Hi, we are Ruth and Hal, designers of Cryptid and deduction game enthusiasts. Cryptid is our first published game, though we've been tinkering with designing games about as long as we've been playing games together. Many deduction games we love, such as Letters from Whitechapel, Zendo and Tragedy Looper, have asymmetry baked in, where one player knows all the information. We both enjoy the deduction side of these games, so we started to think about ways games like these could be made into a symmetrical experience, with all players engaged with the satisfying puzzles that they offer. After getting home from an evening playing Zendo in the pub, we discussed whether it would be possible to programmatically generate statements about a map such that the intersection of all the statements leaves only one valid answer. Ruth is a computer scientist, and thought it would be possible, and a fun problem to work on to boot.
The core of Cryptid fell into place quite quickly. In the next week we created the first prototype, which Hal drew with colored pencils on a sheet of hex paper printed off the internet. For this first attempt, he picked a space on the map, and built the map around forcing four clues to lead to that arbitrarily picked space. Ruth was busy working on a program that could algorithmically generate sets of clues, one for each player, which would narrow down to exactly one space, but we wanted to start playtesting as soon as possible. The first playtests worked, and showed there was a satisfying puzzle to solve at the heart of the game, but threw up our first significant design challenge. Because the game is completely focused on solving a puzzle, the natural conclusion of the game was one player announcing a space they believe is the habitat, and all other players being unable to disprove this. If all the clues allow this space, then it must be the correct space. We needed to decide when players could announce such a guess.
Our initial rules had two actions: ask a question or spend a search team. "Ask a question" is an action that has remained through nearly all iterations of the game: You pick a space, then ask one player of your choice whether that space can be the habitat, according to their clue. Our first version also had a method by which players would gain search teams. Search teams could be spent to ask all players whether a space can be the habitat. If they all said yes, and the space fit your clue, you could announce victory. The search team economy was frustrating; if you knew the answer but could not win because you did not have a search team at the right time, it felt unfair. This led to our first design principle:
1) Do not place barriers between the player and the puzzle.
When another player answers your question and says the space cannot be the habitat, they place a cube of their color on that space. We quickly decided to expand this into another design principle:
2) Information is permanently visible.
Players place a disc on the space if they say the space can be the habitat. We wanted the game to test deduction skills rather than note-taking or memory.
Once we had established this visual representation of each player's information and removed the need to manage other resources, we came across our next design principle, an intrinsic feature of the game:
3) Information is the only currency.
This helped solve some of the outstanding questions we had about our rules. Searching became an action that was always available, but required the player to spend information by placing a disc, which were considered more valuable. We added a rule where causing another player to place a cube always means you must place a cube yourself. Cubes were then plentiful and discs were rare, so giving up a disc is giving up rarer information. This provides a deterrent for haphazard searches, and makes getting a disc in response to a question feel more like a reward.
Our final design principle is probably the most obvious, and also the one that took the most tinkering:
4) Players have to feel like they were given a fair chance.
Each player's randomly assigned clue allows them to eliminate a section of the map. These clues need to feel balanced; when the players reveal and discuss their clues at the end of the game, we don't want anybody to feel like they started with significantly worse information than another player. However, there need to be enough potential clues for there to be multiple intersections reducing to one space. Initial clues varied widely in coverage as there were multiple versions of each clue with varying coverage, e.g., a player might be told that the creature's habitat is within one space of mountains, or within two, or three. While we allowed clues to be chosen only if they eliminated a roughly similar number of spaces, it took a few rounds of playtesting to determine exactly how rough this threshold should be. We decided not to allow redundant clues (where one clue only eliminates a subset of the spaces eliminated by another player's clue) as it was clearly quite unfair if one player started knowing everything another knew and more. Eventually we removed the variability of the distances so that all clues about any one feature of the map use the same distance.
We also needed to balance the number of possible clues to make deducing other players' clues a satisfying difficulty. We started with a few features that might be mentioned: four types of terrain, one type of track, and three colors of structure. For many rounds of playtesting, we tried to add more clues by adding a compass, giving clues like "east of the green standing stone". While this solved our problems numerically, we were never quite happy with the complexities this introduced as there would always be some hexes on the map where the direction wasn't obvious. It also started feeling like a difficult type of clue to be dealt as the straight line marking the boundary between allowed and eliminated hexes was visually obvious and hard to bluff your way out of. Ultimately we scrapped the compass and rather than add an extra type of clue for players to get their heads around, we split two of the features (animal territories and structures) into two subsets. Players can now be told that the habitat is within a small distance of a general feature (e.g., "within one space of animal territory") or within a larger distance of a specific type of the feature (e.g., "within two spaces of bear territory"). This means that all of the clues are related to the distance between a hex and a feature, so players have only one concept to deal with when assessing potential spaces, which makes the game simpler to explain.
During playtesting, people mentioned that it could be intimidating to be dropped into the full game. New players were particularly struggling when given negative clues, such as "the habitat is not within two spaces of a standing stone", so we split the game into two difficulties: normal mode, with no negative clues, and advanced mode, where all clues are allowed and an additional structure color adds more possibilities to the clue space. After some initial feedback from publisher Osprey Games, we also added an optional hint system that tells all players something about the clues in general, like "there are no 'within one' clues". This can be decided during the game if any players are having a tough time narrowing things down, letting players adjust the difficulty on the fly.
Another addition requested by Osprey was a way to distribute clues to players without using the website, which we had been thinking about ourselves for a while without a satisfactory solution. Many of our initial ideas were complex and overwrought, and ultimately they didn't work. We tried a system in which cards would belong to multiple groups such that if you took any subset of a group it would give a functioning set-up, but after running the numbers these groups ended up being too small for this to really work. We also tried a set of card sleeves with holes cut in different places so that putting a single card in each player's sleeve in turn would give out one clue each. This was also too complex and too prone to people seeing clues by mistake. In the end, the simplest solution was also the best; together with Osprey, we settled on a booklet of clues for each player and a deck of cards showing which number clue to look up on one side, and how to set up the map on the other.
We're excited for deduction fans and puzzle lovers to be able to play Cryptid, especially with the lovely production Osprey has put together!