Ask Dr. Sudoku #13 – Puzzle Hunting

A few weeks ago Giovanni P. asked what our visitors might think about “other puzzles.” His question was specific to word puzzles, but this week I put the question to the test when I posted our first “Puzzle Hunt” puzzle. The Monday puzzle was not announced as such. Instead it was meant to just look like an April Fools joke. But it already has the most FAVEs of any puzzle on the site. So what are “Puzzle Hunt” puzzles and what was really going on with that word search?

Most Puzzle Hunts — sometimes called Treasure Hunts but the overlap is not perfect — involve solving puzzles to find a prize. They are an adventure where the puzzles tell you where to go next. Usually each puzzle solves to a word or phrase. In some hunts, the answer could actually be the next location like “FOURTH AND MAIN”, an intersection of streets. In others, there might be a different mechanism to take a more regular answer word like “IVORY” and turn it into spot 12 on a map. Perhaps there is a crossword clue for site #12 on the map that says “Key ingredient” and that fits pretty well when you think of a piano key. If crossword clues aren’t for you, perhaps you give solvers a list of instructions to perform on the answer words. Maybe something like “take the longest intact roman numeral in the answer and multiply it by the number of other letters” to get 12 as well from IVORY. Make a map with lots of numbered places, only a subset of which will be visited, and you have the framework of a Puzzle Hunt.

But this is talking about the logistics of a Hunt and not what a Hunt puzzle is like. While you can make a hunt out of simple riddles and shaded cells in crossword puzzles — and this is exactly what many “Treasure Hunts” do — the puzzle community in places like San Francisco and Seattle has a very different view of what a Puzzle Hunt puzzle is like. For that definition, I want to take something Mike Develin contributed to a Bay Area puzzle hunt mailing list last year that fits perfectly:

Imagine a word search.

Now imagine you aren’t told what words to look for.

Now imagine you aren’t told it’s a word search.

Now imagine it isn’t a word search.

The essential quality of a Puzzle Hunt puzzle is that it stretches your mind in unexpected ways. It may look like one thing but probably solves as something else. Most of the time it will not replicate a puzzle experience you’ve had before (but there are only so many ideas in the world so this, in practice, is hard to achieve). Puzzle Hunt puzzles almost always present you with a lot of data, be it a grid of letters, a set of images or sounds or smells, or even something in the environment around you at the location where you find the puzzle. Transforming the data down to a single word or phrase is the basic goal of a Puzzle Hunt puzzle, but what that involves is rarely spelled out clearly at the start.

See, Puzzle Hunt puzzles almost never come with instructions and certainly not with clear instructions. Sometimes they have vague hints or partial messages along the way but sometimes they have nothing at all. Figuring out what to do is often a large part of the fun. The AHA moment — when an indecipherable image suddenly starts to show some reason or when a string of gobbledygook starts to actually resemble words given a new frame of thinking — is the joy that these puzzles frequently provide. Puzzle Hunting is my favorite form of collaborative puzzle solving by far as it often requires quite different perspectives to see everything you need to in a puzzle.

I play in about a half dozen puzzle hunts every year. Some are only hosted in one place so those not in the Bay Area might miss an event like BANG (I led the construction of BANG 23 which featured a mix of puzzles including logic puzzles that actually built the map). One recent hunt, DASH or “Different Area, Same Hunt”, however aims to host the same event in over a dozen cities. While this format loses something in environment specific clues which are some of my favorite puzzle hunt clues, it gains a lot by making puzzles of this style accessible in many more places. This year’s DASH is even going international with a London hosting in addition to many in the US. And not all Puzzle Hunts are played on foot. Some take place in vans over the weekend. Others involve much less travel like the MIT Mystery Hunt that is more of a “conference room” game where you have a ton of puzzles to solve but it is not as much about getting from point to point but about cracking “meta-puzzles” built out of many other puzzles. Winning the Mystery Hunt means writing the next one; I made the “mistake” of winning the Mystery Hunt the first two times I played with a team of about 25 which is actually small for the MIT Mystery Hunt (must puzzle hunts have teams of about 4-6).

Puzzle Hunts of all sizes allow for a lot of creativity on the part of the puzzle constructor; they are where we can break the rules most. A good number of the puzzles I’d judge as my own favorites amongst the thousands I’ve written are those I wrote for puzzle hunts: “Rewriting the Record Books” from the 2007 Mystery Hunt is near the top of my list although it is insanely difficult and requires you know a lot about American sports history. You are allowed to require certain knowledge from your solvers with puzzle hunt puzzles which separates the style a fair bit from most of the logic puzzles here. Slightly more approachable but still difficult is the “KenWord” puzzle from BANG 23. This puzzle is from a Hunt where the main villain imagines a New York Times puzzle page where the Crossword and KenKen have come together to form the ultimate single puzzle. How this helps him take over the world is unclear, but a one-of-a-kind mix of these types gives a satisfying and similarly one-of-a-kind experience.

So that discussion brings us to our first Puzzle Hunt-style puzzle at Grandmaster Puzzles, an easy one as these things go.

