Thumbnails

I have to admit, I was a bit worried, knowing about Gary's musical bent. Today he gives us five songs that start with DON'T, all clued as a plea, encouragement, etc. And I knew four of the five songs! I consider that a tremendous victory, given my kindergarten-level knowledge of pop music. More to the point, thank you Gary, for choosing songs that if I know them, most people probably will too. Big sigh of relief!

It still was difficult for me, in that while I know most of the songs, I couldn't tell you who sang them unless I had a million monkeys playing a million guitars. Even then I wouldn't be able to put Journey together with DONT STOP BELIEVIN. (But on the plus side, I would have a monkey army to do my bidding.) Anyhoo, perhaps cluing with a famous lyric would have been more memorable? I really enjoyed one of Gary's older puzzles for this reason. I dare you not to have fun when completing: "We gonna rock down to..."

The theme did feel a bit repetitive for me, given that after two themers I could automatically fill in the first word of the rest. Generally Will doesn't allow repeated words within a single crossword, so there usually needs to be a strong reason to do so. I like the change of pace. Perhaps only four themers would have felt less repetitive?

(That last sentence sounded less idiotic in my head.)

Man oh man was I stymied by RINGALEVIO. Even after reading the wikipedia entry, I still wouldn't know how to play. Perhaps this is something like Canadians not knowing DUCK DUCK GOOSE? Might it be a regional NYC thing I really should master? I'll get my monkey army working on that.

Assuming people do actually play RINGALEVIO, I liked Gary's execution. It's a rough job to fill a 11/12/15/12/11 grid. It actually helps to have a grid-spanning middle themer, as this gives you more flexibility in placing your black squares, so bravo to Gary for choosing this central entry. I would have expected many more gluey bits, but there's nothing that jumped out at me as I solved. Impressive given the degree of difficulty; a testament to Gary's skill.

Where I expected to see some roughness was between DONT BE CRUEL and DONT WORRY BABY, as there are so many down answers which must cross both themers. And I did have all sorts of trouble with PLUTON. I sort of enjoyed learning the word, but I would have really enjoyed it if it hadn't crossed TRAM, or TRAM had an easier clue. (I personally get TRAM / DRAM / DRAY all mixed up when it comes to mine vehicles.) In the opposite spot, I had a similar difficulty with DUMONT, but here I was really glad to have easy crossings. If it were a Monday, I might balk at the ETATS crossing, but NYT-readers probably ought to know Les Etats-Unis.

Now someone go write a song called DONT IGNORE ME OR FACE THE WRATH OF MY MONKEY ARMY. Hmm, that splits nicely into 14/14/14...

– puzzle by Gary Cee

It's time for another edition of JEFF VS. DAN! For those of you who don't know of this magnificent contest, I race Dan Feyer, the reigning ACPT champ, at my own puzzles. In the past he's stomped me into dust, often by a factor of two (a factor of three, one time). So after a while, what could I do but up my level of cheating, going one step above knowing the answers? By studying the entire answer grid for five minutes and memorizing as many answers I could, I proceeded to trounce Dan (by two magnificent seconds).

So I decided to be the bigger man and give "the crossword titan" (note the theatrical air quotes) a handicap. I'm eventually going to beat Dan fair and square (by only knowing the answers — more than fair, really).

My time on today's puzzle: a blazing 3:56! My hand shall be victoriously raised!

Dan's time, listed at his blog... 1:49. Dagnabit! He beat me by a factor of two. The sun was in my eyes. And the clouds. And these grapes are awfully sour. Ahem.

BTW, fun to work with Dick on this one. We had a lot of back and forth, brainstorming and revising Dick's core idea. The grid came pretty quickly after finalizing the themers, as 15/14/14/15 is a pretty easy pattern to work with. And I enjoy the exercise of fitting in good short fill like ONE TWO, NO PROB, BAD RAP, R KELLY. BTW, I actually don't know who R KELLY is. I mean, I know the name. But if you put Eminem, Psy, Lil' Wayne, Yo-yo Ma and R Kelly together, you could easily convince me that they were The Village People. R Kelly is the construction worker, isn't he?

– puzzle by Dick Shlakman and Jeff Chen

What a neat idea! Congrats on the debut, Luke. When I got to that central entry, I smugly sat back. And entered MOUNT RAINIER. So what if it didn't fit? There had to be some rebus-y shenanigans, right? Some Seattle-ite I am. I wasn't aware that MT ST HELENS spread ASH into 11 different states — fun to learn!

As with any established theme type, it's important to do something new, something different, or at least incorporate snappy themers which add zing. Art must evolve or it dies. And I like what Luke has done here. Instead of just saying "Why don't I do a rebus with ASH squares?" he uses that fact about spreading ASH into 11 different states as a rationale. I like that step to go above and beyond. It would have been really cool if the eleven states had been incorporated somehow. Perhaps if 1A had been something like (WA)LTER and 1D (ASH)CROFT — a two-way rebus? Those extra layers are tough to incorporate, but how cool would that have been?

The grid did feel a bit restricted to me, so it wasn't a surprise to find out it was an 80-worder. I didn't mind that so much since there were still a lot of long answers, but the abundance of three-letter answers (25) did make it feel a bit choppy for me. Will typically doesn't allow more than 22-ish three-letter words in a puzzle, and for good reason. Each three-letter word typically has been used so much that it's tough to come up with strong clues for them (that haven't been used before). And to me, so much short stuff brings a feeling of inelegance, both by making a puzzle look constricted and producing a choppy solve where you have to switch from word to word more quickly. Subjective, of course.

A couple of rough spots, not surprising given the 13-letter central answer plus eleven instances of ASH. Even six-ish rebus squares can be difficult, so having eleven is quite the challenge, especially when you consider you can't duplicated answers (if you have ASHEN, you can't use ASHES). The NW and SE corners are where I'd expect to see some difficulty (considering they're the biggest chunks of open white space), and there are some chunky bits. By fixing ASHCROFT, ASHAMED, and TEXAS HOLD EM (great entry!) into place, you've constrained three sides of that subspace. Tough to fill cleanly from there, and OON / FLIC / AREOLE are not a great trio. I wonder if moving that first ASH square to the start of 5D would have been better (ASHRAM, A SHARP, etc.)? Leaving that long 1D unconstrained might have given better flexibility in filling that tough region.

Finally, some great clues. Luke's already pointed out the clever clue for NOOSES and LOO — bravo, sir! I also liked the one for EVEL, which had me thinking about Olympic long jumping.

Congrats again, and I hope to see more from Luke! BTW, Will recently put out a note that he's in greatest need of Sunday-size puzzles and non-rebus Thursdays.

– puzzle by Luke Vaughn

★ I pity the poor fools who have puzzles near Patrick Berry. (Cue the sad violins.) Some thoughtful readers have told me that they don't like the fact that I pick a Puzzle of the Week, and I appreciate that feedback. But 1.) I like pointing out fantastic work and 2.) that's what some (many?) daily solvers tend to do anyway. For me, it's a good reminder that there are other people out there with much, much better construction skills than me, and if I want to be one of the greats, I need to keep working at it by studying, practicing, improving.

This PB was no different, giving me such unadulterated pleasure. So instead of qualitatively analyzing the puzzle as per my usual, I'm going to do something different: attempting to QUANTIFY why this work is so good.

