Last Tuesday’s freestyle solution
Get the PDF here!
Get the PUZ here!
Word count: 70
Mean word length: 5.43
Have something you wanna say? Got a question? Want to do a guest freestyle? Want to collaborate on a freestyle? Want to just say hello? Hit me up by email!
Fellow crossword blogger Evan Birnholz’ 8/29/15 New York Times grid was a partial inspiration for the structure of this grid. I don’t know if he meant it that way, but I had to count it twice because it didn’t look like it had a 72-or-less word count. It didn’t have any showy stacks or ultra-wide-open sections; in a way, though, this kind of grid gives the constructor a better opportunity to get more nice answers into the grid without worrying as much about having to hold stacks or seas of white space together with questionable entries. Of course, I do have some long downs (5, 9, 21, and 24), but this design allows for “staggered” stacks that allow me to put marquee answers next to each other without having to stack every letter next to each other as well.
My favorite thing that I learned from this puzzle is the entry at 53-Across. I experience this phenomenon from time to time, and it’s good to know that there’s a word for it.
As always, I’d like to know, folks… comment is welcome! Come say hello! What did you like? What could I do better?
As always, share this link! Pass it around! New puzzle on Friday!
Hey, don’t go away yet, though! As promised in the last post, below is an explanation of the process I went through to build the last puzzle. So if you haven’t done it yet, get it here (PDF) or here (PUZ), and don’t scroll down quite yet!
I started constructing this grid with a very specific block design. I wanted it to be unusual-looking and eye-catching. However, just as the forces of nature go, so go the forces of words, and the block design had to change.
The original design is below. Notice the grid spanner across the center.
I started to construct in the upper left. In the process of building in that section, I discovered that I could alter the block pattern slightly one way or another (as shown below), depending on how the stacking was going, without altering the flow of my original design.
After a while of playing around with the upper left (and not being satisfied with anything I was coming up with), I had one of those “stupid constructor me” moments. Remember that grid spanner in the middle? THAT is where I should have been starting in the first place! So there I went.
I picked out a 15 from my seed list that I was feeling good about — NO JOB IS TOO SMALL turned out to be the one — and went to work. I looked through my seed list for an 8 that began with N, O, or J or ended with A or L (to seed the southwest or northeast, respectively), and out popped JAMAIS VU. I had my opening! I was off and running in the southwest.
I wasn’t too concerned about the middle chunk yet — it was pretty closed off and would be easy to fill pretty much no matter how the SW or NE corner came out. So next was the entry to the left of JAMAIS VU — I wanted at least two really nice long entries in each quadrant, so I started down the list of nice 8s starting with O. Luckily, I didn’t make it too far down the list before I encountered OXICLEAN; I figured it would work well because it had the opposite consonant/vowel pattern as JAMAIS VU. Well, it worked — not as well as I thought, because the only thing that stacked with OXICLEAN/JAMAIS VU was NET SALES. Not awesome, but not terrible… the crosses worked, too. I mean, with that three tucked in the corner, I had to either use HEB or be forced into LESE, but I could live with that (I obviously chose HEB because I definitely don’t like entries like LESE that can only be clued one way).
The acrosses intersecting that stack fell into place — EXAMS, TIME IN, SCATTE(D or R). ?OMET obviously took a C, and I was left with ??BSIT?. Normally I could have used WEBSITE or JOBSITE, but I couldn’t use JOB (we don’t want a dupe now, do we?) so WEBSITE it was. Here’s how it was looking at this point.
