Musings on the hunt so far

One of the reasons I was interested in the GAR was a chance to how a different, independent strain of puzzling works.  A lot of what we do is convention (standard encodings, the way a puzzle is laid out, the way a hunt is run) and sometimes it's hard to realize just how arbitrary that might be.  One point of view is that the evolved Stanford/Microsoft/MIT puzzle community (which I will call "West Coast" for want of a better term) is basically a pinnacle of evolution, that any sufficiently perceptive puzzle hunt author will eventually arrive at a similar style, and that anyone who does things differently is pretty much a backwards provincial heathen.

At this point I think I can rule out that point of view.  There's a lot that Ravenchase does very differently.  I like some of it very much, I like other parts less, but overall it has been a really mind expanding experience to participate in a hunt that's a whole lot of fun and also very different from hunts as we know them.

First of all, consider this: Basically one dude (Josh Czarda) is writing and running, entirely on his lonesome, a 7-day, 6-city hunt, with a good 6-8 hours of puzzling every day.  And (so far) it's totally fun, the clues do not feel "phoned in", the sites are interesting, the puzzles are often deeply environmental, and each day brings unique surprises.  I am really impressed.  If nothing else, I think we may be missing some opportunities to make awesome experiences with much less GC cost than we normally put into things.

So what is different?

First of all, there's a visibly different aesthetic sense.  Ravenchase cares more about some things and less about others than "West Coast" designers.  Every day the opening puzzle is rolled up in a neat little parchment scroll tied off with a ribbon.  Clues are always presented in rhymes with an old-timey font (and when you get a phone message, it's also in rhyme).  There are neat little trinkets and clue boxes that are well made.  I bet he would think our clue presentation is a little dry.  At the same time, if they're indexing into text to get a message, and they can't get certain letters, they'll just drop them directly into the cipher (which will read something like: 1-17-5-3-V-27-A-15-...).  If there's a meta that uses an acrostic or otherwise gets data from input answers, there may be inputs that are completely unused -- but there's a 30 minute time penalty for missing/wrong answers so you need to solve them anyway.  These are things that I think "West Coast" designers and players would consider "cheesy", though I'm not sure it actually matters to the experience.

Another striking difference is the code systems in play.  Conventional "West Coast" puzzling is all about codesheets and indexing.  GAR (outside of the starter puzzle) hasn't used a single codesheet clue.  Ravenchase uses Vigenere all the time, and very occasionally some historical cipher (like Bacon's cipher).  They also use indexing very heavily in several ways: Pure letter count, row-word-letter (Beale-type) lookup, and row-column lookup.  As mentioned above, they'll stick literals in their index sequences.  This all takes some getting used to, but it clearly gives them huge flexibility in that they can get basically anything from any input.  Again there's an aesthetic difference where that feels "inelegant" to our sensibilities, but I bet Ravenchase would think that we're just weirdly obsessed with Braille, Morse, semaphore and binary.

But an even bigger difference is that Ravenchase has a huge focus on environmental puzzles.  If there's one thing I take away from this hunt (so far) it's this.  On the "West Coast" it feels like real environmental puzzles are something of a dying art.  This is basically the Ravenchase bread and butter and it's a really nice change.  You might have gotten the impression that they're just asking us to go to some transparently clued statue and read a few words off a plaque.  They do that sometimes, no doubt, but they do a lot of more interesting things too:

- The photo or hint chases are often nontrivial in an interesting way.  We have many clues that are not obvious, nor are they some obscure detail only a local would know, but they are something that if you sit down and think about and spend some quality time with Google or ask around you can find.  Where is a building that looks like this likely to be?  Given this view of a street with a landmark in the distance, which part of the city is it in?  Whose monument in D.C. might include statues of homeless people?

- The selection and use of highly puzzly items is often creative.  The Wishing Tree was a great find, and hiding blacklight-visible "wishes" on the tree was a clever mechanism.  The hanging monkeys were just awesome, and identifying their shapes made you really pay attention to how they're formed.

