Log in

No account? Create an account
21 January 2013 @ 02:40 pm
Too Big to Solve?  
Not my tagline, but a good description for the Mystery Hunt that just happened. One line of dialogue after last year's Hunt that I led with in my wrap-up was a question of when is too soon for a Hunt to end. I said, in this era of a few competitive teams trying to grow to get over the winning hurdle, constructors aiming bigger was a mistake. The Hunt ending after 36 hours (Midnight Saturday) is fine if that makes the solving experience stretch over the weekend for everyone else. I won't comment generally on this year's effort but it seems a great example to point back to of too much ambition by too many people towards the further militarization of the size of Hunt so that by 2025 the team "The whole of new USA" can go after the coin against "USSReunited" for at least a month. The sense of "puzzle" versus "grindy work" is also a discussion I have every year and I don't choose to repeat myself. I've felt since 2008 that the Mystery Hunt is far from an event I'd regularly attend in person although I'm glad to have finally been onsite to play with Team Luck with whom I've been a "free agent" now for three years.

I had a good solving year as things go relatively, but it was mostly demoralizing personally. I soloed Palmer's Portals, for example, but spent many hours after basically solving 8/10ths with a need to tweak a very small and underconstrained set of things to get from that hard work state to a finished state. At some stage I told the team "I'm going to solve Portals and the Feynman meta and then go sleep" and I met this goal but in many times the expected time when I gave the statement. I led the solve of both Danny Ocean (with zebraboy stating the most necessary last bit to get my work over the cliff) and Richard Feynman (with Jasters). I obviously co-solved lots of the logic puzzles and other puzzles, and gave various finishing help to a range of things too. I think I did this best for "Kid Crossword" once when he had spent a lot of timing mastering the hard steps of a crossword/scrabble puzzle -- and could quite impressively fast rewrite out the set of steps I wanted him to do about the puzzle -- and the follow-up steps were not obvious but I led the killing of the beast. This was too often the feel for these puzzles, and my assassination rate was far lower than I wanted. My Sunday was spent earning 3 puzzle answers by actually going to an event, and then falsely believing the power to buy some answers would let me finish solving the Indiana Jones mini-metas -- where I had already mostly soloed Adventure 2's snakes with 5/8 answers, but then killed myself dead on #1/Ouroboros for the rest of the day for so long solving, as many solvers will say in hindsight, the puzzle that was meant to be in one of a dozen ways and not the puzzle it was. Let me state here as I did for hours with my team, the phrase "I'm not cut out for this" is horrible flavor. It implies both cut this out and, in a different way, also don't cut this out. This makes you want to cut it out, which takes a lot of time, but also to not invest too much time in cutting it out, so as to save the wasted time of doing a task you are being told not to do. Other wordings are far safer, and implied negatives within positives is one of the five worst flavor failure modes in my opinion. Puzzle editing and flavor text is an art and is certainly the biggest variable from year to year and constructing team to constructing team.

So yeah, Mystery Hunt happened. And there were the usual share of overwhelmingly incredible Aha moments. Endgame seemed very fun and I wish all teams could do just that for the weekend or at least a lot more things like that. More of that, and more sleep, would have both been some good choices this year. If only the puzzles solved on schedule.

ETA: And as I added far below around comment #300, as a solver who was both frustrated yet had fun in this Hunt, I do want to thank everyone on Sages for the incredible effort they put in. Making a Mystery Hunt is a gift for all solvers whether it matches expectations or not, and as a mostly thankless job I do want the constructors and editors and software engineers and graphic designers and cooks and phone center workers and everyone else to know I appreciated all you did over the last weekend to give us several days together for puzzling.

Further, as I was asked to write a larger piece elsewhere that has given me personally a lot more attention as the face of the criticism, and as I use the phrase "My team" a lot in general as solving forms this kind of bond, I want to be very clear: since Bombers broke up after 2009 I have been a free agent. I have solved recently with Team Luck but am not a core part of their leadership and these opinions I state are my own. I intend to form my own team next year to go after the coin again, and if you have a problem with what I have said anywhere on the internets, please hate me for it. I believe in my posts I have been offering constructive criticism, but even what I have said is without all the facts of what went on inside Sages so I could easily be speaking from ignorance a lot of the time.