Now, onto the southeast. Of course, the goal is to seed a corner with a nice entry and then work from there, but none of my ten-letter entries were working, for some reason. I just wasn’t getting a satisfying result with any of them. So I backed up and decided to seed the corner with a 6-letter entry (at 42-Down in the grid above) — NEKKID! I was definitely going to use NEKKID, but that presented a problem, namely at the 51-Across position above. I had EK?? or EK???. With EK??, I had four options: EKED, EKES (neither great entries), EKCO (not a great entry, plus 48-Down would be RC??? or DC???, neither of which would be particularly flexible), or EKGS (DG??? yields nothing, and RG??? yields RG III… but then 60-Across would have to start with II, so not a good situation). With EK???, I had EKING (not a good word) and EKE BY (Google didn’t support its use as a phrase that much). So what to do with 51-Across? Only one option was left: move that block over farther to the right and make it EKE OUT! However, that would create a big ugly L shape, which was seriously killing my original vibe I was creating with the black squares. I wasn’t having that. So I decided to pop open that corner, exposing it to the middle (which created better flow in the grid anyway):
(This meant, of course, that I had to take out WEBSITE, but that was not a problem: BOMB SITE fit there perfectly.) Now, it took me a little while to manually fill the southeast, so I helped it along by inserting IN ONE SENSE at 63-Across; it had a lot of common letters, plus it was still a nice phrase. I still wanted to get one debut in the 10-stack, and my manual fill led me to KID LEATHER (which worked beautifully with DAMASK ROSE in the bottom row). The KID LEATHER/IN ONE SENSE/DAMASK ROSE stack also — bonus! — allowed me some flexibility in the remaining space in that corner (45/46/47/51-Down, 45/50/57-Across). Luckily, GOSH NO, BO TREE, and EL OSO all fit in there (ASST is so-so, but not terrible-bad), which was a nice treat. Here’s what it looked at that point:
Now, to the middle, because I felt like I had to build the middle at this point. Why? The CM???? at 33-Across is a limiting factor. It could become a problem if I fill all around it and then get to the middle, and — oops! — it doesn’t work with that constricting pattern. So I went through the list. C MAJOR? Doesn’t work. C MINOR, C MINUS? Blah. CM PUNK? Now, there we go! I decided now to tempt fate — could I now actually get another 6-letter entry from my seed list at 41-Across? Aha… SHAWTY! One problem: 31-Down now had the pattern ?NTH?, which was not good. Ugh, would I really have to use a partial? But then a magical (magical may be a melodramatic word, but I’ll roll with it) observation came to mind: if I pinch in the symmetric block pair, the central down entry would read NTH, and I could miraculously now put an S at the end of SCATTER to make SCATTERS/SODOM! That was pure luck, but it enabled me to keep my two 6-letter gems in the middle and not resort to partials (which I never use anyway) or ugly cheaters (which I normally don’t mind, but in this case would make the middle look REALLY awkward). Plus there were a lot more cluing options for SODOM as opposed to ODOM. Now, here I am with this grid:
Two options now remain at 21-Down: ENLISTEE and ARRESTEE. I didn’t really have a sparkling seed entry 8 letters long that ended in A or L, so I started the manual fill at 35-Across. During the manual fill process, I noticed I had a couple of options with TUNA ROLL at 13-Down, which I liked, so I stopped the manual fill process and restarted it with TUNA ROLL there and went on from that point. THE CONGA and SPANDREL came from that endeavor — SPANDREL is one of those things you know when you see but you don’t know there was a word for. (In case you don’t recall, it’s the space between the arch of a bridge and its deck, or an analogous space above an architectural arch.) I wasn’t crazy about ATTS, of course, but I’ve seen that abbreviation in headings for football box scores, so I rolled with it.
Between ARRESTEE and the stack at 12/13/14-Down, the only thing that fit at 26-Down was LOAMY, which left ???RLORD at 25-Across. It was either OVERLORD or DEAR LORD, and the latter was the obvious choice: DEAR LORD is more colorful, plus OVERLORD would force a terminal V at 9-Down, which is severely limiting. So this is the grid at this point:
Two things were constraining my final corner to construct: the terminal A at 21-Across, and the terminal B at 23-Across. Now, the ????B pattern normally isn’t too constricting, but when you’re building a stack from the 10-Down/20-Across area and working around to that 23-Across corner, that terminal B can shut off a promising alley very quickly. So I started the manual fill at STREAK at 10-Down — the S-T-R at the end of the 10-stack would keep my possibilities open as much as possible. Then came a lot of manual fill. I mean, a LOT. On my iPad app on which I was constructing the puzzle, what I do when I come upon a fill that I find intriguing during the process is to save a screenshot of it and move on. I don’t just save the first fill I come upon… who knows, there could be a better fill that comes up that I’d completely miss! Similarly to the southwest, I hit upon HOME PLANET/AMEN SISTER (AMEN SISTER, surprisingly, was a debut!) at 15/17-Across (after I’d screenshotted maybe three or four other fills), on top of THE SEA at 20-Across, as a workable stack. But, unlike in the southwest, I had several different options for 1-Across. I had SPORTS BARS/CARS, STATE FLAGS, BLACK/WHITE FLAGS, and SCOTT ADAMS. (It’s good to have options!) SCOTT ADAMS was the one I found to work best for the 19/23/27-Across segment (plus I even got PLAN B to work, which was my best possible option at 23-Across!) and that was that! The final product, for your reference, is below.