Help with decentralization of poker?

2»

Comments

  • PokerPlayerPokerPlayer Member Posts: 52
    It is difficult to jump in and out of sideways conversations, but I don't mind trying.
    I am assuming that there is no "RANDAO" and that an external decentralized random seed is not available. I do believe it is the most difficult and important advance society could make and so I think we underestimate the power of the most "truest" randomness available. For this I wonder if rather it is another economical problem in that you must combat those that want to corrupt a random pattern and those that have incentive for a "truer" or fair randomness. If this might be true then I think I have some ideas and it might be that poker players for example could be used to "financially" keep together a random system.

    I do agree with much you allude to about using a blockchain for verification, it all seems quite simple and direct (from a non programmer side). As for "colluding", yes its important to separate it from the problem of gameplay because it is not realistically a technical problem in the programming sense-it has its own solution separate from decentralizing the deck.

  • PokerPlayerPokerPlayer Member Posts: 52
    FrankHold said:

    Maybe we have to split the task. Let ethereum do what it can best – transfer and organize the money flow. And let a poker routine run on a classic server outside of ethereum.

    Its not really acceptable in the spirit of decentralization. Player will want a mental poker process that can be completely secure.
  • innovator256innovator256 Member Posts: 5
    Hello there interesting discussion you've got going on there I have worked on this problem for more than a year now and I have detailed an extensive plan using smart contracts to provide a solution to the mental poker problem. There is also a team working on the project along side a client and a DAO in formation to manage the poker DAPP. I have posted the project as a new discussion , hopefully it gets approved soon. But in the mean time yo can find it here : http://www.reddit.com/r/ethereum/comments/2wumjj/pokereum_an_ethereumtelehash_based_provably_fair/

    We should all collaborate to make this happen as a community, each person contributing where they can. Also free tokens :) Thanks
  • PokerPlayerPokerPlayer Member Posts: 52
    edited June 2015
    I didn't fully understand the pokereum project. My understanding that I had plus my understanding of the poker economy and the future of it led me to this:

    https://i1.wp.com/s23.postimg.org/69611n4ff/DPoker_Diagram.png

    And this explanation:

    https://thewealthofchips.wordpress.com/2015/06/15/3194/

    Just posting it all wherever relevant for discussion.
  • PokerPlayerPokerPlayer Member Posts: 52
    Another thought I've been pondering:
    We give power of the network to the losing players. This is actual quite interesting and as I understand it NOT Pokereum’s solution.

    Interesting thoughts arise…and we start to wonder if that could possibly be the solution WITHOUT any sort of system of public/private tables etc.

    We think about a certain malicious party trying to lose enough money to take over the network…impossible or favorable for the players in the long run, in order to secure a network that will lose its integrity ultimately anyways.

    A malicious party losing to only its own accounts cannot be a very net loser.

    We can also think about possibilities for bots and bot pools to arise that are “losing players” on average, but receive enough supplemental private game rake (in exchange for acting as a random node) that being a slight loser plus the additional control of the network might be enough secure and equilibrium.

    A winning bot must necessarily PAY for some control of the network.

    A losing player gets a piece of the network and they can sell it.

    You want your poker network/stake to grow or shrink in relation to “domestic deposits” or “lost monies”.

    The incentive to own stake in the network can be collecting “rake” by acting as juror for private games.
  • PokerPlayerPokerPlayer Member Posts: 52
    Here's some somewhat organized infor and points. Basically things seem still stuck in design phase, but it seems to me that lightning channels might be quite applicable, and cancel a lot of complexity out:

    A Relevant History of Mental Poker:
    https://thewealthofchips.wordpress.com/2015/09/21/4022/

    A Relevant Overview of Mental Poker Implementations:
    https://thewealthofchips.wordpress.com/2015/09/21/a-relevant-overview-of-mental-poker-implementations/

    Solving Implementation Problems of Mental Poker via Lighting Channels: https://thewealthofchips.wordpress.com/2015/09/21/solving-implementation-problems-of-mental-poker-via-lighting-channels/

    I don't understand them perfectly though but I am going to go through the examples soon and relate them to a poker hand the best I can. I am anticipating a possible issue I am not sure we have but it might be a simple algorithm can be set as the final issue of reneging is solved.

