In Spirit Island, Fear is a good thing - it's you striking terror into the hearts of the Invaders. Exactly how they'll react isn't predictable, but it often involves some of them fleeing the island, and as you earn more Fear Cards the game's victory condition becomes easier + easier.
Fear is one of the few major systems in Spirit Island that wasn't present in the initial design. After a month or two of testing, I realized that the Invaders felt not merely implacable (good) but inhuman (bad) - here they are facing down all these inexplicable disasters, and it doesn't affect what they do? At all?
I briefly tried working Fear just like Damage - affecting only the Invader pieces in a given land. But littering the board with fear tokens made things super-cluttered, involved way too many rules, and didn't really represent the "Invader morale as a whole" I was looking for.
So I swapped to a Fear track. Powers could do Fear alongside their other effects. If they targeted a land with Invaders, the Fear track would advance; destroying Cities also did Fear. If the Fear track hit its end, the players would earn a Fear level with some static benefit and reset the track. The first versions of this required earning 4 Fear/player to gain a benefit, but that didn't happen often enough, so I swapped to 2 Fear/player:
I demoed Spirit Island many times at Origins that year, and one playtest resulted in a huge amount of insightful, useful feedback, including about Fear. (Thanks, Jared + Luke!) They wanted Fear to be more relevant: they'd won while only reaching one or two Fear spaces. That didn't feel right or fun for them... and they wanted the rewards to be game-to-game variable, for replayability. I also realized the effects should probably be hidden (mostly for thematic reasons), so the next Fear Track featured spaces for 7 face-down tiles:
This worked better, but the big challenge became remembering. Players forgot to take their bonus Fear when destroying Cities, or to move Fear back when new ones arrived. Players forgot whether they'd done Fear. Players would jostle the track, and not be able to remember where the marker should go. And still, Fear only really felt relevant when Spirits went heavily into Fear-inflicting powers.
The two parts of this problem proved fixable with a single change: make destroying a Town (then called a Settlement) cause Fear. Towns get destroyed much more often than Cities, and that repetition helped players internalize the rule. It also added much more "baseline" Fear (ie, Fear earned without explicit Fear-dealing Powers), making Fear feel more relevant in every game. With a few other tweaks (dropping the "-1 Fear per City built" as too hard to remember), we had:
By this point, I was testing ways to boost difficulty, and ran into a problem: making the game harder didn't make a Fear victory much harder. So as the game got more difficult, "just do Fear" became a more + more dominant strategy. I started increasing the # of tiles you had to work through as game difficulty rose, which for clarity's sake resulted in consolidating each level of Fear (by then called "Fright Level") into a single stack of tiles.
Eventually, I got tired of making and stacking tiles, and shifted Fear effects to cards. This also let them be a little more complex - the tiles were extremely space-limited!
The "track" mechanism was still problematic; new players always had fencepost errors with whether the "reset" space was an actual space that needed to be traversed, and bumping the track was a constant issue. I tried substituting a pool of tokens that got moved, and found it worked much better: it was bump-proof, the slightly greater physicality involved in shifting a few tokens (rather than just advancing a cube) let players recollect better whether they'd tallied the Fear for a given act, and there were no more fencepost errors.
I was also trying to simplify game setup - partly for player experience, partly out of simple self-preservation, as I was running so many solo tests - and Fear Deck construction was one of the bigger hassles. You had to shuffle 3 separate small piles of cards, which was irritating. But Fear effects were generally pretty short (due to having been on tiles once upon a time), so I consolidated them 3 to a card (thematically related, but increasing in effectiveness), so any card could give an effect for any Terror Level. Now there was just one deck of cards to shuffle - and dividers between Terror Levels avoided needing 3 separate stacks:
...which is the system in the final game, though there was still a lot more iteration in getting the individual Fear effects right. (There are a lot of very interesting / thematic effects which either aren't the right power level or are too swingy in how much they hit/miss. It's OK if Level 1 effects aren't too reliable, but Level 2 and 3 effects should be useful most of the time.)
Oh, and the front of a final Fear card looks like this:
Main victory: Early versions of Spirit Island simply had "Destroy all Cities" as the victory condition - the theory being that once the Invaders' main power was broken, you could mop up at your leisure. (And mechanically, playing whack-a-mole with a few Explorers in endgame wasn't super-fun.) Once Fear came on the scene, "Frighten the Invaders away" become a second means to victory.
Somewhere along the line, I started having trouble with certain Spirits being able to achieve blitz victories: if a Spirit was fast enough out of the gate, it could take down the starting City on its board before another could conceivably be built. For a while I just accepted it as a Spirit constraint, but it kept cropping up. After some brainstorming, I tied the victory condition to the Terror Level - after all, accidents do happen, and just because one City sinks into the swamp is no reason to abandon this lovely island.
This worked great (and integrated the Fear victory nicely). The only changes made from then on were details of what exactly you needed to destroy.
The Reckoning: For a long time, the Invader Deck had four stages: after the double-terrain Stage 3 cards were The Reckoning. Each one caused the Invaders to do something terrible - add a Blight to every land with a City, or destroy a Dahan in every land with a City or Town - and then test a loss condition of some sort. If you survived all the Reckoning cards in the deck, you won.
The trouble was, games hit that showdown vanishingly rarely: players would usually win or lose long before that time. I was concerned about stalemates with strongly defensive strategies (which I wanted to be viable) - but eventually concluded that "lose when the deck runs out" was simpler. Defensive spirits need to manage some sort of pivot to offense (be that damage or Fear) in order to win.
Bonus Victory Conditions: For a while, some Fear cards gave you bonus victory conditions - things like "Dahan outnumber (Towns + Cities) in all lands" or "Disease in every land with a City or Town" or "Earn 5 Fear Cards in a single turn".
The trouble was that how hard these were to achieve depended *extremely* heavily on the board state, the Spirits being played, and - most importantly - when you happened to draw them. Balancing them so they were interesting alongside the regular victory condition was nearly impossible. However, the core concept migrated over to Scenarios, where you know the alternate victory condition from the start.
As I post this, some Kickstarter backers have started getting their copies. Hooray! It's exciting reading reactions, and that many people are having a good time with the game.!
Mostly for design diaries and retrospectives, perhaps branching out into posts on more general design thoughts.
- [+] Dice rolls