for TRESPASSES, [Class for model students?] for ART, [200 at a 500] for LAPS, and [Grp. of 300 people?] for PBA. I'm also glad my clues for NAP, MAWS, DATA PLANS, and OLIVE TREE could stick around for publication. Unfortunately, my original clue for KIM JONG-UN, [Whom Dennis Rodman called "a friend for life" in 2013] didn't make the cut.

On a completely unrelated note, I really hope North Korea doesn't hack NYTimes.com today.

In the Xword Info constructor notes for the 13 September 2013 puzzle, Patrick Berry commented that he has "come to like making themelesses with wide-open centers, because if you can complete the center you'll probably make it across the finish line." My initial response to this was something to the effect of "Bah! You tease us lesser mortals." Yet, my curiosity led me to take his advice, and, by Jove, he may have been onto something.

Fearing the inevitable open-square bonanza, I preemptively added the "staircase" arrangement at the top and bottom. The 15-letter center spanner, LIVINLAVIDALOCA—which just screams for themeless usage in so many ways—came into play when I noticed that my designed grid made for decent crossings with the entry (i.e. it has lots of vowels and a few mid-word letters).

Besides that setup, there's really not much else to discuss. Because of the low word count, I didn't have too many options for the surrounding fill. I'm not overly fond of the ROLLOVERIRA/OVERDIDIT dupe, OMS, RFDS, PERMALLOY (though I do like its portmanteau-iness), SOO, and maybe MLLES (the French equivalent of the crossword-y SRTAS), but I think 6 or 7 clunkers is not bad given the wide-open grid.

If you like listening to music while solving, you'll no doubt like this blast from the past. Or, for a more recent video:

This was my second submission to Will and my first acceptance, so it will always hold a special place in my heart.

The puzzle began as part of an actual midterm I made for a DeCal (a program at UC Berkeley that allows undergraduate/graduate students to lead a self-designed course) on how to solve and construct crosswords. After the inimitable David Steinberg encouraged me to send Will some of my work, I thought of using this puzzle.

Well aware of the theme's relative simplicity, I checked XWord Info to see if the same theme had been done previously in the NYT. Indeed, it had! But, wait ... I had added an extra theme wrinkle that Mr. Bessette had not implemented: having the TERMs in the exact center of each entry and not just somewhere in the middle. Awesome! That seemed to me enough of a variation to reprise the theme some five or so years later.

After I had spent more time perfecting the fill (due to the theme density, many of the downs cross two or more themers and were therefore heavily constrained), I sent this off to Will and received an acceptance email in early November of 2013. Sometime in 2014, after learning of Will's new assistant, I happened upon a lovely article detailing Ms. Shechtman's "puzzling origins" and experiences working with Will. In the third paragraph were two of the same theme entries in my puzzle — plus one I had overlooked — and the same added theme wrinkle! At least I now felt qualified enough to utter the tired cliché, "Great minds think alike."

Still, the coincidence got me thinking: Is an idea or product still original if it was fully developed without prior knowledge of a similar, preexisting idea/product? Or, because of this time factor, would the idea be labeled a variation, a reinterpretation, or, worse, a rip-off? Then, to what extent is originality and craft affected by this race to beat others to the punch? Would Albert Gleizes have been a cubist had he known Pablo Picasso would eventually become the major pioneer of the art movement?

Anyhow, this is my long-winded way of saying that IMO an original execution is just as important as an original theme. Though we might try, we can't all start art movements. Even Ralph Waldo Emerson admitted, "All my best thoughts were stolen by the ancients." Sometimes, originality is simply a fresh look at an old idea.

Before submitting crosswords to the NYT and other venues, I tried filling a number of what I call "practice grids" to empirically find the limits of my fill capabilities. Many of these practice grids still (rightly so) remain unclued and unsubmitted mostly because my early fill attempts were not up to par. Despite the "wasted" effort, this practice served as a valuable learning tool and helped form the foundation for one of my personal constructing mantras: YAHOO (which I'm using as the backronym You Always Have Other Options). A revision can be as simple as a cheater square or as involved as a total revamp; regardless of the way one finds it, great fill always waits for those resourceful enough to seek it out.

Today's puzzle, a revamp of one of my "practice grids," serves as a nice example of YAHOO. In my "practice grid," I tried to fill the black square pattern in this grid to mixed results. After some analysis, I felt that the fill was a bit heavy with proper names and crosswordese-y stuff, e.g. TARTE, O-CEL-O, NEN, LEONORE etc.; however, I did like the stack in the NE corner and used it as a starting point for another iteration. It was at this point that I added two symmetrical cheater squares to (1) help eliminate LEONORE and (2) segment the grid just a skosh to allow for a little more fill flexibility.