EFTA: Thanks to tablesaw for pointing out this chronologic feature of posts. If you want to see all the additions to this post in time sorted order, go here http://motris.livejournal.com/181790.html?view=flat. We're on page 14 at the moment.
Alisonlandofnowhere on January 25th, 2013 09:12 pm (UTC)
Creating this as a sub-thread for members of Sages to ask for feedback on puzzles that haven't gotten much. (If you'd rather not do so publicly, send me an LJ message, or send the feedback to Sages to pass on to everyone involved).

My requests:
Transit Links (co-authored with MellowMellon)
Mergers (editor)
Stakeout (editor)
Grandson of the Realm of Unspeakable Chaos (editor, though not the main one)

(boing!) Cnoocy Mosque O'Witzcnoocy on January 25th, 2013 09:22 pm (UTC)
We figured out a lot about Grandson, but the people working on it weren't the people who worked on Son, and when I took a quick look at Son it looked different enough that I didn't go down that path. Did it get solved?
noahspuzzlelj on January 25th, 2013 10:06 pm (UTC)
Grandson was the first of the Chaotic puzzles that Plant got the answer extraction on! I was only involved at the end, but as far as I can tell from talking to the people who had worked on the whole thing it was an excellent hunt puzzle.
nameelectricshadow4 on January 25th, 2013 09:35 pm (UTC)
For Mergers, the reason Codex (or at least the people I was working with) never got anywhere was because there were too many choices and too few places for confirmation in the middle of the puzzle. We had a few of the ones in the solution on our spreadsheet, but also a bunch of merges that weren't the ones used.

The two puzzles from the past it feels similar to were http://web.mit.edu/puzzle/www/12/phantom_of_the_operator/functions/ and http://web.mit.edu/puzzle/www/10/puzzles/1752/string_maze/ . Both of which have a similar branching factor, but both have more confirmation steps in the middle.

Having something like "each level of the tree has a different color for a different type of merger" or even "here are the 5 types of mergers we use, and we use each of them 3 times" would have made it easier to ensure that the solver is on the right track.

Also, the Aardvark+King->Arthur and Arthur+Cheetah->Chester associations seem a bit rough. Perfectly fair if they're clued or if there's some reason to believe those go together, but it doesn't jump out the way it should if you're picking 2 from a list of 10. Similarly, at the top, it's quite a leap to get to "copper + tin", and there's no real confirmation before that point.

Also, as a general note, for a lot of puzzles, there's a bit of "if we didn't have 50 other puzzles open we might have looked harder and broken through". The overwhelming mass of "too many puzzle-hours" puzzles means that even puzzles which would have been fine on their own might have been too complicated here.
Alisonlandofnowhere on January 25th, 2013 09:45 pm (UTC)

I did get the impression your final paragraph was what was happening in general with the puzzles I haven't heard much about.

And I did push for more intermediate-stage confirmation on Mergers (the original version had no tree), but we ended deciding that this version was good enough, possibly because we were running out of time.
Cody B.: contemplationcodeman38 on January 26th, 2013 03:46 am (UTC)
Listing the types of mergers would indeed have been useful on Mergers--because otherwise, there are just too many ways to get unanticipated mergers out of some of the term pairs.