People often ask me how they can get a themeless puzzle into the NYT, so I've given this a lot of thought. I've come up with a formula that I'll revise and evolve over time, hopefully keeping it simple enough for the non-mathy types. As a finance guy most recently, I liken the evaluation process to the decision whether or not to acquire a company. You buy something for its ASSETS, ignore the neutral stuff, and discount for its LIABILITIES. You can then put a price on ASSETS minus LIABILITIES, yeah? (Roughly.) For me, I think the odds of an acceptance become high when:

1. LIABILITIES < 5 and
2. ASSETS minus LIABILITIES > 10.

What do I mean by ASSETS? Stuff that sings. This is subjective, of course, but here's my assessment of the snappy answers Patrick provides us today, each of which I'll count as one point each:

• ON A LEASH
• IM BUSY
• AGRICOLA
• MARGIN OF ERROR
• BEFORE I FORGET
• BATTLE SCARRED
• I ROBOT
• IRISH PUB
• PET PEEVE
• BATTER UP
• VIES FOR
• HAVE A NICE TRIP
• TURNED RED
• ESCARGOT
• RYE BEER

And the liabilities? Things like partials, abbreviations, esoteric foreign words, pluralized names, etc. Each one will count as one point, except for "puzzle-killers," ug-ug-ugly answers which effectively take a puzzle out of consideration all by itself (RSI, for example, which killed one of my themeless submissions). Here's my assessment of Patrick's liabilities today:

• (insert sound of crickets)

The final count: ASSETS = 15, LIABILITIES = 0. So, Patrick meets the first criteria with flying colors. And the second criteria? ASSETS minus LIABILITIES = 15. As an analyst, I'd put a STRONG BUY recommendation on this one. (Never mind the fact that there's no price already set, you smart-aleck broker/analyst types.)

Will, if you're reading this, perhaps you could comment? Am I close in my assessment methodology or way off?

It's a thing of beauty, especially considering it's a wide-open 66-worder. (That's another point in the ASSETS column, actually.) And the cluing for IRISH PUB, ESCARGOT, BIPED, POT... For all those constructors looking to get published in the NYT, I'd suggest studying this one in detail. Try deconstructing and reconstructing it to see what you can learn through the process. Many of the great artists copied the masters for years before finally coming into their own, and that process was key to their emergence, right? Well done, Patrick, another beauty from the master.

– puzzle by Patrick Berry

Another 62-worder from David! Impressive. These low-count grids are extreme challenges. The huge amount of white space in the two big corners is daunting to think about filling. It's pretty unusual to be able to cleanly fill a region of such size while stuffing it with good material.

I was particularly delighted by the center section. I smiled upon uncovering DIVE BAR, and then further tickled to see it crossed by BIKER CHICK. Then when FAKE IDS and JIM BEAM and RATED R came into place it felt like I had stumbled upon a real treasure. It's tough to jam in that much great (and related) material into one place. And doing it without any real compromises is very impressive. That entire center section sings.

I might disagree with David on BIKEL though. Perhaps it's "clean," but I would argue that it's not "preferable." The difference to me between the central area and the NE is pretty big, perhaps because of how lively and clean the center is, but perhaps because BIKEL and BOMBE feel a bit esoteric to me. It's one thing to have some basic knowledge of a lot of different subjects, and another thing to know a specific area in such detail. Because it's a Saturday puzzle I believe that area is fair. It just didn't feel as fun to uncover as the central section, to me at least.

Like Will mentioned in his comments yesterday, I think that if having that BIKEL and BOMBE made the central area possible, that's well worth the price.

I expect that a NYT audience ought to at least have some knowledge in most areas. A Renaissance man/woman, right? If BIKEL had been a more famous Tevye player, like TOPOL (who got nominated for an Oscar in the role) or if BOMBE had been a more prevalent dessert, like TORTE, I would have liked that corner better. As it was, I had to guess whether it was BOMBE or BOMME or BOMTE, all of which looked French(-ish) to me. Totally subjective, of course. There are undoubtedly people who are indignantly reading this while watching a video of BIKEL or eating a BOMBE.

Ah, puzzle flow. One puzzle Jim points out help demonstrate my point. Note the seemingly small difference between the similarly shaped one from May? Today's puzzle stymied me in the NW corner, as there's only a single answer that can help you break into it (GENDER BIAS). If you can't grok that, like I couldn't, you're forced to work a mini-puzzle completely separated from the rest of the puzzle. That's not always a bad thing, but for a themeless puzzle, I feel it's much more elegant to have a high level of interconnects all throughout the puzzle.

Now check out the previous puzzle again. If you can't crack a single answer in the NW, you have multiple shots on goal to work into it — EMBEZZER and NOGALES both give you opportunities to uncover a little bit, giving you a hint to what LATE AUTUMN might be.

But today's arrangement does allow for some impressive fill. That center alone is gold. And both big corners are quite well filled, if not star-studded with marquee answers. I liked the NW in particular, even though I couldn't actually solve it (I gave up after 40 minutes). And I appreciated David's note about how he came up with BINGED ON, a great entry. Too often people think crosswords are generated by computers, but I find that a mix of computer assistance and pattern recognition (and trial and error) produce the best results.

Today was a tale of three cities, with one mini-puzzles that I solved (the SE), one that I loved (the center) and one that asked me to practice TOLERANCE. Still, a well executed puzzle overall.

– puzzle by David Steinberg

Ah, THERE'S the Fourth of July puzzle I was expecting! An homage to THE STAR SPANGLED BANNER today. One aspect that I really enjoyed about this puzzle was a post-solve analysis of the notes. Check out how much care Dan put into his layout of the musical STAFF. Not only does he place the notes in the correct musical intervals (SOL to MI is a third, SOL to DO is a fourth, etc.) but he gives us good approximations of the note durations (half notes, quarter notes, etc.). I like those careful touches.

Tribute puzzles can be hard to construct, because you run the risk of people thinking it's a straight forward tribute puzzle. Sometimes they tend to get boring if they're straightforward, more akin to reading a Wikipedia article instead of doing something fun. I like that Dan gives us interesting bits of trivia in addition to the obvious FRANCIS SCOTT KEY. Learning about the PRISONER EXCHANGE and the BRITISH PUB SONG was pretty fun. Who knew? Might explain some things about America today. Huh.

A few things did give me pause though, like what makes WHITNEY HOUSTON's performance better than ROSEANNE BARR's (don't answer that). Seriously though, WHITNEY HOUSTON but no JIMI HENDRIX? Sigh. I suppose there's no accounting for taste.

A lot of tough crossings today. Maybe not for experienced crossword solvers, but things like TITI/ISERE and ERST/AIT are going to be rough on some poor folks. When a puzzle has this many constraints, those types of things are going to happen. Going up to 140 words (this grid is 138) by breaking up FACTOTUM could have helped the former, but the latter is right in the midst of the second MI. Very tough. ASTI / STOA is right in that region too. I might have liked it better if two themers were eliminated (EIGHTEEN FOURTEEN doesn't do much for me, personally), and the scale notes spread out from the themers more. Just one man's opinion of course.

About Sunday puzzle word counts. Will's maximum is 140, and it's hard to adhere to. (Rich Norris at the LAT and Patti Varol at the Crosswords Club go up to 144.) So many times I wonder if ugly crossings and glue-y entries could be avoided by going above 140. But I do like how it forces constructors to work in longer fill, which usually enhances the solving experience. In today's case, as much as I like the word FACTOTUM, I'd much rather see it broken up, as the price of CRO, TITI, OTTS, ISERE in concentration was too high for me. It might have been just fine if there weren't already other compromises in the grid due to high themage — tough to say. Always difficult trade-offs in construction, especially since that opposite corner (the SW) turned out quite nice. You hate to break up a strong region like that.

Finally, a couple of beautiful clues. [Penn station?] repurposes a famous train station to clue IVY LEAGUE (it stretches the definition of "station" a bit, but in a fun way). And [Long arm] seemed so innocuous. I figured it had to be some term relating to NBA players or policemen (the long arm of the law). But it's simply a long "arm," in the sense of a firearm. Beautiful stuff.

