Way back in 2011, I had an email conversation with Jeff Allers in which he told me about his game Heartland. It's a tile-laying game with a farming theme, with the twist that tiles are placed on top of other tries to simulate replanting a field of one crop with a different type. It's just been reworked into a feudal Japan theme and published by Renegade as "Gunkimono".
Anyway, what this discussion sparked for me was the awareness that tile-laying games don't have to strictly remain two-dimensional. And thinking in a third dimension led quickly to the thought of a 3D game in which you start with a stack of tiles and remove them progressively to uncover what's underneath. This obviously suggests an archaeology theme, and that's obviously where I went with it.
For some reason, I decided that because worker placement games were becoming the popular thing at that time, I needed an elaborate WP system to give you a lot of different things you could do. It actually probably could have worked, but digging was very swingy, in that you were rewarded for whatever you uncovered as you removed tiles, and also the game was a bit dry. So it got shelved.
Last week I had an idea for a mechanic that I haven't seen previously: roll dice and assign them to different actions, and then resolve actions in die order. Thus there needs to be some reason why you want to go early in some actions, but also that there's an advantage to going late. I quickly whisked through all of my design ideas for which I've hit a snag and thought about whether this could possibly unsnag any of those games. And luckily, I think for the archaeology game, there's a potential fit.
The idea is this: the stack of tiles is laid out in a 4 by 4 grid, with tiles having 1, 2, or 3 squares. (All tiles in a layer are functionally the same) At the start of each turn, randomly activate three sites in the grid (e.g. C3, A2, B1). Then each player rolls dice and assigns one die to each active site. Then, resolve actions in die order. (If two players pick the same die on the same site, just use a 'tiebreaker track' to settle who goes first).
Each tile, when it's first uncovered, receives a number of cubes, randomly drawn, in an amount equal to the level on which the tile is found. These come in four colors, representing different kinds of finds: artifacts, inscriptions, human remains, and structures.
There are five different actions you can take when your turn comes up: (1) remove the tile at that site (costs money), (2) sell one of the cubes for money, (3) pick one of the cubes and increase your "knowledge track" in that color, (4) publish a book about one of the cube colors, or (5) announce a Major Discovery that includes one or two of the cube colors.
Here's the twist: for options 2-5, the 'value' of the action you're taking is buffed by the number of dice on the tile. So if you're first to act on the tile and (for example) choose to sell, you'll have your choice of which color to sell but won't get much money for it. Whereas if you act later, there will be more dice on the tile and thus selling a cube will be worth more, but the cube selection may be more limited at that point.
This is a pretty simple turn mechanic so I'm optimistic that it could work well, and certainly it's an enormous improvement over the bloated complexity of the WP conception of the game. Possible concerns include that the randomly-assigned dig sites will be too restrictive, that the die-rolling will inject too little player control, and that the game won't play equally well at all player counts. But we'll leave it to playtesting to determine which of these concerns materialize and how to address them. A bigger concern might be that the earlier concern, that the game was rather dry, may still exist: there isn't exactly an element of removing a tile and finding something really great underneath it. Instead it's about positioning yourself to benefit from whatever is uncovered, and to decide which sites to prioritize each round. I think it will be fun but definitely not the "thrill of discovery" fun of Thebes for example.
For now, I like that both the physical arrangement of the tiles and the turn mechanic have not yet, as far as I know, been done previously, but both should make pretty good intuitive sense so hopefully the game is easy to learn and play.