Topic: Poker and the shared pot at the table in a decentralised network - page 2. (Read 14727 times)

Well, that's interesting! Looking forward to seeing what you come up with Sergio. You might get the honour of making the first alt coin that's actually technically interesting (well, NameCoin was also quite interesting I guess).

That said - a patent? Why? Do you think that's for the best?

Are you planning to implement atomic chain-trades with bitcoins to obtain your new coin?

For an update on the feasibility of online p2p poker, I'm  working on QixCoin.com since half a year ago.
It solves all poker-related problems: collusion, pot, etc.
It does not require a TTP nor an escrow: it's pure P2P.
It's based on my patented (pending) mental poker protocol.

The problem with using an escrow is that, under many regulations, the escrow service is facilitation online gambling so it must have a license to do it.
So nobody would be able to (legally) provide such escrow service.

Since the protocol itself handles all money-related actions, legal responsibility relies solely on players.
And that it the reason why I had to create still another alt-coin, since Bitcoin scripting language could not support game specific verifications.

QixCoin is salable and performant (the block-chain is NOT cluttered with ZNPs). Bet payments are done off-chain instantaneously. Pots (new games) can be created in less than a minute (the block interval is as low as 30 seconds). QixCoin code is not based on Satoshi Bitcoin code (I think it's the first alt-coin built fully from another source tree).
 
If you have a specific question regarding how I solved some of these hard problems, please ask me here on in another forum, since it belongs to an alt-coin.

Best regards,
 Sergio.

I'm doing a related experiment. I have a thread in the games and rounds section for a correspondence style forum based poker with 72 hour time outs.

However, it is not strictly Mental Poker, but an implementation of what I consider as Provably Fair shuffling, with me as a non-playing dealer. Someone mentioned escrow. That's also my function.

I'll try to get the link, but I'm on an iPad now. I didn't want to necro this thread but, someone did it already.

Lately I've been exploring how to use SCIP/TinyRAM with e-Passports. The idea here is that modern passports have a chip inside them that can be read using an Android smartphone. The data is signed by your government and is thus hard to forge. You could use a modified/improved version of SCIP to create a proof that you have a valid passport that hashes to X (including a public key), thus creating for yourself an anonymous identity that is nevertheless as "bannable" as a real life identity.

It could be applied to poker like so. Your software searches out other people who want to play. They are identified only by their hashes+SCIP proofs (i.e. large opaque numbers) and a nick they chose. If there was a suspicion of cheating, you could choose to add the player to your own personal blacklist. People could share or publish their blacklists of cheaters (assuming they get caught) to make it harder for them to re-offend. People are nevertheless still anonymous.

If you wanted to, you could run games alongside a group video chat to raise the bar for cheating a bit further (i.e. people are not actually sitting next to each other, you can see their hands+keyboard), and you could pay someone independent to randomly audit peoples computers using screen-sharing software to verify they weren't communicating.

All this is a lot of work though.

Do you think that with current knowledge this is now possible to develop it?

A created a new thread for the MAVE paper just published. See https://bitcointalksearch.org/topic/mave-digital-signature-protocol-for-massive-bulk-verifications-76073

The patent was applied to sell the technology to some online casinos.

I would issue a unlimited license for the open source network that solves the p2p poker problem right.
But it can't be GPL because GPL has some issues regarding patents..

The MAVE protocol is unpatented (and, anyway, it's no big deal)

Sorry for the out-of-topic response.

Sergio.

Too bad then. I don't see many people willing to accept an alternative currency that is produced through a secret process.
I thought first that the patents he was referring to were only for the poker (not that I approve them for any purpose, I'm plainly against so called "intellectual property"), but it seems that I misread him. The currency simply NEEDS to be open sourced.

Sergio, I invite you to read this:
http://www.gnu.org/philosophy/not-ipr.html
and this:
http://mises.org/journals/jls/15_2/15_2_1.pdf

Two different and strong perspectives on "intellectual property". I propose that we start calling them Intellectual Poverty laws.

Sorry for going off-topic.

Even if it is "free" as in free beer it won't be "free" as in free speech.

Sounds great. Yes, please post here.
It will be free software, right?


You mean something like standardizing merged mining so that there's no "father chain" ?
Maaku said that his company will launch a bunch of currencies that merge mine that way. The first, "freigeld" (very similar to my freicoin proposal) will be launched soon. You can find maaku in this forum or in freicoin's forum, in case you want to know more, but I don't think he will give you many details before the launch (which is supposedly imminent).

The largest issue remains collusion between players a new blockchain doesn't solve that.

You sit down at a virtual table and 5 of the 9 oponents are all part of a team sharing cards, strategy, etc.  You have no hope of winning.

Two years ago, back in 2010, I thought deeply about all this interesting things.

And I thought much needed to be done. At that time, I designed the first mental poker protocol that is both realtime and have drop out tolerance (that is required for p2p playing). By drop out tolerance I mean that if a player drops out (accidentally or intentionally) the remaining players are able to keep playing without problem. (the quitting player cards get automatically shuffled in the deck).
This is important because, if not present, a colluding player can drop out (and block the protocol) to "save" another player from loosing money.