- A couple times now he has used the shape of a window or grating as a transposition grid.  That's an amazing a-ha the first time we came across it; maybe it's a one-shot trick but I think it can be elaborated.

- Lately we've been getting more complicated environmental clues.  For example at the Washington Monument in Baltimore, part of what we had to find involved the phrase "Foreword Racer".  We knew right away that had to be an anagram of something, but what?  The monument is pretty elaborate, so we spent a bunch of time crawling over its grounds (and in the process filling in other parts of the clue).  One thing you couldn't help but notice were four intriguing sculptures, at four corners of the traffic circle around the monument, depicting "War", "Peace", "Order" and "Force".  (I actually found "Order" a little creepy, this guy sitting on top of a writhing panther.)  Eventually someone (Rich or Jonathan?) noticed that "Foreword Racer" is an anagram of "War", "Order", and "Force" -- leaving "Peace" as the odd one out and the answer to that part of the clue.  I think this kind of thing strikes a nice balance between an annoying needle-in-a-haystack site-search and a simple go-get-it plaque lookup.

- The environmentals are often set up so you can ask locals for help.  In fact he often starts us at a visitor center, or has us go through a visitor center, probably to give an opportunity to ask about things we might have missed.  It's actually kind of fun to show a museum docent or visitor center staffer an oblique clue, and then you see their eyes light up and they say "oh, you're probably looking for the Pride of Baltimore" or whatever.  There was a clue at Camden Yards that depended on identifying several famous Oriole players, and Rachel said that her mom "made some friends" and asked passersby to ID them.  Basic sports trivia is a fun thing for passersby to help out with.  (It's also eminently Google/Goggles-friendly if you'd rather go that route.)

Another difference is the hunt structure.  At this point it looks like Josh's main technique is a string-of-clusters: A linear series of clues all within a walkable area.  This is of course not unknown to us either, since it lets you park your van and get out on foot to explore an area.  The clues in a cluster are often presented in what we think of as "Chinese New Year Hunt" format (though IMHO they're way more interesting than CNYTH's clues), where you get all the clues for an area at the start, solve enough of them to plot an efficient route, and then start walking.  There's usually a meta of some kind for the cluster, which leads you on to the next cluster.  This is a robust structure because generally you can get the meta with maybe 75% of the clues, which means you can miss some of them (as mentioned earlier, there's a 30-minute penalty for doing so).  Clusters have a recommended end time so you can pace yourself.  He seems to set up the clusters so that your likely route will take you past some of the hard-to-ID locations.  That way you get the nice experience of "oh, _there's_ the statue of a dude with a piece of paper".

I think a key advantage of this structure is that GC can monitor a whole slew of clue sites in one place.  Josh hung around the Baltimore inner harbor yesterday, which was the day's primary "cluster", and I'm sure he was ready to help teams or check on any issues.  (And of course he's on the phone if teams elsewhere need help.)  So far, this hunt has been logistically really smooth.  I complimented Josh on this last night, and his response (like any experienced GC) was to immediately knock on the table and give me a "don't-jinx-it!" look.  He visits all the clue sites the night before to check if anything changed.  One place in DC he had to change the hunt on the fly (we didn't know this until he told us, it was smoothly done); another location in Annapolis required errata.

Finally, I thought it was interesting that if Josh needs to deliver a clue somewhere that doesn't have a secure location, and he can't get the cooperation of a local vendor, he'll just hire someone off Craigslist.  He mentioned that to us last night (we were asking who the "woman in red" was), and indeed I found some of his ads, which are very straightforward.  Looks like his rate is $12/hour for 6 hours of standing around and handing out a clue.  Probably more if they actually need to act?  This isn't something I would ever think to do as GC, and probably needs experience and maybe a corporate shell to do well, but I think it's a nice option.

There's more, but it's time to get ready for the Philadelphia-to-Princeton leg.  Still, I feel like this has been a really educational experience for me and I wanted to share some of my thoughts.