I wrote some puzzles

[dr_whom]

Here I've got some comments on the specific puzzles I wrote, or contributed to the writing of. That's actually not really a very well-defined category per se, inasmuch as as lead puzzle editor I had a hand of some kind in a large majority of the puzzles that made it into the Hunt. So what I'm going to be talking about here is the puzzles that have my name listed as author or coauthor. As I'll mention below, that's not a very principled category; there are certainly authors who put my name on their puzzle as a coauthor for contributions that I would have regarded as basically just editorial. But it has the advantage of being a well-defined criterion, at any rate. This list is still very long, as it turns out!

• Question Mark. Oddly, both this year and in 2006 I was the author of in some sense the "first" puzzle in the Mystery Hunt. (Question Mark was one of seven puzzles unlocked simultaneously at the beginning of the Hunt, but it was first in the sense that it was first on the list and extracted the first letter of the meta. 182.2 Smoots was one of two puzzles unlocked at the beginning of SPIES, and it was the one most teams solved first because it was the easier one.) I didn't do that on purpose; it's just that at the time when I needed an answer to assign to this puzzle, OYSTER was the only unclaimed answer that didn't have an A in it, and A is (almost?) impossible to produce given my answer extraction. (I think okosut has a slightly better accomplishment, in that he wrote the first puzzles to get testsolved in both this Hunt and SPIES.) Anyhow, the origin of this puzzle idea actually comes from the editorial discussion on Chris's They Might Be Giants puzzle, Study Materials. We were discussing how solvers would get the aha on Study Materials, and Chris said that he thought having a clue whose answer was JAMES KNOX POLK would be enough to point people towards TMBG. I said there were lots of other facts about Polk that could form the basis of a puzzle—he's the subject of a Thurber essay, his name contains JKLMNOP.... And then I was like, wait a sec, that last one sounds like a pretty good puzzle theme! Thus the idea for Question Mark was born, although I didn't end up using JAMES KNOX POLK in it, which I kind of regret. The title Question Mark was suggested by David S and is perfect, in that it both contains QRSTU itself and gets the Mario round, where the puzzles are represented by little ?-block icons, off on the right foot. It also has MARTIN LUTHER KING JR in it, who I think should appear once or twice in every Mystery Hunt.
  
  
• Keyboard Cat. This puzzle was originally proposed by redcat9 and Reid as just 'give solvers a sequence of letters; they have to realize that the shapes the letters form on the QWERTY keyboard traces out the answer'. So we gave that to testsolvers, and they were like 'cool answer extraction, but where's the puzzle?' So we talked for a while about what sort of puzzle might produce this answer extraction, and eventually came up with this "typing game with spelling correction" concept as something that would be more or less appropriate, and I wrote the list of words used in the game. (It turns out to be quite difficult to come up with a list of 76 words that can each have a plausible single-letter spelling error on a specified letter!) What we weren't expecting was how much people would like this bare-bones typing game—testsolvers were literally arguing over who would get the chance to try it next, and preliminary post-Hunt data shows it to have been the most popular puzzle in the Mario rounds! The authors actually thought it seemed pretty tedious.
  
  
• Rivalries. This puzzle was inspired by a bunch of Sporcle quizzes using the same general "antonyms" gimmick—in fact, I see that there's actually one on the exact same topic using some of the same clues, though I wrote Rivalries before it appeared. I apologize for the erratum on this one during the Hunt; the initial grid I made left the positions of CAVALIERS and CANADIENS interchangeable, and I did fix it even before it went to testsolving; but somehow what I submitted for deployment on the Hunt server itself used the original version. Anyhow, for those of you who are interested in such things, the solution file for this puzzle includes a rare naturally-occurring example of a three-deep center-embedding.
  
  
• Pipe Dream 2. I really don't understand what went wrong with this puzzle. It got through testsolving fairly easily—not that testsolvers thought it was an easy puzzle by any means, but they solved it smoothly, and liked it a great deal, and only suggested minor changes to some of the clues; it's not like it had to go through many cycles of revision and re-testing before it hit testers who found it halfway palatable. And yet during the Hunt it only got a few solves, and was mentioned very frequently on the "least favorite puzzles" survey. I guess I just don't know why the testsolving audience was so atypical, I suppose. This was a real collaboration from the writing perspective: I suggested 'let's have a new Pipe Dream puzzle', Lanthe proposed the warp-zone mechanic, Seth built the grid, I tweaked the grid a bit to make the pipes symmetric and wrote the fill, and Seth wrote the clues. It was my first time writing a crossword grid fill and I was pretty proud of it, but now I dunno.
  