At that time I did my thesis and wrote a patent, and both documents are online (check US patent 13086208). The main problem that I found was the requirement for a time server or better, some timestamped message exchange service to prove a message has been delivered. Since Bitcoin blockchain is not fast enough to serve as a distributed time server, I thought the the technology was not mature enough.

Recently I developed a new p2p signature protocol called MAVE (for MAssive VErifications) and a new P2P currency based on it (yes, yet another blockchain!) which can process thousands of transactions per second with an average personal computer, using 50% of the average PC resources (CPU and RAM) and almost no hard disk space.
But still is not suitable as a time server, because it still need "confirmations".

I'll be publishing the preliminary paper in my cryptocurrencies blog late this week (http://bitslog.wordpress.com) and I'll post an update here so everybody can see it and check it.

I'd like to implement the whole platform, but I have no time. Maybe crowd-funding it?  I really don't want to mess up with regulation and get notices from government agencies, so I won't host any of this work myself. But I'd love to see it working...

I'd be glad to license the patent if someone works out the rest!

BTW, I think we should standardize the block chain, so that single mined block works for different block chains of different p2p networks, something like a Meta-blockchain. Did anyone posted about it?

Sergio.

I sounds kind of strange to talk about friends playing poker together as a bad thing, whatever happened to weekly poker nights amongst friends?

Maybe one could start with two extremes: either you all agree together that you want a table, or each player plays as a lone human against a table of artificial intelligences that do not collude with each other. Basically either you want a particular group of players to play together because they all agreed exactly who they are and that they want to play together or you insulate everyone from each other with layers of artificial players.

Then you can get into securing the code of the AI players to be sure what code each is actually running.

Or you could digress into the fun game of each writing their own A.I. to do their playing for them, teaching it to recognise and collude with A.I.s written or deployed by friends or sockpuppets and so on.

You could do like blackjack, where there is a fixed set of rules the house has to play by, so examination of the hands afterward can check all the A.I. players strictly followed their rules.

Also, isn't anonymity in the sense of not having live video you can zoom in on each player's face in real time kind of against the spirit of poker? It is all very well not to know the player's real name and profession and place of birth and such, but surely a key element of poker is high resolution real time observation of body language and facial expression and so on?

Being able to cheat by keeping the other players from watching your face eyes body language mannerisms etc is surely so much of a cheat in itself that worrying about people colluding is kind of secondary, since if you cannot watch them in real time you cannot see whether they are chatting with colluders during the game?

If you eliminate the human element of actual observation of the skin tone, perspiration, respiration, eye movement, twitches and all that, then you might as well play A.I. opponents. Otherwise you are deliberately setting up almost ideal conditions for cheating then complaining that people might cheat! (You need good audio too of course, in case they try to mumble under their breath into a concealed mic or something...)

-MarkM-

Blockchain everywhere! 

It can't be done in the current Bitcoin system as the scripting language is insufficiently powerful and cannot introspect (i.e the script can't read the block chain). A Bitcoin-like system with these features could be used to faciliate secure poker games and other contracts.
As a "decentralized betting framework", it would be vulnerable to legal attempts to disable it. As a "more powerful Bitcoin" it would have "substantial non-infringing purposes".

ByteCoin

Do you think that this can become something bigger as a "totally decentralized betting framework" ?

Well thanks, this is by no means a complete solution of course, I just wanted to present this idea of having own block chain exclusively for poker chips and poker hand data and examples of how that can be used to approach most of the big problems that arise in distributed poker. I hope that together with the real geniuses out there we will work out all the bugs and come up with a practically working design which will finally  get implemented!

Oh, by brilliant, I was referring to your post. By chip stealing, I just mean the standard issue of people needing to protect their private keys.
If you can manage this, I encourage you to actually do it. People will get excited about it.

Are you referring to the general idea or to my last post?
Because based on my design,
nonrandom shuffling is not a problem:
If we use approach 1 with commutative encryption, shuffling is always random, as each player shuffles cards themselves, and even if all but one players would shuffle in a strange way, the one shuffling truly would make their plans break. There could be a case where all the players shuffled cards in some strange way, but what do they win by that when they don't know how other players are shuffling?
If we use approach 2, that is indeed a problem. I suppose it could easily be fixed though by not using only the random value generated by the dealer as a seed, but instead a [random value from the seed + random value from each player], and a hash it, so that one would not be able to predict and thus manipulate the outcome of the shuffling.
One step further would be to first make the hash of the secret dealer value public first, THEN get the values from the users, THEN shuffle cards based on hash([those values + the secret value]). That way the seed will be made of both information from users and from the dealer, only dealer would have both pieces of information, and by the time it actually gets both pieces, it would be too late to change the first piece (hash has already been distributed). This is probably an overkill...

Friends at a table is the same thing as colluding, I have explored this in quite a detail in the previous post, basically the idea is to fix this with a beginner-fee and keeping tack of people always playing at the same table with each other. (And also semi-random table assignment.)

Can you elaborate more on what you mean by chip stealing?

Yeah, that sounds brilliant. The software appears very complicated. There are possible exploits [nonrandom shuffling, friends at a table, chip stealing] and implementation difficulties [random p2p matching, information pruning]. A secure, usable system would be an awesome feat. Can't wait to see this.