Some of the Codexians might have heard me shouting out that "Mac OS X" and "King" should totally be given a missing-link merger to get "Lion"...which, of course, turned out to not even be one of the merger types that was used.
Doug Orleansdougo on January 25th, 2013 09:49 pm (UTC)
I was happy to see Stakeout; it was an idea I had had myself, but I hadn't come up with a way to build a puzzle around it. (I was also happy when I figured out the significance of the title.) Our team (Central Services) ran into a dead end at the extraction step, but I'm not sure it's the puzzle's fault; we just have a habit of forgetting to look at the original puzzle presentation once we've entered all the data into a spreadsheet. Also, it didn't help that one of the extraction methods we tried (indexing the number of given jigsaw pieces for each album into the title of the other album) gave us the letters APORRST (once we used an alternate release for one of the entries) which happened to have two anagrams, one of which was the answer to a different puzzle in the round!

One way I thought of to clue it would be to print one letter on one piece of each album. But that might be too much of a hint?
Ali LloydAli Lloyd on January 25th, 2013 10:44 pm (UTC)
I testsolved Stakeout but never worked out the significance of the title... care to enlighten me? :-)
(no subject) - dougo on January 25th, 2013 10:50 pm (UTC) (Expand)
(no subject) - (Anonymous) on January 26th, 2013 04:36 pm (UTC) (Expand)
noahspuzzlelj on January 25th, 2013 10:04 pm (UTC)
I didn't work much on Mergers, but Malia did for hours. Plant got stuck on the very last step. First, they weren't sure that the rest was right because a lot of the steps felt ambiguous. Second, the last step is a bit rough, previously the append/prepend examples had all been full words, and in the last step it works very differently (where it isn't even a prefix, let alone a word).

Usually puzzles of this type involve combining 3 or 4 things so that the confirmations are stronger. There was a recent hunt that had basically this puzzle but with sets of three (I can't remember which one, all I remember is that Yom Kippur was one of the intermediate answers), and Connect Four from SPIES used sets of four.

If you want to do it with 2 instead, then you need to have some kind of extra mechanism for confirmation. Otherwise it'll end up either very easy, or too ambiguous. Preferably this should be a mechanic that's not very helpful at the beginning but which locks things into place at the end. Here are a few possibilities: give the lengths of the unknown words, give a list of how many times each rule (prepend, etc.) was used, give the multi-set of letters used between all of the answers. There may be some even better confirmation idea I haven't thought of.

There were a couple specific steps that people were unhappy with, but other than the last step I think the biggest complaint people had was the CHEETAH step and that's because they hadn't put together that the score was 10.0 and not 10. So as long as there had been some sort of confirmation mechanic that wouldn't have been a problem.
nameelectricshadow4 on January 25th, 2013 10:21 pm (UTC)
http://web.mit.edu/puzzle/www/12/puzzles/phantom_of_the_operator/set_theory/ was the one you were thinking of with sets of 3 (well, 2 and 3 intermixed).
Dr. C. Scott Ananiancananian on January 25th, 2013 10:40 pm (UTC)
Codex never solved Grandson, but we got very close. I wasn't involved too much on the solving side, but I edited Seth Schoen when he wrote the conlang puzzle "Sounds Good To Me" for the Producers hunt. So I can talk a little bit about editing conlang puzzles.

First, be sure you have enough test solvers and editors. Codex had barely enough solvers able to tackle the "Sounds Good", and although I'm not a conlang expert, I was editing the puzzle in part because our better conlang'ers needed to be saved for the test solve step. We all knew the puzzle was on thin ice because of the thinness of our solvers, and we took great care to make sure that it was in final shape before it went to the test solvers. If there had been any issues found during the test solve, the puzzle would probably have had to be cut, because we didn't have enough solvers for another go-round. We all accepted this up front.

I think we also finessed the lack of test solvers by testing the extraction step and clue phrases separate from the puzzle. Not really possible for 'Grandson', but it helped us gain confidence in the "Sounds Good".

The other thing we were careful to do was to provide confirmation of the essential ahas, as high up in the puzzle as possible. "Sounds Good" is written in toki pona transliterated into hiragana. We added a clue phrase explicitly warning that this was not Japanese, to forestall a bunch of wasted effort which a solver might spend on hiragana cryptograms. The words "toki pona" are the very first words after that. I also made sure that Seth's introductory text included every hiragana syllable so that the desired mapping could be confirmed.

