A Relevant Overview of Mental Poker Implementations

After taking a look at A Relevant History of Mental Poker one might be inclined to ask if there are any viable implementations of the solutions proposed. First we might look at one last paper and one special patent to get a little more understanding of the difficultly of implementing mental poker.

Mental Poker Over the Internet

A Secure Mental Poker Protocol Over the Internet attempts to further address the problems of actual implementation of mental poker.  The obvious application of this implementation would be internet poker but this implementation too comes with its own problems.

The suggestion here by Weiliang Zhao, Vijay Varadharajan, and Yi Mu is zero-knowledge proofs proposed by Crepeau:

…are “not practical and are often very complicated and messy.  They need a fairly long computation time to shuffle a deck of cards.

The proposal to deal with this and other issues is the solution whereby a dealer collects the private information at the end of the hand, and only the dealer ever knows every players strategy.  This point may or may not prove useful in the future but for now avid wealthofchips readers or cyrpto enthusiast already have alarm bells going off since Trusted Third Parties Are Security Holes.

If we can understand a poker hand to be some kind of complex transaction (with some form of probabilistic events involved) it shouldn’t be hard to see the poker does in fact suffer from the same byzantine general’s problem that that a p2p e-cash system would. Poker sites do transact a lot of value throughout the world and in this regard they function nearly identical to banks.  But if one wants to create a bank, or implement a solution to a p2p banking system it should be hard to convince one that this solution must be ultra secure-it must literally stand up to armies and attack of the largest magnitude.  A trusted dealer cannot do this, nor could a trusted messaging channel or payment channel. This is a clue that Satoshi’s solution to the byzantine general’s problem is, in some form or another, a necessarily component to the implementation of mental poker.

But for now we are still exploring the “decentralization of cards” or rather the actual dealing and card play protocols themselves.

In “Contributions to Mental Poker” Jordi Castell`a-Roca addresses this problem as well as some other key points from the above paper:

In Chapter 7 we have presented a new TTP-free protocol which allows the game to continue after player dropout without diminishing player security.  This is an unusual functionality. Our solution is based on zero-knowledge proofs, while previous dropout-tolerant proposals were based on secret sharing (cards were encrypted with a shared secret key and, if some player left the game, the rest of players could still decrypt the remaining encrypted cards).  One drawback of using secret sharing is that a sufficiently large collusion of players can obtain all cleartext cards; another drawback is that the number of dropouts that can be tolerated is fixed. Our new dropout-tolerant solution does not allow any player coalition to find out the cards in the hands of the rest of players. Moreover, the number of players that can leave the game is not limited.

A Mental Poker Patent

Sergio Lerner feels he has a solid enough solution and implementation to patent his design. Lerner has some ideas on a type of coin and possibly blockchain that might facilitate this implementation (Quix Coin slideshow) but the author here has not seen or heard of the implementation of either of these projects.  Lerner did admit, sometime within the last year, he is working on multiple projects that are implementing these advances (or some evolution of them probably).

The Two leading Mental Poker Implementations and Their Restraints

The two leading Mental Poker Implementations the author knows about, and feels are most worth mentioning are Pokererum and Pangea. Both existing in the forum of nxt.

Pokereum

Pokererum has a nice 30 page whitepaper with a lot of promise but also a lot of promises.  The paper focus’s on a random player seating algorithm that uses player heuristics, a randomly selected, jury pool, proof of stake, and smart contracts, as a solution to the difficult problems a mental poker implementation faces.  If we take the solution as valid, there IS still the problem of only seemingly offering randomly seated tables (although this might be somewhat mitigated by the possibility of facilitating private tables).  The solution is succinct in some ways but all in all there needs to be much development done and a proof of concept before the system it creates can be called sustainable.  Vitalik Buterin did quickly comment

I’m not comfortable with this, it sounds very gameable…~Vitalik Buterin

The author here proposes a possible fix to the Pokererum solution that might have an interesting ramification, but it is still dependent on Pokereum reaching launch.

Pangea

Pangea’s solution is less intricate but admittedly requires technology not yet developed. A head developer has been tasked with created a 3 second blockchain to facilitate gameplay.  Theoretically this solves many or all of the mental poker implementation problems, however it is yet to be proved or experimented on in relation to how fast a blockchain could should be in relation to the security needed for the transactions it holds. Contingent on such a development mental poker is of course quite solvable.

Do we have the necessary components for poker’s version of bitcoin: secure p2p poker with no trusted third party?

It seems we are close enough to a solution that there are already multiple groups working on the development of a practical mental poker implementation, however, unfortunately such an implementation does not exist today. One such group seems to suffer from the issue of only presenting a niche solution (randomized table seating) and the other group suffers from the lack of an existing block-chain that is fast enough to facilitate a poker hand at a reasonable real-life speed.

This is where the author would like to turn to the possibility of lightning channels and their evolution duplex channels for a possible solution to either and both of these issues.