No surprise I wanted to make a board game inspired by Carcassonne. That means a world that grows organically during play. In Edge of the Sky, you play an air balloon that is placed on the world. That balloon travels, and if it travels past the edge of the world, he has to add tiles around it's location, effectively growing the world.
Chance
Probabilities play a huge role in Edge of the Sky. There are two types of tiles: air and land. When land tiles are added to the world, it's added with it's ?-side up. That side lists all possibilities what the other side of the tile could be. My original idea was that after visiting the tile, depending on what kind of land you found would immediately translate into score. But that undervalued the strategic aspect of the game. Tell me, would you visit a tile which will score you either 2 or 5 points, or one with 2 or 6 points? Even while it's possible the 2/5 contains 5 and 2/6 contains a 2, players will always go to the tile with the highest values. Which makes the game boring...
I wanted to keep this probability-based mechanic, but this specific scoring system wasn't gonna work. So I changed it to a system that is best compared to a scratch card. (You have 6 or so scratch spaces, if you find the same number three times you win that amount of money. Often you scratch open all spaces, rendering the game essentially useless, but there are also variations in which you can't open all spaces.) In Edge of the Sky, islands can be a maximum of seven tiles, and there are four different land types you could 'scratch'. You can scratch for as long as you want, and since the probability indication is available, you can quickly estimate the chance that an island will be worthwhile. But nothing is definite, and seemingly valuable islands might attract concurrence.
This new scoring system added a whole new dynamic, namely that players will invade an island someone else is working on, and will claim the points as a result. This became the major feeling I wanted players to have, and I quickly began comparing it to the Indiana Jones antagonist Belloq, who steals the idol Indiana had just risked his life for to obtain. The problem was that players thought it was safer to keep exploring their own islands, thus never interrupting each other and making it almost a single player game! The major problem with this is that the distance to travel was too big, making it a non-viable strategy.
Interaction
I tried several things to change that. First was Boost movement- provided you only traveled on sky tiles for that turn, you could move much further than normal. This was an improvement, but since players couldn't enter land in this way, their plans would often still fall apart. I replaced this by rolling a die to determine the amount of fuel they could use. (Before this rule change, players always moved max. two spaces per turn.) Different types of land had different fuel costs, making it easier to travel through the sky and making land expensive. This was a good change that adds more luck: with a high roll, a player can execute a really cool plan, while players with a low roll aren't harmed by it because they are often still allowed to do an useful move.
The final addition was teleporters. These were added organically during the game (they appear on some sky tiles) and allow players to travel cheaply from A to B. This allows hijacking to become truly useful: since teleporters appear when players are close by, it can easily be used by other players to go there and invade the island, also adding more risk to adding air tiles to the map.
While figuring this all out, I also tried other ways to make players interact more with each other. One was the tile swap, where you could swap out an existing tile of the island and swap it out for another one. This feature was rarely used. Looking back on it, it was too much of a micromanagement thing, just swapping out a single tile. Not every idea you have needs to make it into the game. Later on, I found a much more effective way to get players to interact that would fit better in the metaphor. Players could attack other air balloons, pushing them away from them. While this too is rarely used, it is better and players seem to like it, often using it to get rid of their spare fuel.
Rulesheet
Explaining a game on paper is very difficult to get right. Once again, I used Carcassonne as example, in this case it's manual. Carcassonne explains itself in five pages, with lots of examples on how the game (especially scoring) actually works. Carcasonne uses bold text to note important information, or simply information you could easily miss (that is why the word NOT is often in full caps in manuals). It took me a lot of revisions to get entirely right. To make a good manual, Trello/Notepad simply isn't enough: I used Google Docs to format it and add tables that acted as explaining examples and for adding other graphs. Make sure you summarize the game early on, then explain how the game is set up (e.g. how the board must be built up) and then summarize the turn flow (how a single turn looks like) and THEN proceed with the actual full instructions, with examples for difficult-to-understand rules.
Setting
I wanted a slightly uncommon fantasy setting, which are floating islands, held aloft by magical forces. The original idea was that the world was just created and consisting of just a small patch of land, and the player to claim the most valuable land would be crowned ruler. But then the types of land were replaced by elements, of which you had to find three on a single island, and this theme would be unlogical. I changed it by making it the king's death wish that the one with the largest elemental power would become the next king, but that didn't feel like a very nice theme for a family friendly game, and the king's wish was a bit unbelieveable anyway. I did like the idea that players would become a king (or queen) upon winning the game, so I changed the story into a giant tournament that would be won and become king if a player can channel three elements, making it more lighthearted.
Edge in the Sky is a world-building game, but it has become an entirely different thing when compared to Carcassonne: it is a pretty unique game. I hope this post explains most of my design intentions. You can get the print and play version of Edge of the Sky here.