– puzzle by Daniel C. Bryant

A few months ago, I joined CrosSynergy, a syndicate of crossword constructors providing daily puzzles for the Washington Post. Bob Klahn, who runs the group, has been extremely helpful in getting me oriented, and Lynn (a long-time CS member) has been great as well. It's been a nice reminder to me how important it is to know your audience. I had started by aiming too hard, since that's 1.) the type of puzzle I like, and 2.) NYT-ish. It was great to get the feedback that for the WP daily solver, I was making (cluing, especially) puzzles way too tricksy. Thanks, Lynn!

Onto the puzzle. I like how Lynn casually commented how yeah, maybe she'd toss in BEAUTY / PARLOR. Typically it's difficult to do such a thing, especially in the method she used. Placing that pair of answers on the west and east sides of the puzzle constrains them quite a lot, and that can often require glue-y fill to execute. But look how beautiful that west section came out. Maybe THEIST isn't a "Monday-ish" word to some, but I'll reiterate my point about knowing your audience. A NYT reader might not be expected to know this word, but shouldn't an educated person be expected to know the word "theology"? It's not a big step from theology to THEIST. And ELISHA Cuthbert may not be someone everyone knows, but she's a costar of a pretty popular TV show ("24"). Plus, Lynn's been very careful to make the crossing answers transparent.

The east side isn't as pretty to me, as TRURO strikes me as both 1.) one of those names that either you know or you don't and 2.) a bit esoteric, especially for West Coasties like myself. That said, Lynn's been careful to make the crosses very gettable, so I don't find it unfair. Just not as elegant as the west section. Simply my opinion, of course.

I personally had a tough time grokking the theme, as the sequence WASH, CUT, SET, DRY, COMB is much more complicated than my personal system (CUT). But I liked seeing that sequence after finishing the solve. As Lynn mentioned, fun to get that a-ha moment. I unfortunately stumbled upon BEAUTY and guessed PARLOR very early on, so it didn't have quite that impact I would have liked. That's why I prefer to have revealers (if they're necessary, and here I think Lynn made a good decision since it's a Monday puzzle) toward the end of the puzzle, where you won't accidentally run into them. Fitting in BEAUTY PARLOR at the end is not easy, though, especially considering Lynn's already high theme density.

What with the very high theme density (five themers plus BEAUTY / PARLOR, it's not surprising to see a few rough spots. The center in particular struck me, what with ERGS, EPEE, and REDYE all crunched up. Perhaps I wouldn't have thought as much about it if the theme hadn't been about the BEAUTY PARLOR. As it was, I had a hitch, wondering if REDYE was a secret Easter egg or a bit of accidental themage?

Finally, I really enjoyed the clue for YARN, especially given that Monday clues tend to be straightforward by necessity. Not only is it playful, but it might subconsciously give a little lift to a solver at the beginning of a work week. I think it's fun to aim for that in generally, personally. What should crossword puzzles be, if not uplifting? As with this clue, I had a ball with Lynn's puzzle.

– puzzle by Lynn Lempel

Two more CrosSynergy members on the byline today! Yesterday Lynn Lempel, today, Gail and Bruce. Nice. There are fifteen of us. By my math, three CS members will have written tomorrow's puzzle. Numbers don't lie.

One thing I like about good "word that can follow both words of X" type themes is when I don't see them coming. That usually means the themers must be snappy and not feel contrived, otherwise running across a made-up-sounding entry is a sure giveaway. I enjoyed most of the themers today, SPORTS NUT in particular. HIGH ENERGY is also a fun one, evoking images not only of energy bars but of yappy dogs and certain people on my Ultimate frisbee team.

COFFEE ROLL befuddled me, but it does appear to be "a thing." Perhaps I would like it better on a later-week puzzle though, as something that gets less than 100K Google hits when in quotes feels a bit esoteric to me. Generally, that's the bar that many of my constructor friends seem to use. This does serve as a good example of potential issues with this type of theme — there are only so many words that can follow BAR. Pairing them up into snazzy theme answers which will also fit symmetrically into a crossword grid can be tough.

As with all puzzles containing a central nine-letter answer, the fill is harder than average to execute on. There's sure some nice fill in the corners, which can always be tricky. And I absolutely loved what Bruce and Gail did with the SW corner. That's the way to fill an OPEN SPACE, not only selecting juicy entries like GUFF, JIFFY, EDIT OUT and RUN-OFFS, but tossing in a J without needing a glue entry to do so. Well done. As a solver, I appreciate having a couple of Scrabbly letters here and there. Spices things up. I'm not sure if this is a universal desire (probably not), but I do get the anecdotal feel that solvers generally are on the same page as me.

The north is a different story for me, when it comes to Scrabbly letters. It's fun to toss in a Z as a constructor. But having an awkward partial like RUE DE at the top of the puzzle, plus ARROZ crossing EZER... that would be tough for some solvers to suss out. It's one thing to know MEIR, and another to know EZER, methinks. And I think it's fine to expect a solver to be able to figure out some foreign words (especially in a Romance language), but ARROZ is pretty tough to derive. Along with URIEL, that north section strikes me as inelegant, especially given there are few constraints up there and plenty of other ways to fill that region.

As always, simply a personal opinion. During my solve I felt like there was quite a bit of the ASCAP, REATA, GAPERS kind of stuff (some of which is expected due to having five themers, especially given the center themer splitting the puzzle in half), and in retrospect, if that north section had been cleaned up, I don't think I really would have noticed the crosswordese overall. Anyway, I'm sure people will be writing in from the RUE DE la paix, saying that they both loved the Z and are aghast that I don't know who EZER Weizman is.

Finally, a comment on CGI. A friend of mine is in the computer graphics area, and when I used the term CGI the other day he got all riled up, saying that the proper term is VFX (visual effects). Terms and acronyms go by the wayside all the time, but I still wonder whether CGI is perfectly fine. Hopefully Tony's not reading this post...

CGI!

– puzzle by Bruce Venzke and Gail Grabowski

A STAR-STUDDED puzzle indeed! It confused me a bit that this puzzle would run today rather than… well, whenever the Oscars are. But I liked the revealer a lot, pointing to the fact that the word STAR is hidden in entries throughout the puzzle. The "word hidden within themers" can be a bit hit or miss, and the big factor that makes this work for me is the quality of the revealer; such a red-carpet answer. Would have been very different if the last word had been STAR with a [Word that can follow X, Y, Z…] type answer.

Interesting layout today, Bruce working in a lot of long fill. I was confused where the themers were during my solve, so I highlighted them below. Now that I look at their clear placement, it makes me feel silly. Erp. Anyhoo, turns out there are seven of them — quite a feat. Many of the themers were strong, TOURIST AREA and COSTA RICA echoing each other nicely. JUST ARRIVED was my favorite, not only a great entry in itself but an uplifting one, which will likely evoke many strong memories for parents. Good stuff.

ASTARTE… such an interesting entry. A few years ago there was a STAR rebus puzzle which featured ASTARTE, and it completely baffled me. Even when I saw the answer, I was sure it had to be wrong, or perhaps it was an insider's joke, two pieces of crosswordese (ASTA and RTE) jammed together. Turns out I learned something, as today ASTARTE fell like a domino. As my wife always tells people, crossword solving is all about practice.

I wasn't wild about the fact that ASTARTE is the only themer where STAR isn't broken across two entries. (Someone correct me if it really ASTA RTE.) I often like seeing high theme density, but adding those two extra themers confused the picture for me because it felt like I kept on running into starred clues (how meta!); plus, the price of BAL / ETE / TOR felt high. I liked the SW corner better, only having RLS and EST as a relatively low cost, but uncovering ASTARTE was only fun for me in that it made me feel like maybe I do learn something from all these crosswords.

I like long fill. It typically adds a lot for my solving experience, affecting it positively if there's great bonus material. I love both MADAGASCAR and WISECRACKS. Great entries, both of them. But already having a little confusion on what was a themer and what was not, having two entries that were actually longer than several of the themers felt a bit inelegant. I totally see where Bruce is coming from though — as a constructor, there's a strong drive to get extra material in the fill. It's a huge bonus. And with his layout of themers, it would have been difficult to work in long downs more than the THAT'S A GO and CREATURE spots.