    A Possible Solution to Habitual Renegers:
    https://thewealthofchips.wordpress.com/2015/09/21/a-possible-solution-to-habitual-renegers/

    It might be wrong to think but it seems to me simplicity is a hint things are developing in the correct direction. Any thoughts are very welcome!
  • PokerPlayerPokerPlayer Member Posts: 52
    Also this:

    Two Possibly Useful Thoughts for the Implementation of Mental Poker: https://thewealthofchips.wordpress.com/2015/09/22/two-possibly-useful-thoughts-for-the-implementation-of-mental-poker/

    I'm just brainstorming. I don't see any implementations, so I am gathering the info and thoughts I can.
  • AFDuldeyAFDuldey Member Posts: 3
    I was working with a team of people to solve the "poker on ethereum" problem. There are a number of teams working on it now. I assume a solution will be available soon. :D
  • PokerPlayerPokerPlayer Member Posts: 52
    Is that promising?
  • PokerPlayerPokerPlayer Member Posts: 52
    I didn't see this before:

    “How to Use Bitcoin to Play Decentralized Poker” proposes a protocol solution whereby any player that refuses to participate for selfish gain forfeits a substantial fee to each player that is playing honestly. This removes any possible economic incentive for such dishonest action.

    http://people.csail.mit.edu/ranjit/papers/poker.pdf

    To me this means the philosophy is complete:

    There is still the difficulty, from the players perspective, in understanding how a P2P environment can police collusion, bots, hud’s, and other types of cheating. A secure implementable P2P mental poker protocol provides the necessary infrastructure and tools for many competing poker sites to arise and to solve these problems. Rake can be paid to these sites to provide any, all, or none of the historical solutions that the centralized server model generally included.

    This suggests a conclusion in this direction still using third parties for realistic implementation, but not for trust in the tradition way:

    The introduction of a secure mental poker protocol which handles security of chips, dealing, gameflow, and cashout/deposits removes the need for a trusted third party. The problem of player collusion is solved removing the barrier to entry to provide a poker site solution thereby fostering a competitive market place for players. The implementation of the protocol is secure, trustworthy, allows effectively instant game flow, and is exponentially cheaper for emerging sites who can now offer a superior product at a reduced cost because of a drastically reduced overhead.
  • PokerPlayerPokerPlayer Member Posts: 52
    ^The insight here would be p2p poker would be initiated by a protocol enforced by a contract, and all of the other solutions come from innovation. If Satoshi introduced a p2p money that projects have been developing services for, Poker Satoshi would be offering cards, chips, and game flow, for Poker Sites to build their services around.
  • PokerPlayerPokerPlayer Member Posts: 52
    I don't know if ur really solving the problem at all by not at all solving it though. Also you are wrong about collusion and I certainly have a solution for that aspect, but I'm getting the impression you aren't interested in reading.
  • PokerPlayerPokerPlayer Member Posts: 52
    ah sorry, wait....g416g are you a programmer? maybe you are just not very familiar with the problem. perhaps you are stronger in the language and translation, then we could be helpful to each other.

    But please...don't tell me your solution to decentralization is centralization... we won't get along... :) just allow for the possibility I have insight in this regard
  • PokerPlayerPokerPlayer Member Posts: 52
    Thx. Lemme know when ur caught up: https://thewealthofchips.wordpress.com/

    You are starting from a conclusion, and its not correct. Use the search for "collusion" as well. But don't ignore the rest of the material :)
  • PokerPlayerPokerPlayer Member Posts: 52
    I'm not explaining anything unless you show sincerity. Why would I argue with a conclusion?
  • PokerPlayerPokerPlayer Member Posts: 52
    edited March 2016
    it's a sincerity filter.
    Post edited by PokerPlayer on
  • PokerPlayerPokerPlayer Member Posts: 52
    edited March 2016
    From the players perspective if I can offer a more profitable game that is colluded vs one that isn't then I have effectively solved collusion.