If the filled NW section, particularly GAHAN, NESS, and the dreaded EMAGS (inelegant in the singular and just plain awkward in the plural), of that iteration had not bugged me so much, you would have likely solved this second grid. However, even after submitting that version and receiving Will's acceptance email, I still tinkered with the NW until I eventually found the fill you see in the published version. Incidentally, this new NW, which contained the word ONES, also meant that I needed to scour for a new SE (since the SE section in the previous version had the entry ONE AM). Fortunately, because I was only dealing with triple 8 stacks, that section wasn't too difficult to rework.

Looking back, the SW section, in particular SML, ABOU, and MEDI, stands out as a slight sore thumb, but I hope my extra efforts in other parts of the puzzle still make for a delightful solve.

Lastly, in case you're curious, Will/Joel didn't really change many of the clues this time around. I counted at least forty clues that went completely unchanged and only eleven new clues that went in a totally different direction from my original submission. Not all of the unchanged clues are "originals," but I will gladly take credit for 15A, 16A, 17A, the 33A/12D combo, 37A, 4D, 14D, and 46D. However, IMO, Will/Joel take the cake with 25D's clue. Until next time, YAHOO(ooohoooh)!

In my themeless solving experience, I've noticed that many constructors start with a seed entry. However, I'm not entirely satisfied mimicking what my predecessors have done before me. So, my goal for this themeless puzzle was to find two stack-able, seed-quality entries that wouldn't overly butcher the crosses.

I remember picking WII SPORTS early to fill one of those two spots, because (a) it would be an NYT debut, (b) it would probably be familiar to anyone who owned a Wii (the game was bundled with the console, after all, and even entered the pop culture sphere via a few episodes of The Big Bang Theory and the hilarious Tropic Thunder), and (c) I thought I might be able to form some nice bi-/trigrams from the crosses. After looking through a few candidates for the second entry, I latched onto DR SCHOLLS, because (a) that long string of consonants at the beginning seemed like it might just mesh with WII SPORTS and (b) the brand, though somewhat old, is still alive and (dare I say...?) kicking.

At this point, I modeled the black square pattern around these two entries and (what I thought would be) good potential bi-/trigrams. I immediately knew that the bigram "WD" would really only work at the end of a word or in between words (e.g. CROWD, LEWD, RAWDATA, SHREWD); however, once this crossing was dealt with, I had a few more versatile combos ("IR," "IS," and "SC") nearby. Anyway, long story short, if you had a slight inkling that the black squares in the NW area seemed conveniently placed, you have a strong intuition. = )

My original grid had some problem entries in the NE, and, though he didn't initially accept the puzzle, Will suggested that I fix the area and resubmit the puzzle. In addition to patching the NE, I also took some extra time to search for a more interesting south section. Fill quality can become subjective to a point, but I was happier with the trade-offs in the printed version.

As for the clues, Will and Joel changed quite a few — for the better. I'm happy to see my clues for 13A, 19A, 21A, 46A, and 42D survive with little editing, but Will's/Joel's clues for 26A, 48A, 2D, 32D, 35D, 38D, and 50D are undoubtedly awesome. I'd also like to highlight the clue for 14D. The clue [Hawke of "Boyhood"] looks innocent enough but actually serves a very important role: it solidly places this puzzle in the now. Not a year ago — which would be too early for a "Boyhood" reference, not a year from today — at which point "Boyhood" will have run the awards circuit and, unless successful, become slightly less relevant, but now — when the film and a few of its actors are currently up for Best Picture and (likely) all over the news. I believe this little bit of currency really adds to a themeless, and I'm glad Will/Joel agree, at least for this puzzle! =]

I hope your solve was enjoyable; if not, I'll try to get you next time.

Yay! My first themeless publication!

Though all my NYT crossword publications thus far have been themed, I think I more prefer experimenting with the concept of what makes a "good" themeless puzzle. (If you call it a freestyle, I would advise that you watch less "Dancing with the Stars" and/or Olympic figure skating.) Is there really such a thing as a "final evolution" of themeless crosswords, as exemplified in the PBTI (Patrick Berry Themeless Ideal), an acronym coined by Rex Parker? Or, might there be newer, undiscovered, perhaps seemingly inconceivable grid patterns that, when filled diligently, might also create a(n) equally/more pleasurable, albeit bizarre, solving experience? And, how does the great unicorn that is "style," a.k.a. one's personality as a constructor, fit into all of this? I hope my themeless puzzles make these questions even harder to answer for future generations. =)

As for this puzzle, which I jokingly labelled "MTV.Movie.Thefts" in my files, I toyed with a few ideas: (1) using 14s as semi-spanners and (2) having intersecting marquee entries. Though Will accepted my original submission (see right), I became increasingly dissatisfied with the huge black square formations on the left/right and the fill quality as a whole.

After some further experimentation, I found a nicer looking black square pattern but had to sacrifice my plan for having MICROSOFT, MTV..., GRAND..., and ADULT... intersect in such a way. However, the new grid allowed for better crossings with MTV... and GRAND... Also, oddly enough, ADULT SWIM reappeared as a fill possibility in a slightly different spot! (Prior to this, I had only heard legends of the crossword gods' mysterious ways.) Once I found that BRICK OVEN, in the spot symmetrical to ADULT SWIM, gave some good fill possibilities, the last task was to keep as many bland entries out as possible.