So perhaps I personally might have executed a little differently, but that's what's great about having such a wide range of constructors. I'm sure many people will read this and much prefer the way Bruce executed it. A great revealer and so much excellent material packed inside.

Finally some great clues for KEG, DOODLES, and BEER GUT. I especially liked how [Marginal things?] made me remember the story of Fermat doodling in his margin that he had an elegant solution to his theorem, but the margins weren't big enough to write it out. And he left the math world hanging.

I have one final note of unparalleled, elegant, clever, genius-inducing brilliance, which is

– puzzle by Bruce Haight

There's some brilliance behind this puzzle. I wouldn't have fully grokked it if I hadn't gotten a heads-up on the multiple levels of theme:

In the solution to the 7/10/14 puzzle, the answers to 5-, 54- and 63-Across, and 4-, 12- and 50-Down, are preceded by the invisible word SILENT. Additionally, each answer word crossing these contains a silent letter at the crossing point.

I've highlighted the theme answers and ghosted out the silent letters that head them. I wish I had been able to comprehend all the layers myself (stupid brain!), but I'm glad to see and admire it in retrospect. All the interconnect, too. John could have satisfied himself with only using short words with silent letters like ISLAND, but he decides to go big and utilize ELLIS ISLAND, so much better an answer than the plain old ISLAND. Not just HONEST, but HONEST WOMAN. I admire the big-time thinking.

The "go big" approach did produce some compromises, as John noted. It's one end of the spectrum. Other constructors might have chosen a less ambitious approach, perhaps only using short words or only packing in five theme pairs (or both). I usually don't mind a few bits of crunchy glue when a concept merits it. The frustration I had in trying to suss out DEMESNE for example though… it's a tough call. I also had a rough time in the DPT / STYLO section — usually I might suggest a set of cheater square s in the NE and SW corners, but of course that would have made the placement of MOVIE impossible. Tough trade-offs.

I usually like to see symmetrical themers, and today was no different. There's something so elegant about having not just the crossword grid be symmetrical, but having all the themers too. It might be too much to ask to have all pairs of themers be symmetrical, but I might have picked this puzzle for the POW if at least the "words that can follow SILENT" were symmetrically placed. There is something to be said about having asymmetry though, since it makes the puzzle even more challenging to solve (no freebie of knowing where another themer is once you've located one).

Finally, I'm a bit mixed as to the theme not revealing itself in a more natural way. It's too bad that a lot of solvers will likely finish the puzzle and not see all its grandeur, and then they'll miss the notepad tomorrow. Such a shame. Of course, having SILENT as a revealer in the puzzle would have made things clear, as its clue could have explained everything, but that's a pretty blunt force instrument. I don't have a great answer, unfortunately.

Overall, a memorable idea that will stick with me as much as another one that I loved (after someone told me what the heck was going on).

– puzzle by John Guzzetta

Fun to see another one from a strong pair, wizened Judge Vic and the young Skywalker padawan. Er, Ezersky. If history tells us anything Vic, you might want to check Sam's dark-side pockets for light sabers. Seriously though, I enjoy seeing collaborations, and an established one that has this much fun working together gives me a great big smile.

A lot of strong material today, anchored by the nice central grid-spanners, ANAKIN SKYWALKER and FACEBOOK FRIENDS. They go big with the grid design, not just satisfied with four additional sets of triple-stacked corners. It would have been easy for them to section off the SW and NE corners to make them easier to fill (similar to how the NW and SE corners dead-end), but they chose to go for 1.) more puzzle flow and 2.) two additional long answers in JOE CAMEL and KING SIZE. I like how open those two corners are, giving the solver multiple ways in.

That decision did cost them a little, in various entries such as ARNE (not the most famous of cabinet members) and ONE AM (an arbitrary time) and JER (as much as I love Seinfeld, this answer looks so odd to me). Still, the overall effect is well worth the trade-off.

Speaking of the dark side, finding an ENERGY BAR in a SEWER LINE (and eating it George Constanza style) TASTES BAD all right. Not an image I can get out of my head unfortunately now. (Someone pass me a light saber for an emergency lobotomy.) It won't bother everyone, but if it grossed me out, I bet others will feel similarly. TASTES BAD also seems a bit arbitrary to me, opening up the door to TASTES GOOD, TASTES SALTY, TASTES FISHY, TASTES LIKE AN ENERGY BAR I FOUND IN THE SEWER LINE. All in all, it's a reasonable answer, just not one I would consider stellar at the big-impression 1-Across location.

Normally, I don't care for cross-referenced clues, which make you jump all over the grid. I ignore them most of the time, unless the two entries are close in proximity. But today I appreciated the SITS / BANC cross-ref, since it helped me finish the tough NW corner (in which I was stuck). It helped improve the quality of my solve, and that's really the most important factor to any puzzle.

I loved seeing PO-PO in the grid! I only heard the term a few years ago, but it stuck in my head. I like getting a bit of slang worked in. Too much goes away from what makes this the NYT puzzle, but every once in a while a piece of good (and up-to-date) slang does wonders. To my delight, WHAT'S UP G? has been used (although I've only heard the "more correct" WHAT UP G?). The first time one of the kids I work with said YO WHAT UP G? to me I was baffled. Gave him the silent head nod, and that seemed to be acceptable.

Finally, some beautiful clues. Some of my favorites were actually DERIDE and ONUS, with two seemingly innocent clues. [Knock sharply] threw me off track, thinking about RAP AT THE DOOR, and [Large charge] felt like it should be AMEX BILL or MEGACOULOMBS. But my top choice today was [Outerwear for moguls?], which misdirected to ITALIAN SUITS or something. Great a-ha when I figured out that the moguls were referring to ski slope moguls.

– puzzle by Sam Ezersky and Victor Fleming

Crosswords are like cars. There are some universal assets for both: great themes and strong fill for crosswords, and gas mileage and reliability for cars. But it makes a lot of sense for car makers to create "concept cars" to show off what might be possible. These concept cars may not be very practical, but they sure are cool to look at. It wouldn't be much fun if EVERY car were a concept car, but one every once in a while creates a lot of buzz.

Today's puzzle is in that vein for me. It's crazy-looking pattern. Eye-popping. Those giant white spaces in the top and bottom scream LOOK AT ME! As a constructor, I look at that and shudder for fear of execution difficulties, but for most of the NYT audience, I bet more of a "hey, you gotta come take a look at this!" is in order.

With any themeless featuring a lot of 15's, those grid-spanners must pull their weight, since they won't leave much room for other good fill. If you hit the "Analyze" button down below, you'll see that the longest entries after the 15's are only seven letters long. It's generally harder to come up with great IM IN AWE type entries when you're constrained to only seven letters. More often than not, more neutral words like SCOWLED or INSEAMS will fit in better.

Out of those six grid-spanners, I thought LIBATION BEARERS and I DONT FEEL LIKE IT and EARTH SHATTERING were fantastic. SEMI SOFT CHEESES and ARIZONA CARDINAL lean more toward neutral for me. And TAPPAN ZEE BRIDGE... yikes! It took every cross to figure out, and even then, TAPMANZEE (with MATTI) sounded reasonable. (I think too much about all the fun practical jokes I could pull if I had a chimpanzee. Even better if I had a tapmanzee.) Overall, not bad on this front.

Showy concept cars often have compromises in order to fit everything under the hood, and that's seen in some entries like DERAT, SOLFA and ARRAS and their crossing, the little corners filled with CPL / URI and MTA / SSN. Not what I'd like to typically see in a themeless. Additionally, the puzzle's flow wasn't ideal, being broken into three mini-puzzles: top, center, bottom. I was stymied by the top (which is fine, as I'm sure some people will know the TAPPAN ZEE BRIDGE off the top of their heads), without any way to break in.

