I created a solution to the prior authorization processes in healthcare that has saved Americans several lifetimes worth of waiting for insurance approval for their medically necessary procedures, prescriptions, and treatments. What would otherwise be a several day to several week long process where the patient has no recourse but to wait and continue to suffer their illness without treatment, my products have shortened substantially, and sometimes even resulting in on-the-spot approvals before they have even left their doctor's office.
It's still relatively small (/early) in terms of rollout/adoption, but I'm always proud to be able to make a real, positive difference for humans in their times of need.
Not in a dismissive way like a statement of idle non-preference, but rather as a hyper-pragmatic statement along the lines of ''doesn't matter how weird/silly the solution, if it solves something important, then it is valuable"
absolutely! I was thinking about this also - there would almost certainly be some tradeoff between 'paying' for more information by attempting potentially disallowed moves. For this initial implementation, I figured I would just allow that to be balanced by (being lazy and not trying to code up some special way to handle that, and) just saying 'if you input a legal move, that is your move and no take-backs' :)
Yep, it's difficult to think how to handle impermissible moves gracefully in the general case.
I suppose that one option would be for the game state to diverge for each player, and only after both of them have reached a conclusion could they both be told the move at which that happened.
(and in that case, ensuring that both players reach their respective game's conclusion near the same point-in-time -- hiding the magic trick -- becomes a separate challenge)
I mean, technically i implemented it with a few different board states to help organize the 'common knowledge' board, the 'up-to-date' board, and each players' respective information. :)
Beyond that, what you described is basically how the implementation operates anyways. it notifies a player if they are put in check (without identifying the threatening piece, though I could change that pretty easily), likewise if theres a checkmate/stalemate. And in a situation where only one move is possible, it just auto-uses that move and notifies the player.
So the two players dont really 'discover' the outcome of the game out-of sync, per se.
An especially fun part is related to piece capture - eg if black captures white's pawn, then on white's turn, they will simply see their piece missing. So technically they would 'know' about the capture before their opponent would, which...I cant really even imagine the implications of yet, since I haven't had a chance to playtest this variant yet :)
While pokerstove does essentially the same calculations, I don't think it has a web interface. My goal was to make an easy to use app that didn't require you to install anything or have any programming background. But thanks for the note, and I can definitely use pokerstove in the backend to run faster simulations.
