Designing a game is a journey. It starts with a spark, a catalyst. For Cursed Court, it was playing a prototype of a poker-like game that used the Fibonacci sequence. It reminded me how much I like Texas Hold ‘Em. I love the interpersonal challenge of trying to read, understand, and predict people. I also love the accessibility. It's a game I can get my non-gamer friends to play.

I was also reminded of my biggest problem with Texas Hold 'Em: It’s nearly impossible to get people to take it seriously without money. The game simply isn’t engaging without an external motivation. I think the biggest factor is that players must fold a lot to play well, which robs the game of drama.

So I set out to create my own game. I set my sights on a design that didn’t push good players toward folding. I started with a small deck of 13 suitless cards and said the highest card wins. I then gave the lower value cards special powers to increase their chances of winning. I tried a lot of different powers. I made Merchants that were stronger if multiple copies were in play. I tried an Assassin that would win if only two players were left. There was even a Priestess that would make the lowest card win the hand.

That version succeeded at increasing the odds of winning with any given hand. It also succeeded at teaching me a very important lesson: The harder it is for you to judge the strength of your hand, the harder it is for you to wager well. In other words, the closer I got the possible hands to an even win rate, the less interesting the betting became.

It became clear that wasn't going to work. The goal I had set out with was fundamentally at odds with the rest of the game. Fixing it took two large leaps of faith. The first was to replace a fundamental aspect of poker, the ranking of the hands. If you think about poker hands, they’re really just a complicated set-collecting game with unbalanced odds. I wanted something more balanced. I found it in a tic-tac-toe grid. I drew up a little diagram with the cards arranged in a 3x3 grid, and I told players to forget normal poker hands and instead try to collect three cards that would form a line.

Players clearly found a glimmer of potential in this new ‘Frankenstein’ of a game. But much like most amalgamations, it had some grave problems, serious enough that I needed a second leap to make any more progress. That progress started with removing the numbers on the cards. I also enlarged the 3x3 grid to make it big enough to allow players to place bids on myriad different combinations of cards, a little bit like roulette.

To make that fun I needed to make one minor tweak to the scoring of a hand. Rather than only being able to create a hand out of your own private cards and the community cards, I changed it so you can use anyone's private cards. This added a whole new aspect of deduction. You now had to guess what cards other players had based on how they bet.

Even with the core deduction on the 3x3 grid, there were still a lot of missing pieces. How should bidding work? Could you share spots, or bump players off? Was scoring winner-take-all, or did players split the pot? What if you didn't have enough money to bid on the location you wanted?

These questions all took many iterations to work out, and they left many good variants that are now in the back of the rulebook. And while figuring those out was fun, I still marvel at how far the game came from its original spark of an idea. The original poker-like version of the game served as the phoenix that had to die and be reborn before the final version of Cursed Court could take flight.