That said, the overall effect is still pretty neat. I wouldn't want to see this type of puzzle every week, but a few times a year it provides for a great visual effect.

Finally, a few notes on clues. [They run up legs] for INSEAMS is awesome, exactly the type of clue that resonates for me, personally. (I was thinking about spiders running up legs, or better yet, miniature tapmanzees.) The clue for SCOWLED was less interesting to me, going for an esoteric definition of "scowl" (I looked up "scowl," and one of the last definitions was "lower." Then I looked up "lower," and one of the last definitions was "scowl." Some help, stupid dictionary.). This kind of cluing feels too SAT-ish and unsatisfying for me, but I'm sure others will enjoy it. Even prefer it.

Personally, it was a really nice touch to end on [Bright spots], which I thought were going to be something like HIGHLIGHTS, but was indeed, SOLARIA. Nice misdirection.

– puzzle by Tim Croce and Alex Vratsanos

★ What a perfect title for this clever idea. Self-evident, indeed! I'm impressed that Tom was able to come up wtih so many snappy phrases that fit the pattern. The Yogi Berra quote in the middle sings, and how could you hate HATERS GONNA HATE? But the real topper is the quote from Gertrude Stein represented in a repeating circular pattern in the center. So many levels of delight today.

In just four published puzzles, I've picked two of Tom's as Puzzles of the Week now. Not bad sir, not bad at all. As Jim and I discussed, Tom's a constructor to watch. I've enjoyed my correspondence with him — seems like he has the right attitude: humble and willing to listen, learn and drive himself to improve. Hoping to see a lot more from him.

Liz Gorski's rebus interpretation of the Stein quote was another fun one. I appreciate Tom's new interpretation, taking things a step further.

Another thing I admired about this puzzle was its scientific tone. It's not going to RESONATE well with everyone, but I personally enjoyed seeing ENTROPY, LIGAND, and TITRATES in there, triggering good memories of college chem and physics classes. I bet it will trigger shudders for others, but you can't satisfy everyone. HATERS GONNA HATE, as they say.

I did wonder if this would have made a better weekday puzzle. A 21x can get a bit tedious to solve if there's not some factor that forces it to use an oversize grid. A visual element often does that for me. Grid art is another reason I find compelling. For me, the best Sunday puzzles are those that absolutely, positively, cannot be done in a normal 15x. All in all, I thought it was really nice to get all those snappy theme answers today, but it did get (pun intended) a little repetitive.

Neat idea, well laid out (great spacing between his themers and the central element), some strong, smooth fill and cluing, and a neat visual element. A winner of a Sunday in my book.

– puzzle by Tom McCoy

★ A fun theme, a great reason for a "word hidden within phrases" type puzzle. This theme type gets done often, so to have a fun addition like these "eerie encounters" was really enjoyable. I would have LOVED to see the UFOs "land," but I personally found that nearly impossible to pull off (as I'll explain in the final paragraph).

A nice construction today, one with few glue entries. I really appreciate that on Mondays, where I feel it's so important to be friendly to a novice NYT crossword-solving audience. It's nice to get in good long fill like SCALENE, PAGE LAYOUT, FOGHORN, but even nicer to only have a few bits of A DIME, IS IN. Welcoming to a newer solver. It's clear that MaryEllen took care in filling her grid, and the extra effort is much appreciated.

I used to think MITRE and OCULI "aren't Monday words," but my philosophy has shifted over time. Being the crossword for an educated NYT audience, I believe it's acceptable to have semi-esoteric words if the crossings are all fair. And some would argue that the MITRE (the Pope's hat, for example) is something the NYT audience ought to know.

OCULI is tougher — if you don't know Mauna LOA, you might be in trouble. I think this is the one problematic spot of the puzzle. Note how OCULI crosses three themers? Nothing else can fit the O?U?I pattern. So I'd prefer to see the four themers spread out more, which would allow for more black squares separating them, and thus more flexibility in filling. It would likely mean that the long across fill (PAGE LAYOUT and CONFINED TO) would need to be broken up, but I don't mind that, since I found it to be inelegant for those answers to be almost as long as the themers.

Finally, I'm sure friends will ask me if it bothers me that I had a very similar theme in the LAT back in late 2011. (Answers at C.C.'s Crossword Corner site.) The answer is no. Two constructors come up with similar or even identical ideas all the time. The cruciverb.com database is great for checking to make sure your theme hasn't been done before in non-NYT outlets, but it usually lags a few months behind, so I bet MaryEllen wouldn't have found my themers in a search when she was constructing hers. Additionally, there is some overlap between NYT and LAT solvers, but they're largely different audiences.

The truth is out there!

– puzzle by MaryEllen Uthlaut

A business-oriented puzzle, right up my alley. And specializing in entrepreneurs, even better! Back in 2002, I had the good fortune to help a friend start a company from ground zero, Acucela Inc.. The start-up experience was rough, costing me many 80+ hour weeks and taking several years off my life, but I wouldn't change it for the world. Who knows if emixustat will end up working for dry form age-related macular degeneration, but if nothing else, being able to create jobs for dozens of people was pretty cool.

I enjoy communicating with C.C. — it's always fun to hear about her process. She mentioned that she originally wanted to focus on high-tech entrepreneurs — I love that desire for a "tight" theme — but couldn't make the symmetry work out. So I like the fact that she retained crossword symmetry by choosing four titans on industry for the long theme spots (highlighted below). We both lamented the fact that she couldn't make my main man JERRY YANG and YAHOO work. (I didn't know him, but he graduated a few years after me.) What can you do.

Crossing themers will always up the difficulty of a puzzle. C.C. does well to choose the pinwheel layout, which helps a lot with spacing. I really like her use of cheater squares in the very SW and NE — any open 5x5 section gets hard to fill cleanly, and with the J of JEFF BEZOS in place, things are even harder (J in the middle of a word limits one's options). Generally I like to see those 5x5 sections broken up or sectioned off, i.e. with a black square at the intersection of A DARE / OVERFED. This would cause a big cascade of changes in the puzzle skeleton of course, but it sure would be nice to avoid the ODELL / ALTE crossing and the A DARE partial. The rest of the puzzle is pretty nice and clean, so the NE corner stuck out to me.

Business themes aren't going to be for everyone (this one was memorable to me but certainly didn't appeal to everyone), but today's was right in my wheelhouse. Although it would have been fantastic if C.C. could have achieved her goal of being more specific (focusing only on high-tech or dot com entrepreneurs), it still made me smile.

– puzzle by Zhouqin Burnikel

Neat idea today, an elegance to the word TRIANGLES breaking perfectly into three groups of three. There's something so pretty to that. I also liked Dani's natural progression, from a SCALENE (no sides of equal length) to ISOCELES (two sides of equal length) to EQUILATERAL (all three sides of equal length). I really enjoy seeing constructor's background shine through in a puzzle.

It took me a while to figure out what the circled letters were spelling, because I have a propensity to read in a clockwise fashion. I found it to be a bit odd that the TRI / ANG / LES go counterclockwise. Perhaps that's a cultural difference? In any case, I did appreciate the extra layer. Three themers + three triangles + TRI ANG LES fixed into place = a lot of depth.

Speaking of extra layers, I would have loved to see a triangle hinted at in the black squares. If you squint really hard, you can almost see the start of a triangle in the "hill" of black squares (above FLO). It would be very hard to do, especially at the tips of the triangles, but what a neat effect that could be.

Mirror symmetry can be a great tool to have in your arsenal. It's really the only way to pull off today's puzzle, because SCALENE 7 / ISOSCELES 9 / EQUILATERAL 11 can't work with regular crossword symmetry. What a fortuitous coincidence that they're all of odd length, allowing mirror symmetry to work! Another progression that would have worked: SCALENE 7 / RIGHT 5 / EQUILATERAL 11. Hard to leave out the poor right triangle in all its Pythagorean glory, but what can you do.

