It’s seems the philosophy of LC’s suits a mental poker implementation perfectly, with the assumption there is a sufficient escrow from all players that creates a disincentive for a player to hide their private key/secret (at anytime during or at the end of the hand).
Each player, on their turn, adds their action to the hand-history, sign’s it (hashes it with a private key), and passes the action and signed commitment through a fast messenger service.
There is no problem of integrity for the game because a challenge can be verified with a smart contract and escrows lost from the player that breaches the protocol/rules.
An action can be assumed verified if the next player acts upon it so the only consideration is ‘the player who’s turn it is and the next player to act after them’.
The problem then is only when a player, who’s turn it is to act, claims to have sent a fast msg, but the player next to act also claims to have never received it.
We can’t figure out who is the malicious player in any meaningful way, either the player to act or the subsequent player.
However the malicious attack is limited only to encumbering the game flow until the actions/transactions are confirmed via the block (and in relation to time allowed in a poker game to make a decision). <<< sorry for bad technical knowledge here.
In other words the only attack possible is timebanking every hand.
I wonder if problem is well encapsulated and explained because I think there might be a natural solution to it.
(some thoughts on different types of type arise such as being able to maliciously stalling the game beyond what the rules of the game normally allowed for a timebank)