In the original idea, the Christ child would be discovered as soon as the very last tile was flipped over. I play tested that my very first time, then realized that I didn't want the game to end four rounds into the game, so came up with a new end game condition. Inspiration hit me about it being a certain number of blessings a player got, which I first thought 20, then 10. Then I play tested it. And made it 6.
The original idea had the Christ child being found on the tile that fell mathematically on the star path where it first crossed itself twice, but then that was dumb so I made the whole roll a die twice thing, which I think is really fun because you can still use probability to figure out the most likely tile it will be, but it's still unknown until the very end. The idea for the Prediction board was new. At first it was only used for getting points at the end, but then the Shepard spots were just lame, so I came up with the idea for the bonuses you get.
The idea for movement changed before the very first playtest. My original idea was that your piece faced a specific direction on the tile, one of the six, and that when you moved you would have to rotate your piece to face a different side before moving it forward. And your cards would be like, 6 and 7 and stuff. But looking at the size of the board, I decided that that was dumb and instead of having you face a certain direction, I would just do you could move to any surrounding tile.
And your hand would consist of the Rest card as well as a 1, 1, 2, 2, 3, and 3, which then changed to just a 1, 2, and 3. The star path card was created to make the star path more important than just this weird end-game thing. And it worked. Suddenly the direction and path of the star path became a lot more important than just this thing that happened once at the end of the game. The star path became an integral part of the movement and play experience.
The idea for different types fo blessings, small, mediums and large, was in the original idea for the game. I'm not sure where the idea came from that it would be unknown how much you'd exactly get, like 1 or 2 points for the small blessings, 2 or 3 for the next, and 3 or 4 for the biggest. It kept things tight and tense to the end, not really knowing how much your opponents really had.
I knew that there had to be a reason to maybe not get there as soon as possible. But what? Then the idea hit me to have the Christ child's location revealed as soon as the 6th blessing was gotten, and to have the tiles keep their actions. Also, players would get a bonus point for every two gifts they had, plus another bonus point if one of their gifts was the one they started with. And giving them three turns allowed them time to quickly nab that last Donation Request card they needed to fulfill, which was a more fun alternative than "oops, I know you were just one turn away from getting 3 or 4 points but I just ended it sorry haha." And when it comes to game design the golden rule is this: Do whatever is the most fun. And the three turn limit still made getting there the fastest still matter, like you couldn't just say, "oh you got there, well now I'm going to take 12 turns and get three more blessings haha."
On the Prediction board, I used to have it that the first player to get the correct spot would get 4 points, the second player to guess it would get 3, and so on. Playtesting it, no one would go on the same spot because they all wanted that one extra point. I changed it so that the correct prediction would get 2 points no matter when you guessed it, and it was a lot better, because now all you worried about was the probability of a spot being the right one, which I think is really fun, and the bonus it gives you, which I also think is really fun. You no longer had to worry about getting to a spot first. So that worked out alright.
The tile bonuses. At first, the only reason the bonus was there was for you to get it when you flipped it over. That was it. No, actually, they were also usefully because I did have the rule that if you were on a pot and someone else came to that tile you would get the tile's bonus, making it so you wanted to go to tiles others would go to their next turn. Kept up the player interaction. Playtesting it, I realized that it was tough getting gifts if they only came from, like, three spots, and tried it out having it so that players could get the bonus instead of the action by paying a coin. And that turned out to be a lot more fun, and kept the options open more, so I kept it.
Oh, yeah, the starting number of coins was toyed with a bit. It was 8, but then I tested it and no on ever needed the extra storage space, so I changed it to 6. Still the extra coin storage space isn't as good as the extra gift storage space, even after changing the Trader's action from 'you can only sell one gift' to 'you can sell as many gifts as you want.' I made up for that a bit when I had the Inn card give you 2 points for all coin storages, but the one for all gift storages only gives you 1 point.
The Traders used to give you less money (1, 2, 3, and 4; which is now 2, 3, 4, and 5) and you used to have to spend more money to get gifts (2, 3, and 4 instead of the 1, 2, and 3). I changed those through play testing to make the game more fun (golden rule of game design).
The prices for the three Camel movement upgrades changed around a ton of times. As did the way the Rest to Rest-and-star-path power worked. It was took powerful, then too weak, then too powerful, until I wrestled it down to it only gave you the star-path movement if it was your last or second to last card in your hand. It was still not optional, until I play tested that and it wasn't fun at all, and in sticking to the golden rule I did what was more fun and had you moving along the star path optional.
The idea of having three tiles never used per game stayed the same, and allowed me to have some free space for ideas, giving me three more tiles to squeeze ideas into. In the end, it also creates a variable play experience, when you're never really sure which tiles won't be making an appearance this game. I also changed the frequency of tiles based on how useful they were. I only had two expand-storage-space tiles and three camel-upgrade tiles until I realized that storage space was better than camel upgrades, so changed it to 2 camel upgrades and 3 storage expansion tiles. The camel upgrades were 2 per tile, with 3 tiles, making the upgrades different per tile, but by making it 2 tiles I had to put all three options on each tile.
Anyway, I'm sure you're done with me talking about this, so I'm going to end my post here. Anyway, long story short, is making a game is lot of hard work. If you do end up playing it, let me know what you think.