So I have to ask, out of curiosity and because I'm attempting to do a simple bot for fun, starting with bilging because some on this forum said it's easiest...

Does the depth function do what I dread to think it does? As in do a swap of every location, with 4 internal iteration per possibility aka the entire board worth of swaps to the power of 4?

And keep track of the best location score to find out which one scored highest, effectivelly finding the best 4 depth move possible - but then you redo it after every time the user makes a move?

I mean it's about 15k combinations each time, or is it more? Is this the right approach to a simple bilge bot?

I can't imagine it would yield donkeys and vegases with any consistency however, I know that building those usually takes well upwards of 4 moves, and it requires them to be focused. Oceanus doesn't seem to work this way, and has a better shot at donkeys and vegases, what's the difference? Any speculation?