The problem of Mental Poker is fascinating to me. I am the guy over at 2p2 that recently bumped the old decentralized poker thread and got it going again. I'm glad that several projects have popped up since then, as I feel like this is a "killer app" for cryptos if done correctly. I'm not sure if some or none of these projects got started because of the attention I brought to the subject, but that was my intention and it happened nonetheless, so I am happy about that.
I am worried how fast the new projects have got started and are already collecting funds, because I feel like Mental Poker is possibly not sufficiently solved quite yet. Yes, you can play a game of Mental Poker, and you can prove the validity of the shuffle, deal, and actions. However, I have deep concerns about the efficiency of the Mental Poker Framework to provide a fluid real-time game of poker like people are used to on centralized sites.
I direct you to Sergio Lerner's patent to learn more about the MPF. He spent years researching, developing, and experimenting with Mental Poker and I consider him to be the utmost expert on the subject (or at least the most knowledgeable person on the subject that I know of). I will cite a simulation he did with different MPF protocols and the time it took to verify the results. Although he improved on efficiency of prior protocols, it still seems rather slow. You may see the patent here:
https://www.google.com/patents/US20110202766?dq=sergio+lerner&hl=en&sa=X&ei=vv6gU9a4GdSuyASNxoDQBA&ved=0CDMQ6AEwAwComparison of Protocol Times
MPF MPF MPF
over over over MPF
ECC ECC PH over PH
base base base base
VSM- VSM- VSM- VSM-
VL VL VL VL
using using using using
Operation CO-VP HMVP CO-VP HMVP KKOT90 BS03 Cre86 CSD04b
Shuffle Time 14.61 s 27.31 s 36.26 s 57.97 s 333.80 s 273.39 s 415.54 s 102.29 s
All cards draw 0.17 s 0.17 s 0.43 s 0.43 s 21.00 s 35.94 s 17.28 s 46.29 s
time (5 cards
for each
player)
All cards show 0.12 s 0.12 s 0.39 s 0.39 s 0.78 s 46.30 s 0.08 s 46.30 s
time
(showdown)
Total 14.90 s 27.60 s 37.08 s 58.79 s 355.58 s 355.63 s 432.89 s 194.88 s
processing
time for first
game
Amortized 1.60 s 2.87 s 4.06 s 22.79 s 319.58 s 319.63 s 396.89 s 158.88 s
processing
time per game
Sorry that it jumbled up when I copy and pasted it.. refer to the link I provided.
As you can see the most efficient MPF protocol (that I can find) takes 14.64 seconds to shuffle and, more worrisome, it takes 27.6 seconds for a showdown. Obviously this slow of a speed would greatly hinder the game flow of a hand of poker. Consumers will not want to wait 30 seconds after each hand to find out who wins if it goes to showdown, or 14.64 seconds to shuffle.
Have you thought of this problem? If so, how do you plan to improve upon it so that the MPF could provide a playable game in real time, with a normal game flow that online poker players are used to?
Secondly, collusion is another big problem with decentralized poker networks. How do you plan to combat against it? Obviously playing heads up games, "fast fold" (Rush) poker, and multi table tournaments only would pretty much solve the problem of collusion. However, I feel that is only a half-assed solution and I would like to see something better than that come along where people can play full/6max ring games and single table tournaments. Providing heads up games and multi table tournaments only serves a portion of the games available on centralized poker networks and I'm afraid if this is not solved this will not be the "killer app" that I think it may be.
I feel it is important these questions be asked and answered before more funds are raised and you possibly realize that what you plan to do is not really practical competition to current centralized poker sites due to the complications. I am not trying to cause you harm, I support your project and am very interested. I want to see you succeed, and this is why I'm asking these questions.
Thanks for your time,
Ch