Quite a few constraints today, and they force a few compromises. Rich Norris at the LAT limits partials to a maximum of two, so A TAB / OH TO / OF A is an unfortunate way to open the puzzle. I understand the constraints of the EQUILATERAL triangle — you can't move around just one of the three vertices — but the SCALENE and ISOSCELES both have enough flexibility that it feels like some of the partials plus the PSAS / ARIS / OST stuff could have been cleaned up by shifting the position of circles.

All in all, an elegant idea with some compromises in execution.

ADDED NOTE: an astute reader, Jean Cranmer, was confused because the bottom triangle didn't look equilateral to her. I smugly sent her definitions of "isosceles" and "equilateral" and then... realized that the sides of the bottom triangle are 5.83 (using the Pythagorean theorem to calculate) whereas the base is 6. I imagine this is the closest Dani could come to an approximate equilateral triangle in a 15x square grid. Thanks for the catch, Jean! (And a good reminder that I'm frequently wrong.)

– puzzle by Daniel Raymon

It wouldn't be Thursday without a twist in our crossword! Alan brings us "literal" interpretations of X BACK Y phrases, with X being reversed. He also goes down to 74 words, which makes the puzzle more later-week-ish, giving it wider-open feel.

These types of puzzles cause me all sorts of internal debate, as to whether I "fix up" the answers for the database or not. Jim and I have all sorts of fun discussions about this. On one hand, manually changing PMUHWHALE to HUMPBACK WHALE makes it much easier for the solver to figure out the trick. It also means that there's less gobbledygook in our database. However, doing so takes away the link between the clue and the answer. After all, HUMPBACK WHALE is NOT a literal interpretation of [Singer in the sea]. As a solver, I applaud the kooky thinking. As a database administrator, I shake my fist at you all. Er, us all.

Interesting layout today, Alan choosing to leave giant swaths of white space in the NE and SW corners. Tough to fill these spaces, especially after you've fixed one long answer crossing through them. I think the SW came out quite nice, smooth except for STN and NIE, two common enough pieces of glue fill. SPARE TIRE next to TRUE LOVE, that's beautiful work. The NE felt crunchier to me, with ATT quasi-duplicating ATTY, a couple of ILA / SAS / ENORM type answers, and the crazy CDEF. Oof. I realize that some solvers will appreciate this entry since it's highly gettable, but it strikes me as most inelegant. Hard to fix once you place an F at the end of NERF. Look where CDEF falls: tightly constrained right between two themers.

I appreciated seeing a few throwbacks, QUAYLE getting a great clue in [Bush successor] (he succeeded Bush in the vice-presidency). And Lisa KUDROW... funny how poorly "Friends" has stood the test of time. Totally fine to me if LEBLANC and KUDROW become verboten to crosswords. But if George and even ESTELLE Costanza ever disappear... Jeff is getting angry!

Finally, beautiful clue for ALLCAPS. Not FLOP SWEAT or NERVOUS TIC sort of stress, but "emphasis" KIND OF STRESS. Good stuff. And the best today, [Something slipped under the counter?], a great repurposing of the "under the counter" slang for something literally slipped under the counter. So appropriate for today's literal theme.

– puzzle by Alan Arbesfeld

Hard to tell that Ian's a sports nut, yeah? There's only COSTAS, THE REBELS, TERRELL Davis... and TEAM SPORTS, of course. Thankfully I like sports too. Even not being a Broncos fan, it wasn't too hard to pull out TERRELL's name. Poor NFL running backs have such a short career lifespan, often topping out before the age of 30. You have to cram a lot of big runs into a small number of years if you're ever going to become famous enough to make the NYT crossword! What, that's not your goal? Ridiculous.

I like the vibe of Ian's puzzles, skewing both more slangy and more sporty than average. I remember the first time one of the kids I work with saying YA HEARD? to me — I give a lot of knowing nods. I'm not sure if YA HEARD will stand the test of time a la WHAT UP G? or go the way of FO SHIZZLE, but for now it seems to be holding up all right.

I always appreciate the variety in Ian's themeless grids. He's always trying out something new, not content to fall into a pattern of using establish grid patterns. This one has so much interconnect it's hard to describe or even grok it all. Beautiful puzzle flow, a feng shui masterpiece, with each section having so many ways to break into it. And look at that intersection of FIREHOSES / RAP GROUPS / RUSSIAN MOB / DR SEUSS / BED HOP etc. That's just one area packed with all sorts of goodness.

I'm a bit mixed on BEDHOP, actually. As much as I enjoy nervously tittering at answers like SHTUP, there's a quaint indirectness with SHTUP, almost like something your grandma might say (at least my grandma, who is awesome). BEDHOP is so direct, so in your face IMO. A matter of taste, of course, but it feels a bit out of place to me in the NYT.

A little more glue than I'm used to in a Livengood themeless, what with both AMO and AMAT (with the same clue, even!) , OON, AZO, ETD, etc., but well worth the price of all the great stuff jam-packed in.

– puzzle by Ian Livengood

Like Ian Livengood yesterday, Barry Silk is another constructor I admire because of the diversity of his gridwork. As you can see on his constructor page, most of his puzzles are themeless. I enjoy scanning through all his thumbnails, in awe at how many grid patterns he's used. A mark of a real themeless professional, someone who can adjust a grid, even demolish it and start from scratch, to suit the specific needs of whatever seed entries he chooses.

Also like Ian, Barry is a sports fan, Philly sports in particular if I'm not mistaken. I've struggled with a few of his constructions in the past, like SIXERS GAME so it was nice to get answers from a variety of fields today. Some astronomy (COSMOS and CARL SAGAN), pop entertainment (PRETZEL LOOP and FOR YOUR LOVE), zippy phrases (GO TO YOUR ROOM and GO ALL OUT), and yes, even some sports (WILD CARD TEAM and HURLER). As a solver, I almost always enjoy variety in a themeless puzzle, so I thought this mix was well-selected.

Yet again like Ian, Barry's puzzles tend to be wide-open, with an easy flow. Notice how each subsection can be entered with multiple different words? In fact, there isn't a single space that can't be entered by at least three entries. That type of open feel really helps me as a solver, giving me multiple shots on goal to crack into any one region. For instance, that SE corner (and NW, similarly) is the most chunked off of any area, but if you can't figure out any of the answers within that corner, you have FOR YOUR LOVE, ODESSAN, and CARL SAGAN to help you break in. Well thought out. It makes construction more difficult to have such wide open flow, and this solver really appreciates the extra effort.

Finally, in the "nice clue saves a glue entry" department: who knew EL AL had a King David Club? What an interesting piece of trivia. Now if only EL AL would get as big as UNITED or SOUTHWEST and make its common letters more crossworthy…

– puzzle by Barry C. Silk

A big welcome back to Eric! After a three-year hiatus from constructing, he comes back with a brain bender of a Sunday. Eric is the author of the Winston Breen series, a neat trilogy about a kid obsessed with puzzles. A fourth grade teacher friend of mine recently sent me a list of her district's "Best Books List" and asked if I could recommend any of them for read-alouds in her class. I chose maybe ten great ones for her… including "The Puzzling World of Winston Breen"!

Back to the puzzle. How to explain this one? If you look at the grid below, you'll see the central entry, A LITTLE GIVE AND TAKE, parsed into two-letter chunks. Each of these chunks are "removed" from one word and then "added" to another, in every case forming valid words or phrases. As an example, the first two letters, AL, can be taken out of FALLOUT to form FLOUT. They then can be added to HERD to form HERALD.