    There is nothing else to be solved in that regard. You choose to look at the problem from the wrong perspective. Like saying the Byzantine generals problem is unsolvable, yet here we are: https://bitcoin.org/bitcoin.pdf

    Simply higher order solution.

    Spending your time as a skeptic is wasteful.

    ps. btw tell me how it breaks down ;)
  • PokerPlayerPokerPlayer Member Posts: 52
    edited March 2016
    You side stepped my point, that if i can provide a colluded game that is more profitable than your centralized non-colluded version, I have effectively solved what you suggest cannot be solved. I haven't given my proposal. You need to acknowledge what is obviously and perfectly true (whether you think I can provide it or not).
  • PokerPlayerPokerPlayer Member Posts: 52
    edited March 2016
    I am asking you to acknowledge (or publicly deny), that if I can provide a colluded game in which players profit MORE than a non-colluded game then for all intents and purposes I have dissolved the problem of collusion.

    It's quite a clear and succinct point, I should think we do not need to dance around it. If you can't even understand this small point, how could I get you to understand something more complex?

    (also effectively does not mean "in my head" by any accepted use of the word)
  • PokerPlayerPokerPlayer Member Posts: 52
    edited March 2016
    I'm not moving on whatsoever until you can at least acknowledge what any rational person would. I assert: A colluded game that is PROVABLY more profitable than a "fair" game effectively dissolves the issue.

    If you can't see that, you won't understand going forward. There is not getting around this simple statement.
  • PokerPlayerPokerPlayer Member Posts: 52
    edited March 2016
    That is what I thought. I'd appreciate it if you left this thread alone, thank you. good luck!
  • PokerPlayerPokerPlayer Member Posts: 52
    Please no more posts in my thread!

  • PokerPlayerPokerPlayer Member Posts: 52
    This might be interesting to those that are interesting in poker on a blockchain. Just a little tour through mental poker with examples and links to the calculators: https://medium.com/@rextar4444/how-to-play-mental-poker-a-tutorial-by-the-author-of-cypher-poker-loosely-edited-by-twoc-j-smithy-8c1c6f7a8259#.pxhj1tym0
  • PokerPlayerPokerPlayer Member Posts: 52
    edited April 2016
    This post is an extrapolation of this dialogue: https://medium.com/@rextar4444/pb-and-js-discussed-some-fundamental-concepts-of-mental-poker-implementation-61025630141d#.t5xv7o7dz

    There are 2 types of collusion that are relevant to a successful implementation of p2p poker:

    1) collusion such as sharing hole cards
    2) collusion such as not handing in private keys

    In regard to 2:
    We might assume that the table is full of malicious players except the hero. The worst scenario would be: all players are all-in, hero wins the hand, but 1 malicious player refuses to hand in a key in order to verify the game. |

    Hero could have potentially won his initial chipstack times the number of other players. The malicious group must not have anything to gain from colluding in this fashion: 1) they must not have any monetary gain 2) The honest player therefore must get back at least what he was supposed to win.
    We can pose the problem succinctly like this:
    The escrow needs to cover the worst case scenario: the maximum # of malicious players (n-1 for a table of n player) times the maximum amount the honest player can lose (total equity or chipstack).
    And so the succinct solution would be, at a table of n players, that escrow for each player is simply:
    (n-1)(chipstack)
    Does this seem to satisfy "2)"?

  • PokerPlayerPokerPlayer Member Posts: 52
    "CypherPoker is now able to play a game through to completion with full Ethereum support — escrow, status tracking, validation, and payout; the works."

    https://medium.com/cypherpoker/cypherpoker-v2-0-update-5-79192123fa5f#.7prg3w6xl

    This is the project I am most involved with/supporting and believe in.
Sign In or Register to comment.