The former already planted the idea of making a game about the ancient seafarers who braved the unknown oceans to discover islands to settle. It was a great feat, considering they not only sailed into the unknown, but also navigated and established at least a few known routes — all without a compass. Instead they relied on the sun, stars, the color of the water, and even the animals of the sea.
The latter inspired the idea of making a trade route game, that is, a game in which you'd establish routes. If I'm not mistaken, that earlier game uses two currencies, one of which occasionally spoils. I liked that idea so much that I immediately stopped reading anything else about that game so that I could implement the core idea how I would like to use it and not inadvertently copy mechanisms from 'Ohana Proa. (As a side note, when Twilight Struggle came out, I was already working on Wir sind das Volk, so I decided not to read anything about Twilight Struggle until I'd finished my own game to avoid "contamination".)
At this point I had the following basic ideas:
1) Players should have a number of people that they have to send to the islands. The design should not be a majority game since while the players are competitors, the actual people — that is, the Polynesian seafarers — are not. So you should get a point for every island you have at least one person on, with bonus points for islands that are further apart.
2) The players establish routes between the islands and can move along those routes. Moving should be for free (costing only an action), but establishing the routes should cost a currency.
3) There should be two currencies that regularly spoil.
For idea number three, I used my mantra: "Make it as simple as possible." Instead of me creating an elaborate mechanism that details what spoils when, the round's starting player names one currency — fish or mussel — and all players then give up everything they have of that currency. This process elegantly makes any sort of hand limit unnecessary.
As a core mechanism, I used an engine from a discarded game that never quite worked: Each round runs through three phases, aptly named III, II, and I. The roman numeral indicates how many movement points you get if you decide to move in that phase, how many meeples you can place on the board if you choose to do so, and how much a connection costs if you want to built one. The order is reversed since it would be a no-brainer to build first (when it's cheap) and move later (when you move a lot); switching the order makes the decisions much harder: Do you build first? Then you may block players, but you've paid an awful lot for the connection. Or do you put more people on the board first?
The first tests...were not great. The problem was that players got stuck pretty soon. Clearly something was missing. Routes shouldn't block other people permanently, or else everyone becomes stuck in their respective corner of the map.
The solution was twofold, but for both I used my previous guidelines: 1) There are two currencies! Use them! and 2) The Polynesians were friendly people!
If you establish a connection for the first time, you can use whichever currency you want, but you mark this currency on the route, and if someone wants to establish the same route later, they have to pay you for the knowledge in the currency that was used. That way you can use currencies that other players have less of in order to complicate their lives.
But the game still wasn't dynamic enough until I had a lightbulb moment: If you are on the same island as another player's pawn, you can pay that player to force them to bring you over their connection to another island! Both pawns will be moved, which can be helpful or hurtful for the owner of the ferryman (depending on their plans). That's a fun and thematic mechanism — and it was the missing piece in the overall design.
There were still minor issues: Despite spoilage, the game gave players too much fish and mussel, so I just reduced the islands where you get them as income (with "just" equalling a couple of months before I knew how to tackle that problem).
The endgame involved too much calculating, and the first half of the game lacked enough variety. The latter was solved by a Kingdom Builder-like scoring with three cards that change every game, which adds a lot of variety. For the former, I stepped away from the true history of the region in which the game is set and adopted a more "Inspired by true events" vibe, stating that players want to leave the starting islands because a volcano is about to erupt. The round in which the volcano erupts is random, although within a smaller interval, so there is an element of uncertainty, but it won't catch you off guard.
After all this designing, I wanted to stress test the game at the UK Games Expo in 2017 in the Playtest UK area. After the game a playtester said, "To be honest, this is the best game I've played on the fair so far. Every decision is meaningful, every decision is hard, there is a lot of interacting. I think you are already there!" I took this as a good sign...
I'm quite happy with the result of the Polynesia design, which reminds me of King of Siam in the sense that it's a very deep game with a small rule set and a sixty-minute playing time with a lot of interaction and nearly no random factors. They play completely different, mind you, but I daresay that if you like one, than you probably like the other as well.
Peer Sylvester