Each clue phrase in "Sounds Good" also ends with the exact same stereotyped phrase: "this is ______ in [language]". This meant that you could get pretty far through the puzzle just looking up language names in toki pona. Combined with the straight-forward extraction (which was forgiving of some missed identifications) it provided an angle of attack even if this was the first time you were reading toki pona.

For "Grandson", the thing that jumps out at me from the solution is that there are six linguistic rules necessary, which need to be applied in a specific order. As an editor, I would have asked the author to provide the exact contexts which the hunter would use the infer these rules and their order. Perhaps there's not quite enough evidence for some of the rules (or the order), or the evidence provided could be interpreted more than one way. (Certainly Codex managed to translate the text but ended up with a different transliteration at the end.) Without seeing the exact reasoning which the puzzle's author expected to be used (which should really be documented in the solution), I can't tell if this was an ambiguity/flaw in the puzzle or a mistake we made while solving it. This is the same information the editor or fact-checker should demand, in order to ensure that the puzzle is well-formed and the answer is unique.

I guess my basic advice for a conlang puzzle would be to force the author to provide translations and written-down reasoning in the solution such that the puzzle can be reasonably checked by a non-expert.
(no subject) - pesto17 on January 28th, 2013 04:08 pm (UTC) (Expand)
AJDdr_whom on January 25th, 2013 11:38 pm (UTC)
My comments on Mergers and Grandson of the Realm..., among other puzzles, can be found here.
Dave "Novalis" Turnernovalis on January 26th, 2013 04:36 am (UTC)
I did a bit of grunt work on Transit Links and found it to be a perfectly reasonable, solvable puzzle. I did not love the extraction, but I put the blame for that on the hunt tradition which requires that answers be extracted from puzzles; this tradition plays somewhat poorly with logic puzzles.
(no subject) - landofnowhere on January 26th, 2013 05:23 am (UTC) (Expand)
(Anonymous) on January 26th, 2013 07:02 am (UTC)
As a first-time author/editor, I would be grateful for feedback on a number of puzzles that I helped with writing/editing (and, in one instance, test-solving):

* A Streetcar Named... (Rubik)
* A Walk Around Town (Get Smart)
. Eclectic Spatial Geometry (Marty Bishop)
. Houston, We Have Liftoff (Indiana Jones, Adventure 3)
. I Left My Stomach In Salt Lake City (Indiana Jones, Adventure 3)
. Numbers (Indiana Jones, Adventure 1)
* Security Theater (Marty Bishop)
* Stratosphere (Casino puzzle in Danny Ocean)

Puzzles marked * are those in which I had an idea/authorship role, and I'd be particularly happy to get feedback on them.

I'm generally interested in solvers' experiences, because I'd like to hone my instincts for what makes a good puzzle so that I can write more of them. I'm also interested in whether any prolonged periods of not-very-fun frustration occurred and, if so, how the relevant puzzle could have been adapted in order to mitigate these.

(no subject) - zotmeister on January 26th, 2013 09:12 am (UTC) (Expand)
(no subject) - (Anonymous) on January 26th, 2013 09:59 pm (UTC) (Expand)
(no subject) - ppaladin on January 26th, 2013 02:46 pm (UTC) (Expand)
(no subject) - antimony on January 27th, 2013 01:45 am (UTC) (Expand)
Re: Puzzle feedback - (Anonymous) on January 27th, 2013 04:22 am (UTC) (Expand)
Re: Puzzle feedback - antimony on January 28th, 2013 12:18 am (UTC) (Expand)
Re: Puzzle feedback - (Anonymous) on January 28th, 2013 11:05 am (UTC) (Expand)
Re: Puzzle feedback - (Anonymous) on January 28th, 2013 04:10 pm (UTC) (Expand)
Re: Puzzle feedback - snowspinner on January 29th, 2013 04:45 pm (UTC) (Expand)
Re: Puzzle feedback - (Anonymous) on January 30th, 2013 01:10 am (UTC) (Expand)
(no subject) - devjoe on January 27th, 2013 06:03 am (UTC) (Expand)
(no subject) - Dima Kamalov on January 28th, 2013 06:21 am (UTC) (Expand)
Adam R. Woodzotmeister on January 26th, 2013 08:43 am (UTC)
Transit Links was the high-point of the Hunt for me. WARNING: Story alert, contains spoilers.