While the entries HEMPS, ETAIL, ADAK, ADAWARE, SNELL, OSAMA and others may not be ideal, I was more satisfied with the trade-offs in this puzzle over those in my previous iteration. Will agreed and opted to publish this one instead.

As for the clues, I see I have much room for improvement, which is to say that Will and his team rewrote new (often harder) clues for about half the entries. My favorite, completely unaltered clues are [Comes to a sudden close?] for SLAMS; ["Verily"] for TIS; [Playskool product tester] for TOT; and [Quite like] for FANCY (fortunately for the large majority of the NYT crossword solving base, I decided to not clue this with respect to Iggy Azalea's chart-topping single; you're welcome). Of Will's rewrites, I particularly enjoyed [Pre-texts] for IMS, [Limited expense?] for TRAIN FARE, [Big name in air circulation?] for SKYMALL (my original clue didn't include the first three words, which are key in giving the pun a fair chance), and [Like many ideals] for LOFTY.

I hope you enjoyed the solve and enjoy the remainder of the holiday season!

Like my NYT debut, this puzzle also has an Oppositeland twin, Patrick Berry's 5/15/2005 "Words to the Y's" Sunday NYT crossword. After (eventually) solving Mr. Berry's puzzle, I figured there had to be some words, like "hymn" or "sync," that used this same trick in reverse.

Before I began searching for specific theme phrases, I first brainstormed as many "y-less" homophone pairs as I could. Once I had amassed a good number of these pairs, I prioritized finding good entries for pairs with particularly odd/unique letter changes, e.g. SHAYS' --> CHAISE, SORRY --> SARI, or IDYLLS --> IDOLS. Using these "odd" entries as a starting point, I then attempted to find corresponding symmetric answers with as much variety, interest, and punniness as I could muster while also making sure to avoid letter changes that were too similar (e.g. DAYS --> DAZE, RAYS --> RAZE and/or [Willie] MAYS --> MAZE). If the theme entries evoke at least one or two chuckles, then I think they did their job.

Before and during the grid design process, I kept debating whether to use seven or nine theme entries. Since my theme entries were relatively short for a 21x21 puzzle, I wanted nine but also knew that attempting such might require some fancy gridwork or a sacrifice to the mighty crossword gods. Fortunately, there was no sacrifice necessary since SUNDAEBEST/DEVILRAISE intersected symmetrically with TRUSTEESIDEKICK/CHAISEREBELLION. Doing this really helped in spacing out the theme entries and creating some spots for 6's, 7's, and 8's in the fill. (Slight side note: Since the shortest theme entry was nine letters and since there were theme entries in the down direction, I tried to keep all fill shorter than nine letters to prevent "Is this theme or is this fill?" confusion.)

As for the fill, I particularly liked BUTTED (tehee), REDDIT, TOLKIEN, RICKROLL, SIZZLES, BADRAP, and BOOHISS.

If you were tripped up by or liked the clues for 1A, 72A, 116A, or the 11D/12D duo, you can blame/thank me. =) If this also happened for 65A (it's certainly convenient that ROWLING and TOLKIEN have the same number of letters), 15D, 19D, 41D, 56D, or 93D, you can blame/thank Will and his team.

Hello fellow xword(info)ers. Glad to make your acquaintance. I hope today's offering didn't rough you up too badly. =)

The direct inspiration for this puzzle was David Pringle's 2005 NYT April Fool's Day puzzle, but I'm sure other constructors' work seeped into my brain cells without my knowing. (Despite some similarities, this puzzle's conception dates before Mr. Steinberg's ERASE R'S puzzle.) My intent was to create a metaphorical yin to Mr. Pringle's yang: instead of using black squares posing as white squares, I would use white squares masquerading as black squares. The Rolling Stones hit "Paint It Black" seemed like a rockin' way to execute this idea.

To successfully pull off the illusion, I would need to make solvers think the newly "added" black squares could actually be black squares. This forced some immediate restrictions: (1) theme answers could not occupy rows/columns 1, 2, 3, 13, 14, or 15 and (2) theme answers needed to have the word "IT" preceded and followed by at least 3 letters. Furthermore, I knew that repeating the letter string "IT"/"TI" in any entry other than the theme entries would be confusing/inelegant, and thus made it a priority to avoid such.

After I had found a nice collection of theme answers and tried placing some, I soon discovered that placing all themers in the across direction severely limited *real* black square placement in the grid's center. Since my revealer was a devious length of 12 letters, I needed as much flexibility in black square placement as I could get. Hence, the pinwheel arrangement.

Filling the NW/SE center portions of the grid was also fairly tricky, but, once they went down, I was sure the puzzle would make it.

Fare thee well for now, brave solvers. But, beware the perilous puzzles that lie ahead! [insert evil laugh]