  
• Everybody's Got to Be Somewhere. The point of this puzzle was to write a puzzle whose answer was WINTHROP PAROO. The original answer extraction was going to be something much more complicated—the cluephrase was going to be something like JUST EAST OF NORMA CASSIDY, and since Norma Cassidy sings "Chicago, Illinois", and just east of Chicago, Illinois is Gary, Indiana, the answer to the puzzle would be the character who sings "Gary, Indiana"—but testsolvers stalled out on that and I was persuaded to replace it with just a link pointing to a Google map of Gary, Indiana. (I also didn't get to use the fact that WINTHROP is also a placename, and PAROO sounds like one, which is kind of disappointing.) I'm very glad I got to use "14 Dwight Ave., Natick, Massachusetts", though, which is a great song and so appropriate for the puzzle it's ridiculous.
  
  
• Unfair Cryptogram. I'm not sure it's quite fair for my name to be on this as a coauthor, really—David S and I both had the idea of "cryptogram with letters represented by character strings of varying lengths" independently, but he really ran with it, had the idea of using cheat codes, and actually wrote the puzzle. We put it at the end of the Mario round to sort of help transition the theme of the Hunt from Mario in particular into video games in general, and because ASHAMED was a great answer for it.
  
  
• The Mega Man world structure. As I'm sure a lot of you have heard by now, I've been wanting to do a Mega Man theme since 2002 if not earlier. The games seem really well-set-up for an innovative puzzle structure: several rounds, which you can tackle in any order; for finishing each you get a weapon which helps you defeat one other round's robot, but you have to figure out which. My original suggestion—each round's meta answer is a clue to the solving method of some other meta, which would have been really analogous to the games—was rejected as unworkable because of the obvious rich-get-richer problem: the teams that solve metas fast anyway don't need to be rewarded by having the other metas made even easier. Seth's proposal of having the meta answers be wordplay operations that you can use on the other robot's names saved this concept. By the way, did anyone notice during the Hunt that three of the six Mega Man rounds had their puzzles released out of order? The order in which the Bebop, Blackberry, and Craps puzzles are listed is the order they were supposed to be released in, but in fact they were (I think) just released in alphabetical order of their answers. This was a little bit unfortunate in some cases: it meant that some of the harder or longer puzzles, like One More Try, Favourites, and Basic Knowledge, which were supposed to be unlocked early to give teams more time to work on them, were unlocked fairly late in the round and therefore easier to skip.
  
  
• One More Try. A puzzle that a lot of teams got stuck on. Apparently I'm listed as a coauthor on it; my main contribution was helping with the clue-writing. I was never quite sure that a crossword was the right venue for the two-hit-wonders concept; but we were a little short on crossword-style puzzles at the time. I think it would have probably been better if it had had a shorter answer—at least, we probably would have been able to use a smaller grid and/or a better fill. But there was something to be said for putting this puzzle in the Bebop round, and leaving Powder Monkey on the answer DRUGGY (the most convenient alternative answer for this one).
  
  
• Meta testing! The motivation behind this puzzle was fundamentally "an answer like BACKSOLVER deserves a puzzle that fits it." It had other functions as well, though: it allowed us to use a few of our many abandoned meta ideas that hadn't been selected as metas or reworked into standalone puzzles; and putting this puzzle in World 2 also allowed us to hit teams over the head with "The way this Hunt actually works is that there are a bunch of rounds based on different video games." Also, of course, it was meant to be a cute in-jokey peek behind the scenes, and a way of showing teams that have never written the Hunt what meta testsolving actually looks like. The Mario "meta" here was based on a proposal by noahspuzzlelj for World 1-1 that we dropped when we decided to go with tigupine's mushroom-flower-star proposal. The Frogger one is based on Ricky's Bio Man proposal that we didn't use because it wasn't sufficiently Bio-themed; after the way our actual Bio meta turned out, maybe we should have gone with this one after all. The Portal one (an idea Julian put forward for an actual proposed Portal round that never got assembled into a concrete minihunt structure) could probably never work in this form as a real meta: if these were real puzzles, there'd be no good reason to scrutinize the flavortext in trying to solve the meta, but as a "meta" presented out of context, the fact that the flavortext is given at all means that there must be something hidden in it. (I had a ball trying to write plausible GLaDOS flavortext.) The Street Fighter meta and supermeta here was a full gameworld proposal by me, leech, and David S. It was eventually deemed too lightweight to use as an actual gameworld, though we considered throwing it in as a round of micro-puzzles if it turned out we had extra time and more ideas than we needed; that didn't happen, but I'm glad it found a home here. (Easter egg: can you guess the correct answers for Guile and Ken and see their victory speeches?) The Q*Bert "meta" was the only one written specifically for this puzzle; it was meant to be fairly easy, since it extracts the ER of BACKSOLVER, and testsolvers found it pretty easy. I gather that actual teams psyched themselves out on it a bit, expecting it to be harder than it was? In any event, I find myself wondering if Hunt teams found the answer extraction aha for this whole puzzle easier than testsolvers did; actual testsolvers, of course, had already spent weeks testsolving metas with partial data and were accustomed to ignoring the fact that they had missing answers, while Hunt solvers weren't primed in quite the same way. Anyhow, this was voted the most popular puzzle in the Mega Man world, and I'm pretty proud of it.
  
