Dry run

It's not required, but if you can manage to have a dry run in between the end of testsolving and BAPHL itself, it can be very helpful.

Why to have a dry run

• Having teams go through the full solving process in situ gives you a lot of information that will be helpful for running the actual hunt.
  • Spotting red herrings: In the dry run for BAPHL 10, the colored icons (which were used to help place answers on the train tracks in the endgame) were themed after Lucky Charms. This led to teams thinking that the correspondence was relevant for the first city meta, and going way down a rabbit hole that led nowhere.
  • Logistical concerns: how much time should teams spend at each location? Are the location-specific puzzles reasonable? Do the locations provide adequate space for solving comfortably?
  • Get a feel for what the experience of the hunt will be like as a whole, as opposed to the individual puzzles.
• Solvers who can't make it on the date of the hunt can get still get a chance to experience the hunt, while helping to improve the hunt for other teams.

When to have a dry run

• After your puzzles have been thoroughly testsolved, hopefully. You can get additional testsolving data from a dry run, but you should make sure that the puzzles are at least solvable and fair ahead of time.
• Before the hunt itself, obviously. But how far in advance?
  • At least two weeks in advance, so you have time to make corrections and change plans as needed.
  • But not so far in advance that your data isn't useful:
    • You want solving conditions (e.g., weather, seasonal availability of locations) to be similar to what solvers will experience during the hunt.
    • Some site-specific data might be subject to change.
  • Probably between 2 weeks and a month ahead of time, then. For BAPHL 10, schedule conflicts forced us into doing the dry run 3 weeks ahead of time, which worked out quite well.
• Consider having the dry run on a Sunday or a holiday that doesn't fall on a Saturday; some solvers would love to do BAPHL but can't go on a Saturday for religious reasons.

How to have a dry run

• Recruit solvers through whatever channels you have access to. If you have two tiers of difficulty, see if you can get some people with little or no prior solving experience for the easier tier.
• If you have enough volunteers for multiple teams, great. But make sure your dry run teams accurately reflect the typical BAPHL team.
  • The day before the dry run for BAPHL 10, we thought we had 7 or 8 solvers, and thus planned to have two teams. A couple people called in sick, so we had only 6 solvers. The solvers chose to break up into two groups of 3, but we might have gotten more useful data if they had stuck together as a single team.
  • But if part of your hunt involves interaction between teams (e.g., the "plane ticket" rebuses in BAPHL 10), then two or more smaller teams may be necessary to pull this off.
• Try to emulate the conditions of BAPHL as much as possible.
  • Assuming you have already testsolved the puzzles, encourage dry run teams to solve puzzles at a reasonably fast pace, solving in parallel if they have enough people, and moving on to the next puzzle when they get an answer rather than solving it completely.
    • In the dry run for BAPHL 10, as a result of having small (3-person) teams and some of their ingrained solving habits, there was almost no parallel solving, and the dry run took about 8 hours to complete.
  • The puzzles should appear the way you want them to look in the actual hunt. This includes visual design elements, flavortext, and other non-puzzly bits.
  • Dry run teams should check in and out of locations as normal.
    • But if you are planning on having teams go to locations in whatever order they choose, it may be better to have the dry run teams do a fixed order, to cut down on staffing requirements.
• If you have enough people staffing the locations, it can also be helpful to have someone from the running team following each of the dry run teams, taking note of red herrings and other stumbling blocks, and answering questions when appropriate (Is this really how the meta is supposed to appear? Does this trivia question take recent events into account?)