It was after the hint for the puzzle was released that our team's resident cartographer tackled it in earnest, figuring out the stations were to be paired up (and in the process editing an earlier station identification that was the name of a station in more than one city!). Devjoe, who had just walked by and asked about the puzzle ten seconds prior, promptly determined, "You know there's a puzzle type called Number Link, right?" If he hadn't have said that, I could have stared over Mapmaker's shoulder for hours and gotten nowhere, but from that gifted a-ha, all the rest was mine. I pointed out the flavortext implied we'd have TWO superimposed Number Links, and that we should do some math (thanks to that hint) to determine if the grid size supported that theory. I botched the addition, but they didn't, and finding a total of exactly 242, we knew we were on the right track.

Joe was able to get to graph paper before I did, so he made his own copy of the puzzle while I waited for Mapmaker to fill a spreadsheet, which we then discovered couldn't be printed from his machine, so I copied it onto graph paper as well. Joe had already gotten a head start of about five minutes when I sat down at the same table and said, "You know something Joe, if I solve this before you do, despite your head start, I will never hold it against you, but boy will I ever feel good about it!" Joe laughed in the way only he can, and we put our heads down and worked at it. I'm sure that at times we each looked up at the other's paper, but I know the times that I did, either I didn't like what I saw and didn't go with it, or what he had matched what I did already.

Sure enough, I finished first. I was beyond impressed with that puzzle; I wouldn't have expected something with so many variables and so little deduction to feel so... fair. I felt my false assumptions were all very quickly disproven. Answer extraction not being my thing, I handed the solved paper back to Mapmaker for him to do his thing, but I did leave him with a theory; after one false lead was followed, Tyler tested my theory and it proved to be correct. But I'm skipping ahead.

It was not a close race. Joe finished the puzzle at least five minutes after I did, probably closer to ten, and at the end of it he was staring right at my solution, comparing it to what he had tried, in order to finish it. To say Joe is a better solver than I is a massive understatement, but that one time, I surpassed him. I'll take it.

The best part, however, was not long after the solve. Thomas walked by to see what we were up to. Joe told him I just solved a Siamese Number Link. Thomas started to put his hand up, then just slowly turned around and walked away without another word. Needless to say, Joe and I both reacted to this by laughing hysterically.

I thought the puzzle was very well designed. The hint should have been there all along - it was needed to confirm our "a-ha" theory, and it basically meant I had the benefits of a Link-A-Pix to help with my Siamese Number Link, without which it'd have been a real headscratcher, if not actually intractable. But with it, the resulting puzzle was nothing short of elegant, and frankly I'm damn impressed it was ever built. Even with the "burn-offs" involved (by which I mean segments the paths travel through JUST to use up necessary length to get the distances to match, areas that could be left blank if these were just Number Links), I was quite confident the solution was unique, which for a puzzle of intuition with that level of variability is an astounding feat.

In case you're curious: that false answer extraction lead came courtesy of someone who usually nails that sort of thing first guess, when he said, "Shouldn't EVERY cell of the grid have a name now?" Two names, actually, but yes, that made too much sense, and time was spent filling in a spreadsheet with intermediate station names, looking for correlations between the pairings.

I should remember solving this puzzle for a long time. It was a great team effort, and we each had an epic performance enabled by a very solidly constructed puzzle.
(no subject) - MellowMelon [wordpress.com] on January 26th, 2013 09:37 am (UTC) (Expand)
transit links - (Anonymous) on January 29th, 2013 08:13 pm (UTC) (Expand)