It was pretty hard for me to figure out and remember where each bigram was added. Jim and I discussed "fixing up" the database so that the answers accurately reflected the clues, but that seemed like it would take away the cleverness of the concept. For instance, if you were perusing this puzzle a few years down the road and hit the clue [*Royal messenger] and saw HERALD, you wouldn't think twice. In one sense that's great, as it's not confusing at all. On the other hand, if the answer reads HERD (as it does), it'd make you scratch your head. I personally like it when people scratch their heads. So instead of the fix-ups, here's a listing of the answers where the bigram was added:

• AL: HER(AL)D (53-Across: [Royal messenger])
• IT: GRAV(IT)Y (115-Down: [Newton subject]):
• TL: GEN(TL)E (90-Down: [Gentle])
• EG: L(EG)ION (14-Down: [Great in number])
• IV: TR(IV)IAL (88-Across: [Piddling])
• EA: PA(EA)N (43-Across: [Words of praise])
• ND: BA(ND)IT (86-Down: [Old West robber])
• TA: RO(TA)TE (8-Across: [Turn, as a wheel])
• KE: REBU(KE)S (8-Down: [Upbraids])

Considering that just took me 15 minutes to search out and piece together, I can only imagine how rough it's going to be on some solvers. REBU(KE)S and L(EG)ION in particular had wicked clues. I like a challenge, and really enjoyed figuring out several of the "added-on" words. Those two… whoo! "Upbraids" apparently means "rebukes" (not "does one's hair up"?), who knew? And LEGION felt like a noun to me, but the dictionary says by golly it's an adjective, too. Learn something new every day.

It would have been nice to have the bigrams in order somehow, since all the jumping around broke up the flow of my solve a bit. And it might have even been nicer to not have the central revealer; instead, asking the reader to read across the bigrams in order to read the final reveal (something akin to Patrick Berry's recent puzzle). But overall, it's an innovative idea, one I haven't seen before. I always appreciate the creativity.

– puzzle by Eric Berlin

Debut! And we add another constructor to the youngest constructors list. I think about what I was doing at age 16: watching "Three's Company," eating Cheez Balls, attempting to solve "Zork"... sigh. Pretty darn impressive to see young go-getters spending their time more wisely than I did. Although knowing Mr. ROPER does come in handy with crosswords at times.

Synonym type theme today, with PRIVATE PARTS being a neat revealer. A bit risqué, but again, when I think about my interests when I was 16… ahem. What I really liked about the theme was that PRIVATE PARTS not only defined HIDDEN, SECRET, and CLASSIFIED, but it actually described the whole first two themers! HIDDEN VALLEY is a "private part" (maybe a "private place" more accurately, but let's just go with it). And SECRET GARDEN also fits that pattern!

It was a bit of a disappointment that CLASSIFIED AD wasn't something like CLASSIFIED ROOM. Or CLASSIFIED JACUZZI. If only that were a thing.

A nice construction, especially for a debut. PRIVATE PARTS is a tough revealer to work with, since a 12-letter word forces cramped spacing of the themers. But Matt handles this constraint pretty well. The tricky parts are usually going to be where the themers are closest to each other — note where VALLEY and SECRET are above each other? A five-letter word sitting in between them makes it a tough fill to execute. It's no surprise that I found the roughest parts of the puzzle to be the DOA / EIN north area, and the FIVE AM (seems arbitrary to me — funny how much difference of opinion there is among constructors!) / MAS south.

But look how well the big NW and SE corners came out. Not easy to fill those chunks, and Matt even tosses in a V of AVRIL to spice up a silky smooth section. The SE corner does have ARHAT, which some will carp about. It's certainly not a common word in American usage, but all the crossings are fair, and it's a principle that NYT readers ought to know. Why not learn something new, yeah?

And finally, I appreciate the effort to work in longer quality fill. CREW CUT is a nice find in that little seven-letter space. It does force some compromises with EDUCE and ERG and ENS, but those feel like minor costs to me. It could be argued that the EDUCE / ERG crossing is unfair to beginning solvers… hmm, that's a tough call. I find the ERG pretty esoteric even as a mechanical engineer, and EDUCE is not commonly used in everyday conversation. The constructor always faces tough trade-offs.

Congrats again on the debut!

– puzzle by Matt Fuchs

Fun puzzle from prolific constructor and Will's new assistant, Joel. Today's theme uses crossing answers that mean one thing when put together in one order, and a completely different thing when reversed. I actually didn't notice at first, as I tend to skip or gloss over cross-referenced clues when I solve, but it was a neat moment when I realized what was going on. MAN MADE vs. MADE MAN was my favorite. Such a difference to switch two tiny words.

I highlighted the answers below because it was a little hard for me to keep track of where the themers were. It's impressive to see how much Joel packed in! It would have been nice if every one of the themers had a symmetrically placed theme answer, but that would have been awfully difficult to do. Nearly impossible, I would guess, since finding word pairs that fit this reversing pattern AND had a common letter is hard enough. I did like the quasi-symmetry of MAN / MADE and PAN / CAKE; it would have been nice to have more of that.

I'm amazed at how smooth this grid is. With 1.) seven pairs of theme answers, 2.) short themers (which force longer non-theme fill) and 3.) the fact that each pair of themers cross each other, it's an incredibly difficult construction. Very, very difficult to pull off smoothly, much less with jazz. I like how Joel quasi-sectioned off his grid into nine parts which he could (more or less) independently work on. Although that does cut down on puzzle flow a bit, it's amazing how smooth he got the fill.

And to do so with great long stuff like BANDLEADER and DIRT BIKES and even IN CROWD! Pretty much every piece of long non-theme fill sings. Note how Joel prioritizes two-word phrases? It's much easier to get snazz out of phrases like LASER DISC as opposed to REFILLED.

Joel asked my opinion about this idea a while back, and I was skeptical about all the cross-referencing it would put the solver through. But I was pleasantly surprised at how much fun I had, once I took the time to appreciate each clever pair. Enjoyable experience.

– puzzle by Joel Fagliano

Howard! What a pleasure to see this nicest of guys get his debut. Howard finished 3rd at the 2014 ACPT, performing admirably on a tough final puzzle. I'm not sure I would have completed it using the "B" level clues. And to solve using the "A" level clues in front of hundreds of people... impressive!

This is one of the best revealers I've seen in the "both words can follow X" type of themes. At first I didn't quite understand it. But when you look at BLANK CHECK more like: "___ CHECK" it really sings. Not only is it an apt description of the theme (BODY check, DOUBLE check, etc.) but it's a nod to the fill-in-the-blank nature of many crossword clues. Beautiful.

And to top it off, it's always fun to get some of the constructor's personality and life in the puzzle. I smiled to see BODY check in there, knowing that Howard is a hockey player. (I'm also in the "futile attempt to capture the sporting ability of my youth" stage of life.)

It's an impressive debut. The layout of the themers looks good, and generally the fill is really smooth. There's bits of ATRA and ANS and AS AN, but that's all normal stuff. Most everything else is quite well done, and the addition of a natural-feeling Q and Z in the NE is such a nice touch. Often those Scrabbly letters feel shoehorned in, but that's one beautiful, flowing corner.

TOCCATA in the center sure is a nice little piece of fill, but having a seven-letter word sandwiched in between two grid-spanners can often cause difficulty (hello AMB!). Perhaps shifting a few black squares or even going up to 78 words would have made that easier to fill. And I'm glad Howard mentioned the RRN (random Roman numeral). There's a huge range of opinions among constructors about these. I'm with Howard in that I would take almost anything else besides a RRN, but some people like the fact that they can make a devious but gettable math clue (IV * CLI anyone?).

Constructing has definitely improved my (still sorry) solving skills, because I end up looking up and debating things like which is better: DOREN or DORAS or DORAL? Looking up all these esoteric bits tends to stick in my brain and spill out when I see a clue like [R. J. Reynolds brand]. Scary to think how much constructing might improve an already elite solver's times!