  
• The Eternal Struggle. When I originally proposed this puzzle concept it was much less interesting: you were going to be given the starting configuration of cells, and the letters for each timestep were given shuffled, and you'd more or less solve it as a dropquote where you know exactly which set of grid squares each set of letters goes into, but not in which order. David S suggested making solvers actually try and reconstruct the Life game from the t=14 position, and it made the puzzle a whole lot more interesting. (Ricky noted at one point that the puzzle could be solved by running the t=14 position forward to see its behavior, and then Googling for what known Life starting positions stabilize after about 5000 generations. Did anyone do that?) I'm also really proud of this puzzle's thematic unity: it fits the Bio Man round, obviously, and the answer CREEPING INFECTION makes sense for the behavior of the cells in a Game-of-Life puzzle, and moreover the puzzle's title makes independent sense with both the puzzle itself and the answer extraction.
  
  
• Good Vibrations. This puzzle is really Noah and Malia's; I was added as a coauthor at the end because they needed someone to do the audio editing for it. The puzzle isn't really my style, but I had a great time trying to find somewhat suggestive excerpts of the songs to fit the sex-toys theme. (I'm particularly proud of that little moan of ecstasy from "Earth Angel".) Chris suggested the answer for this one, and midnight_sidhe wrote the flavortext, and overall this puzzle turned out to have an unexpectedly nice thematic unity.
  
  
• The Least You Could Do Is Phone Me. Something I wanted to do for both this Hunt and SPIES was to write a really good linguistics puzzle. I didn't manage it either time, but this one's okay and kind of linguisticky. EDIT, January 2012. Apparently I do have more to say about this puzzle than just this, and I said it in part of my writeup for the 2012 Hunt.
  
  
• Bebop Man meta. This started out as a concept by Ricky which involved finding the letters ABCDEFG in the puzzle answers instead of DO RE MI etc., but there were a couple problems with it and even testers who solved it weren't really satisfied by it, so I reworked it into this version. This was the last meta finalized, and at one point we had two other Bebop Man meta ideas floating around, but this one made it out of testsolving first. Noah and Ari had a Bebop proposal that was in some respects a lot more elegant than this one, but no one who didn't already know the answer was able to solve it in the available time.
  
  
• Stagecraft Man meta. This turned out to be a real group writing effort that I think produced a really slick meta. There were actually two potential themes for this round: Stagecraft Man and Cognate Man were both proposed as potential victims of the DNA Destroyer and thematically appropriate wielders of the Word Sword. And although I would have loved to have an etymology-themed round and meta (maybe to make up for my failed cognates puzzle from SPIES), "Stagecraft Man" sounded like a more plausible name for a Mega Man robot, and I thought of a meta idea based on Shakespeare character names before I thought of one that would be appropriate for Cognate Man. While I was trying to come up with an appropriate answer extraction for my character-names meta idea, Noah proposed instead a meta based on the phrase WHAT DO YOU READ, MY LORD, and then Ricky had the brilliant idea for how we could combine the two. I really like this because unlike a lot of metas it's not just a case of extracting a few letters from each puzzle answer and then putting them together in the right order; and it uses Shakespeare in two different ways, which I find quite elegant. (And no one complained about me wanting to use "rapier wit" in the flavortext, which kind of gives away the answer if you squint at it right.)
  