Last Monday, when I was supposed to give you a Sudoku based on my (unannounced but predictable) schedule, I gave what looked like a Word Search instead. One which would, according to the Rules and Info, “have a unique solution that can be reached by observation alone,” a slight play on our normal formula. After identifying the grandmasters in standard fashion, the solvers would see this.

Notice that the leftover letters in order spell “UNUSEDAREASUDOKUDSRNDKKEOEKA”. Which can look like a lot of gobbledygook but hopefully “SUDOKU” is a key word to recognize to eventually give the partial message “UNUSED ARE A SUDOKU” at the start and then still just gobbledygook at the end. Here the solver must take a small leap. While the word search is 11×11, the remaining letters are part of a 9×9 group. And there are 9 unique letters. So they can be placed on a standard classic sudoku grid and solved as a more expected Monday puzzle would be.

Finding a unique solution is very strong confirmation you’ve done the right thing, but in true Puzzle Hunt fashion you still need to find a single word or phrase from this puzzle. Which is kind of a word search like where the puzzle started. There are about 20 good hiding spots in a sudoku — the rows, the columns, and the main diagonals — and the last two are the best most of the time as you can repeat letters in an answer, hiding it from people who can anagram sets of letters. The main UL-LR diagonal hides “RUNAROUND” on it, which is a kind of puzzle found at the MIT Mystery Hunt. It is also how you might feel after going through so many steps without instructions, except that the process of puzzle solving can be fun whatever the reason you are doing all these assorted tasks. So I gave you the RUNAROUND on Monday, and for that I hope I need not apologize.

I promise to have more puzzle hunt-y puzzles here in the future. I may tell you when they are coming. I may not. But I intend to have all the puzzles types I enjoy here eventually. From the good reception I got on Monday, it seems a lot of you like unexpected surprises too.

  • Avatar photo Grant Fikes says:

    I loved this puzzle a lot. Very clever.

    As much as I like aha moments like this, though, and as much as I enjoy constructing word puzzles occasionally, with my many books of crosswords and logic puzzles competing, it’s always the pure logic puzzles that get solved the most consistently, because when solving them on the go, I don’t need to drain my iPhone battery on Internet research when searching for clues I don’t know offhand. (On a tangential note, my favorite Hunt-style puzzles are the ones that derive answers from pure logic puzzles, such as this one, or the amazing “Portals” from 2013, because they combine the victory of an “aha” with the pure logic which is my greatest strength.)

  • Boomer says:

    Yes, but your sudoku does not even exhibit rotational symmetry! Shoddy work Craig, shoddy work.

    (Kidding of course. I especially like that your original grid was larger than 9×9, preventing people from jumping to sudoku immediately.)

    • Avatar photo drsudoku says:

      While I know you are kidding, I did take a first pass at making it symmetric. But having the leftover letters “read” properly, leaving RUNAROUND or something else on the diagonal primarily in unclued cells, and having the eliminated spots be a sensible word search result (part of long enough lines of similarly marked off cells), made it all quite unlikely. So within 5 minutes of sketching the look of the second stage out, my better instincts said symmetry was one constraint too many.

      Also, my name is Thomas. But there is certainly a talented Craig in puzzle construction that we’ll be including in future titles and maybe online. I wonder if Craig could have gotten symmetry out of this?

      • Boomer says:

        Oh, I’m embarrassed. Sorry Thomas… I got the link from a Craig, possibly the same one you’re talking about, and assumed that he was citing his own work. Well, you know what they say about assuming!
        You’d think I would know better, especially on a logic puzzle blog!

        • Craig K says:

          I’m going to assume I’m the person being referred to in both cases.

          I only wish this puzzle was one of mine. My post to the April 1st puzzle was kept deliberately vague to reduce the chance of inadvertently spoiling what was going on, and apparently it was a little too vague in the end. Regardless of this, I’m gald you followed the link and enjoyed Thomas’s puzzle.

          In regards to the what-could-I-have-done question, I don’t believe I would have necessarily gotten a symmetrical sudoku to work out for this. I am not a sufficiently skilled solver of sudoku to really excel at making ordinary sudoku puzzles (though I have constructed sudoku variants in the past and undoubtedly will again). I may not have wanted to construct a symmetrical sudoku if I were Thomas, however, because that might have been a visual cue that something funny was going on in the word search.

  • hagriddler says:

    I like this kind of puzzles very much.
    Somewhat along those lines are the puzzles at my site I made with 2 friends. Each puzzle series contains a number of puzzles without specific instructions. You just need to figure out what to do. You can only advance to the next puzzle once you finish the previous one.
    Although we are most proud of the dutch version, the english version is a nice derivative.
    I don’t know if self advertising is allowed here, please remove this post.

    • Avatar photo Grant Fikes says:

      You mean the kind of puzzles popularized by not_pr0n? I’m not into those myself. I’d rather have my puzzles separate from each other to solve (or skip) in any order I please. If there were a not_pr0n clone with nothing but abstract logic puzzles, though, that might be rewarding for me, because I’d have the confidence of knowing I can do it.