– puzzle by Howard Barkin

★ It took me a long time to figure out what was going on here, but what a neat moment when it clicked. I didn't know the song PAINT IT BLACK, but that didn't take away too much from my solving experience. Once I got over that hump of figuring out the first IT themer, it all fell into place. Great concept.

I couldn't visualize how David put this together! So I reconstructed his puzzle skeleton, which helped me understand much better. It's actually a 72-word grid with crossing themers, a really tough puzzle to pull off. To get this to work, and on a debut puzzle no less... super impressive.

I might have liked the revealer to be placed in the horizontal direction, which is easy to do by "flipping" the puzzle about a line from the NW to the SE (any crossword can be flipped like this and still have all the answers read correctly). For me, it would have been so nice to have the puzzle flipped like this, so that the revealer had been in the usual location. I'm so used to having most revealers running horizontally, located somewhere around the bottom of the puzzle. I'm such a creature of habit.

When I construct, I always look for the most constrained and/or biggest chunk of space I need to fill. Notice how the north and south, with their 6x3 chunks and the themers bordering them, stick out? That's where I'd typically start filling, as they'd be among the hardest parts to fill, if not the hardest. The rest of the puzzle is quite smooth, darn impressive given the 72-word nature of it and the crossing themers, so it was a bit unfortunate that ECARTE reared its ugly head right off the bat, and in the south we get SDI (which Will has mentioned that he's on the verge of not allowing anymore), OSH, TAVI, and the crossing I got wrong, PETER TOSH / SOLANO. I expect to not get a lot of pop music references, but I'm from California and hadn't heard of SOLANO. I don't think I'll be the only solver to have issues there.

All in all, an impressive debut. Great idea and pretty darn good execution.

– puzzle by David Phillips

I laughed at Pete's comment about not knowing Howard Cosell. How could anyone not know Cosell, perhaps the most iconic sportscaster of the 20th century? And then I remembered that Matt Fuchs (who debuted a few days ago) was only alive for about three years of the 20th century. Huh. Thankfully, Pete and I have had some laughs over his constantly changing photos, which usually show him going back in time. I'm demanding a prom photo next.

At first I was a little worried that this puzzle wouldn't have many long entries — puzzles featuring 15's often lean heavily on those grid-spanners for jazz, leaving not much space for other long entries. But Pete gives us a neat optical illusion, making it look like the puzzle has a lot of 15's because of all the open space, but there are actually only two. He includes about the normal number of long entries (a dozen 7+ letter entries is roughly the minimum for themeless puzzle), but all scattered around. I appreciated getting pieces of great long stuff through the puzzle — most themelesses concentrate all their good material in the four corners.

Huge, wide-open spaces give solvers a lot of ways to break into any one section of the puzzle. Today's was almost TOO wide-open, if such a thing exists, as there were very few places I could grab a toehold. Typically there's a small area with a couple of gettable four or five-letter entries that get me going, but there wasn't much that was easy. I struggled to inch into each bit of the puzzle, but with multiple ways to attack any region, it was more than fair. Nice workout.

The wide-open layout does make construction more challenging, because if you change a little section of the puzzle, the effect cascades far and wide. Check out the ON CLOUD NINE and AT ALL COSTS crossing, for example. It causes some difficulty in the east region, with OISE and ASSN and COSA all crammed together. But if you change ON CLOUD NINE to something else in an attempt to smooth out the fill there, so many other entries must also be changed.

Overall, there was a bit too much of the TOD, STDS, ELAM, ASTI sort of fill for my taste, but all the great fill like CHINCHILLA, CRUDE OIL, and OLD GOATS (ahem, not applicable to me or Pete) was fun to uncover.

– puzzle by Peter A. Collins

Nice construction from Julian today, one that features the "stair step" pattern of black squares that we've seen more of these days. You can browse themeless patterns on our Calendar page, hovering over the Fridays and Saturdays to see the thumbnails. Neat way of perusing the archives, yeah? Note that Patrick Berry had a similar type of arrangement on July 4th, Peter Wentz had one on May 24th, Patrick had another one on February 21, etc.

Why? Because this arrangement offers a nice bonus: the potential of three additional snazzy long entries. Many constructors use a basic pattern of four triple-stacked 8's or 9's (one in each corner), which allows for TWELVE marquee answers. But if you choose your central entries wisely, this type of stair step arrangement gets you not only the usual triple-stacks of answers in each corner, but an additional three in the center for a total of FIFTEEN slots. Harder to execute on of course, but it's well worth the extra effort if you can get an extra POSTER CHILD entry worked into your grid.

Julian also adds in a layer of difficulty Kevin Der pointed out in his January 17 puzzle of this year. It's common to see three 8's or 9's stacked atop each other, but to have four of them is something entirely different. Quad-stacking 15's has become a more common sight these days in the NYT crossword, but those usually come with (expected) compromises in the crossings. The real trick of what Julian (and Kevin) attempt, is to pull off quadruple-stacks of longer answers without compromise.

The SE corner is pretty nice, with some beautiful answers. Even the single-word entry, ATOMIZES, sizzles. KNOW BEST felt a bit off to me (KNEW BEST seeming more in-the-language and less partial-like) but still, it worked. SMEW isn't most constructor's preference, but it's a legit entry. And although COSSET isn't a word most people run across in everyday usage, it's definitely legit. It's a reasonable set of trade-offs.

And I quite liked the NW corner, with all its Scrabbly goodness. Julian eased that J in there so smoothly. I did frown a bit at the ST MORITZ / LAZIO crossing (I had a moment wondering whether it was ST MORITS, and LAZIO did nothing to help me figure it out), but ultimately, ST MORITZ ought to be well within an NYT reader's bailiwick.

Smooth solve today, a beautiful grid building off of new developments in xw construction.

– puzzle by Julian Lim

101 puzzles! Impressive. Scan through Randy's constructor page, especially noting how many Sundays he's made. It's hard enough to get a single puzzle into the NYT, even harder to get a Sunday, and he now has 45 of them. Wow!

Fun take on "___ line" phrases today, all reinterpreted as verbal lines from a phone call, a date, from Tom Cruise, etc. I like how Randy stuck to almost all in-the-language phrases. MAY I SEE YOU AGAIN seems arbitrary to me, but all of the nine others are choice entries. And I also like how he stuck with mostly all common "___ line" phrases as clues. The best ones are those that require a jump from clue to entry, i.e. Cruise line (like Holland America) is very different from Cruise line (like from Tom Cruise). Ones like [Story line] and [Telephone line] weren't quite as amusing for me, in that the clues and answers were quasi-related.

Ten themers is a tough task. I appreciate what Randy said regarding the fun being mostly in the theme. Will sometimes runs a Sunday puzzle with less theme density but more juicy fill. More often than not, I tire of those, as they feel like a gigantic themeless puzzle. So I enjoyed uncovering all these ten different "lines."

Always the trade-offs. As much as I like EAT FRESH and ILL GET IT and their respective clues, they sure presented grid difficulties (stacking two theme answers right next to each other will almost always present a challenge). It's hard to say whether or not each of them was "fair." AMOCO, FARFEL, Chris NOTH in the SW. Robert DONAT, Edouard LALO, AGITA in the NE. There's a certain amount of knowledge that a NYT reader is expected to have. These strike me as pushing that fine line (is SEE IF I CARE! is a "fine line"?).

I applaud Randy's use of cheater squares in the very NW and SE. Those regions are nice and smooth, helped out by those two black squares (there's a reason why Rich Norris calls them "helper squares"). A black square either at the D or T of DONAT would at least have opened up some other options to reduce the glue-y entries. Now, I don't advocate tossing in black squares willy-nilly — too many compromise a puzzle's aesthetics — but I personally would like to see them used more, especially on Sunday puzzles, which are so very tough to construct smoothly.

Anyway, a fun outing with a few rough patches. Excellent theme choices and a great title.

– puzzle by Randolph Ross