  
• Mega Man supermeta. I'm really grateful to Seth for coming up with a sensible way for the Mega Man structure to work and produce a supermeta. My main solo writing contribution to this supermeta was just drawing the grid that's used for answer extraction, but I was part of the long meeting at David P & Tracy's apartment in April where we brainstormed possible robot and weapon names, and it was really exciting to watch as they all finally started to click as a coherent set. That was the best puzzle-writing moment of the year for me, I think.
  
  
• Forsaken Fortress. This is a lot like One More Try, in that it's a crossword that's more constrained than it looks, which leads to some so-so fill, which I have a coauthor credit on because I helped write the clues. The way we went about writing it is by brainstorming a bunch of words that could be clued relatively naturally with clues that had a triple D in them, and then figuring we could write iambic-pentameter or country-acrostic clues for the ones we needed. The results are perhaps predictable—the 3D clues are on average a lot more natural-sounding than the other key clues—but I think "a backless stool or storage cabinet" and "Hawaiian high school graduation gift" turned out sounding pretty okay, and "a polynomial of odd degree" is a twofer. (The reason it's diagramless, by the way, since some people were asking, is because answer extraction is by looking at the intersection points of the key clues, and having it be diagramless makes you have to do at least a little work to figure out where those key clues intersect.)
  
  
• Triforce of Fellowship meta. Noah's original proposal for this was a meta based on the actors from the LotR films which turned out to be way too constrained to write. Since David had recently suggested that we didn't have enough pure metas based on wordplay, I fooled around with anagrams for a while and this meta is what I came up with.
  
  
• Triforce of Holiness meta. A third viable set of nine for the Zelda metas was surprisingly elusive. Noah's proposal for a Triforce of Justice (i.e., the Supreme Court justices) was likewise too constrained to write; and the idea for the nine-circles-of-Hell meta turned out to have already been done. Around when this was going on on the meta brainstorming list, Julian happened to be in the Philly area; we met for lunch one day and we started talking about alternative sets of nine, and whether it might be possible to make a nine-orders-of-angels meta work. We discussed a bunch of potential answer words—is CANADA fair for Dominion? Can we think of any cherubim other than PROGINOSKES?—and over lunch it looked like this one was going to be too constrained also, but after I got home I was able to put together most of the letters of CREATURES, and the meta list was able to fill in the rest.
  
  
• The Civilization world structure. I was (via Skype) in on the meeting where the Civ world concept was originally worked out, but my main contribution here was the tech names. It was quite unfortunate that, in a minihunt with a tree-based structure, five out of six metas depended on having a preexisting linear ordering for the puzzles that fed them (and potentially different orders from meta to meta, for ones that shared puzzles)! I suggested using various ways of sorting the puzzles by technology name for whatever metas needed that, and then since I was more or less the only person who both knew what all the puzzles were and cared, I ended up making up most or all of the tech names. I tried to come up with things that (a) were more or less relevant to the puzzles they were attached to, (b) more or less made sense in terms of the tech-tree innovation structure (e.g., Epic Poetry as a prerequisite for Literacy, Legal System as a prerequisite for Oratory), and (c) fit whatever requirements of alphabetization or word length or whatever the corresponding meta orderings required. That's a bunch of constraints, and they're satisfied... pretty unevenly, I guess.
  
  
• Racking Your Brains. I'm not sure I really deserve a coauthor credit here; Jane wrote the puzzle and I basically fiddled with it around the edges to clean up some of the clues and grid, but the excellent (and highly constrained) construction here is all hers.
  
  
• Hints, with a bit of love! Hm, I seem to have written two puzzles with exclamation points in the titles. Anyhow, this puzzle was Noah's idea, but apparently I spend a lot more time brainstorming &lit clues than he does and ended up writing most of it myself. It went through a bunch of revisions and improvements and ended up really solid; I think I've gotten more personal compliments from solvers on this puzzle than any other one I've written. We just got really lucky that there was an available answer word I could write an &lit clue for; otherwise the answer extraction would have had to be a whole lot clunkier.
  
  
• Soooo Cute! Wait, am I an author on all the puzzles with exclamation points in the titles? I didn't have much of an authorial contribution to this one, actually; David wrote the puzzle, and my contribution was just photographing the half of the animals in it that belong to midnight_sidhe. They are, in fact, soooo cute.
  
