Jeff is going through some self-flagellation over at his blog over the 2004 Mystery Hunt, and I have to agree with him — our hunt sucked a lot. On the other hand, since I was responsible for a decent chunk of the suckage, I feel the need to find the silver lining in his dark cloud.
Firstly, many people did have fun. It was not that we wrote a completely awful hunt with no redeeming characteristics. I heard a lot of compliments about the creativity of the arrival skits, the detail in the maps, and the transformation rules (although ORIGAMI PIRATE HAT threw people for a loop). Our innovations in approach were novel and helped expand the idea of what the hunt could be like. It was a far cry from my first hunt, which was Engimatology — paper copies of puzzles, many quite standard, and a theme pasted onto the puzzles like faux veneer on particleboard.
In that same vein, our structure (here I toot my own horn a bit) was pretty cool, albeit totally ridiculous in it complexities. I don’t think people had really considered the idea of double-use metas on the scale that we did. Similarly, the “choose your own adventure” nature of the gameplay was new. We really pushed the envelope in terms of structure, and I think it was our inexperience and lack of foresight that caused the failure rather than problems inherent in the ideas.
Jeff’s main beef is with the testing process, which I agree was terrible. I think the root reason is that we had 50+ people wanting to be a part of the organizing team, but not wanting to do any real work for it. So we wrote 100+ puzzle, nominally assigning 2 to each person. Then testing began, and nobody wanted to sit down and actually test puzzles, so this Wiki thing was set up for people to bandy about solution approaches. This had the benefit of letting authors get inside the heads of the solvers a bit, but then we had people test-solving puzzles over the course of weeks. Test solving groups were small, so people allocated extra time to compensate for size and lack of 133tn355. A far better system would be to have the team divided in half or thirds at most and have them all sit down and toil on a set of puzzles for 5 hours, with a ten minute debriefing with the authors afterwords. Too much accomodation was made for 5 remote test solvers. We had phones — we could have called to get feedback on our puzzles.
To sum up, we had a lot of great ideas with terrible execution. That is not to say that OPS screwed up, or that PUZZ was not on top of things or QA was too slack. We just lacked the experience to make rational judgments about puzzle difficulty, hunt difficulty, and so on. I showed up in Cambridge a week before the hunt and lived in Building 4 the entire time until we moved to the Bush Room, and then I slept in there. I hardly saw any of my friends who were not involved with the Hunt. Of course I couldn’t make rational decisions — I was popping multivitamins all day long and living on coffee and CSS. But then again, the bulk of our team of 50 people was not around, and we were too slow to come to decisions like “this puzzle sucks too much and we’re replacing it with a word find.”
Next time (and there will be a next time) we will do better.