      As for self-advertising, I’m not a comment moderator myself, but I think this one can be allowed. Seems very relevant to the discussion.

    • hagriddler says:

      I meant of course : remove this post if it violates your rules on this blog.

      • hagriddler says:

        Not quite not_pr0n but kind of. If I remember correctly that showed only a picture and you had to figure out the url to the next page.
        In the Netherlands there was a puzzle series called Mzsteriax which was quite popular a few years ago. After playing those series we tried to make a few puzzles series of our own. I’m happy to say those where also very popular in Holland and Belgium a few years ago.

  • Para says:

    I’ve participated in the Australian SUMS and CiSRA puzzle competitions. But a lot of time there is a language or cultural bias/barrier that is hard to overcome for me. The ones I manage to solve I usually like. But sometimes I find the connections too random or weird and depend too much on getting the right terms for vague clues. It’s fun though to work together with others on such puzzles, but by myself I really don’t enjoy them that much. I am always impressed though with some things people come up with.

  • James M says:

    I think the main difference between this website and most puzzle hunts is that here we have people from all over the world, and some will not be so strong at English, but that normally doesn’t matter because logic puzzles are universally solveable. So the challenge is to compose hunty puzzles that everyone has a fair shot at…

    • Avatar photo drsudoku says:

      I definitely intend to have “fair” puzzles, but I can’t guarantee 100% of the time that they will all be language- or culture-neutral. There are simply some puzzle types or even some themes that are helped by a little knowledge of English or American culture. If, say, one puzzle a week out of the six is not approachable by someone who can’t speak English, that still seems ok. I don’t complain that 20% of Nikoli’s magazines have language puzzles I cannot solve. I simply enjoy the 80% that I do understand.

  • *waves* I have been enjoying passively from afar since the start!

    This is your blog to do with as you will, but I am very excited about being a first-time DASHer in London in about seven weeks’ time. (Don’t know how the week in between the US and UK editions will work and whether I’ll just need to keep off the Internet in that time to avoid spoiling.) You probably have a schedule worked out but any sort of DASH preview, or general principles, or warm-up, or introduction you could do would be welcome – and I get the impression that this is exactly along the right lines.

    • Avatar photo drsudoku says:

      About the week-long gap you will have to mind — probably just not reading detailed posts to avoid spoilers is enough.

      I have not planned on this site to post a primer to basic puzzle hunting before this year’s DASH. Of course, with getting The Art of Puzzles in order I’ve hardly planned things out to May anyway.

  • term says:

    I have been regularly participating in MITMH for several years, and I can report an experience similar to Para’s. Constructions often impress, and the moment one grasps the mechanics is exhilarating like little else. I do run up against a language barrier occasionally, which is not fun, but not a big deal given the size of the contest. Cultural barriers are more frustrating when the nature of the impediment isn’t obvious, because such puzzles may still look tractable. Also, solving these with a team is always lots, and a big part of the, fun.

    Far as this site goes, I suspect instructionless puzzles which are solvable by one person on a similar timescale to Monday’s will consistently go down well.

  • Giovanni P. says:

    I definitely enjoyed teasing out this puzzle–it couldn’t have been easy to construct. The promise of more puzzlehunt-like prescriptions in the future is something I look forward to.

  • Like everyone else, I have nothing but praise for the April Fools’ Day puzzle. A fun solve, and the thrill of the a-ha makes me want to go back to solving things like P&A Magazine more regularly.

    A question on KenWord, though: Looking at it now, are the crossword clues even necessary to solve it, or can the TomToms be solved by pure logic alone? (Assuming one is able to intuit the special rules for them, of course.) It’s been a while since I solved it, but I believe my break-in was areas such as the fourth row in the first grid, where the word that could fit there was highly constrained, but then again, I’m rather more comfortable with crosswords than TomToms.

    • Avatar photo drsudoku says:

      Because of what is being placed into those grids, my opinion is that there are way too many answers until you know a few of the words so solving about 50% of the clues is sufficient. As you say, when you have certain words they can only go into certain rows or columns so you can break in that way.

      Of course with most of these puzzles, not being able to solve every clue isn’t necessary a problem. These puzzles can often be solved with incomplete information (like 4 of the 6 total KenWord grids) if you are good at guessing answers. That is one of the skills I have really worked at over the years which makes me a rather good puzzle hunter.

  • Tricia says:

    I can brag about coming in second place for an early BANG – and not having met my partner before the hunt! The most creative puzzle was when we arrived at a diner and were presented with a bag of jellybeans and no instructions. The jellybeans could be arranged to form a four letter word. We luckily hit upon the correct word, but not for the right reasons; the idea was to eat the jellybeans and arrange the letters in alphabetical order by flavor, but instead we arranged them in rainbow order. This puzzle sent almost all of the teams off in the wrong direction.

    Bragging now out of the way, I had a few setbacks on this particular puzzle. First I forgot to find one of the words in the search. Then I forgot to transcribe one of the givens into the Sudoku grid. So, it took me three days to solve. Oh, the shame.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.