  
• The Doors of Cambridge. This is actually a puzzle I wanted to do for SPIES but didn't. It's motivated in part by the feeling that there ought to be at least one puzzle in the Hunt that makes you get off your behind and leave the MIT campus, in part by the fact that Metaphysical has a fairly high density of Harvard alums (similarly, I'm not too surprised that Codex solved this puzzle first), and in part by the fact that I noticed at one point that some white detached house facing Mem Drive is labeled as K-Entry of Winthrop House for some reason. I got extremely lucky in writing the cluephrase: all but one of the letters in RANK OF KFC HEAD is an entryway that's visible from the street or some other publicly accessible location, and in fact it includes all but one of the buildings that have such entryways at all, as far as I can tell. Thanks to leech for the photography; the brilliant flavortext is due to David S.
  
  
• Literary Collection. I don't have much to say about this one, other than that I'd been wanting to write an LC Classification puzzle since SPIES as well.
  
  
• Part of Speech. I'm very pleased with the title of this one. This is a puzzle idea that was proposed by bluefaith which I took over when she didn't have time to write it. I have an unusually comprehensive book on old-school sentence-diagramming at my parents' house, but sadly I didn't have access to it while writing this puzzle; on the other hand, I guess that means that I didn't use any diagramming conventions in writing it that solvers didn't have access to via Google when solving it, so I guess that's something. This puzzle is the other appearance of Martin Luther King, Jr. in this Hunt.
  
  
• The Da Vinci's Workshop meta. My contribution here was the twisty lines that connect the boxes that the tech names go in to the boxes that the puzzle answers go in, not the meta itself. This was a tricky meta to give the "ordering" for because one of the seven puzzle answers has to appear three times in the shell, so just a linear ordering of the tech names wasn't sufficient. The twisty lines weren't supposed to be particularly challenging, and I think they're more or less thematic to the meta in that they kind of look like wires.
  
  
• The St Andrews Links meta. I didn't come up with the main concept for this meta, but I do think I thought of giving arbitrary par scores to the puzzles in order to be able to produce arbitrary indices. Then I spent a while thinking of words that clue golf scores that might be able to fit into some of the other Civ metas.
  
  
• Charitable Collection. I agree with the people who said that this was a little late in the Hunt to have the scavenger hunt puzzle, but in my opinion if we were having a Katamari round the scavenger hunt really had to go in it. (Anyhow, last year the scavenger hunt was actually in endgame!) I was able to snag it a pretty thematic answer, too, which I'm glad about. Thanks to jcberk and Tracy for helping me brainstorm qualities; they did most of the actual writing of the list.
  
  
• Charges. This is a puzzle idea that was proposed by yoz; he brainstormed a bunch of logos that might fit the puzzle well and then didn't have time to write the actual puzzle, so I took over and had a great time writing it. I made up "a cravat nowed" for the Playboy Bunny's bow tie, but the rest of it should be more-or-less legitimate terminology. It's kind of weird how large the percentage of car companies here is.
  
  
• Heard but Not Seen. This puzzle has its origin, very simply, in a time I was listening to NPR while driving home from work and heard a report by Doualy Xaykaothao, and said "Doualy Xaykaothao! What an awesome name!" And then I said it again once I got home and looked up how it was spelled. Apparently "NPR reporters have interesting names" is a bit of an internet meme of some kind, and it seemed like a great topic for a soundalikes puzzle. I have no idea how this puzzle went over, though—nobody's commented on it that I'm aware of.
  
  
• The Baddest Man. I really like the title of this one too. This puzzle was a floater, not assigned to a particular answer until a couple days before the Hunt. It's kind of a shame that this ended up so late in the Hunt, because it means that not too many teams forward-solved it and actually wrote an Encyclopedia Brown story for us. A few did write really good ones, though, and they'll go up in the archive once everything's moved over to mit.edu/puzzle (or sooner).
  
  
• Plotlines. This is a puzzle where Cindy's skills as "harasser" really came in handy. Daniel proposed this puzzle idea, everyone liked it, and then nothing happened with it for three months or more. We brought Cindy in, and she offered to make the diagrams if the rest of us would sketch out rough drafts of some of them. (I did the sketches of Watchmen, A Midsummer Night's Dream, and Wizard of Oz.) I'm very glad that Cindy actually got this puzzle written, since it turned out to be substantially the most popular puzzle in the Hunt: Last I checked, it had more votes on the "favorite puzzles" survey than any other puzzle, even though it's in the Katamari round, which fewer people voted on at all than any other round.

I think those are all the puzzles that have my name on them as an author. In a future post, time permitting, I'll have some comments on puzzles I wasn't an author on, including some of the ones that turned out to be problem puzzles.