Hello, I am a long-time software engineer and I have built a Texas Hold'em Poker platform based on the ICP Blockchain.

The system is decentralized, non-custodian, provably fair, verifiable, etc.

If someone is interested in trying it out and providing feedback, please send me a DM.

Thanks,
--h

Fully decentralized?

Yup, running on several nodes and with every action committed to their blockchain. https://dfinity.org/

So there are no staff once it is running, it just runs?

Yeah - though it's still in development. It's like Defi but with poker.

How do you deal with things like collusion, botting, and other game security related issues if there are no staff and the wallets are non-custodial?

I don't deal with these. Tables are not listed and you need to know their name to join them. It encourages clubs/friends games.
I am concerned about:
- Website disappearing with user funds
- Close source code
- Unfair RNG
- Privacy

It is disappointing to read that the creator of a poker software is not also concerned about game security

I am concerned about these problems but they are very hard to solve. Currently, policing games with a staff comes with disadvantages too.

Such as?

I need to trust that the staff with my funds. I have been on poker sites that suddenly disappeared.
Even if the site doesn't, the staff can make mistakes.

Given you are a new poster with no history and openly admit that whatever site you create will have zero security of any kind - from 1 to 100 where 1 is nobody should take you seriously to 100 where everyone thinks you are the next great thing in this industry - where did you think your posts would score when you made them?

I’m o give op benefit of the doubt although this does seem sketch we deffo need a decentralised poker client run by a decent community in harmony with strong devs

Pls post an option to use site on a testnet

Pls explain how you built it out

I am looking for people that are interested in this style of service to try it, run a few tournaments, and provide feedback. It's not production ready.
Clearly, if people want a staff to oversee tables, this is not it.
I don't have any history in this site of poker but I have history in crypto dev sites (not that it matters).

Fair enough. If your target market are people who want some type of verifiable deal in some way though are not concerned at all about potential cheating from other players or the site's financial viability then this is the perfect experiment for them to try and give their feedback to help make it better to suit their specific needs.

Exactly, however:

The poker d-app shares infrastructure with other d-apps. The smart contract is only billed for the cycles it uses. Therefore tables that are not actively played have no cost. The financial burden should scale up with the usage.

I posted a video on youtube about this project: https://youtu.be/aXvBqNEs3fw

You may not want to put into your strategic plan that you are looking for players who are not concerned whether collusion takes place (as no method of detecting or punishing collusion will exist). Just a suggestion, but the good thing for you is that often times players that obsess about a "fair deal" by their own undefined metrics after they imagine other sites are rigged will not be concerned with actual industry cheating issues like collusion. They will of course believe your site is rigged eventually if they play there, and the irony is it probably will be as they play against teams of players sharing hand information that is openly allowed.

All the best.

I have to agree with OP there isn't really much you can do about collusion on a decentralized site such as this. The reason is really simple - it's decentralized - there is nobody to check whether anyone is colluding, and if you try and write code to detect collusion, you could easily get false positives, so it would have to happen multiple times within a certain number of hands to be confident it's collusion and not coincidence. The colluders could also decide to only collude every once and a while to help avoid this type of detection. A human reviewing probably wouldn't even be able to detect that in a lot of instances. Besides that, what pattern(s) would you even look for that would determine 2 players are colluding? The possibilities on how 2 players can collude are pretty much endless so the only reasonable way I can see to approach the problem would be to figure out what are the most common ways players collude and detect for that. At the same time you can't tell anyone what ways you're looking for otherwise they can avoid those methods of collusion.

What happens to the funds if someone is detected colluding? Do they only lose the funds for the specific hand being played when they get caught or does their crypto wallet get frozen as well? Presumably if it's the latter the colluding players are already cashing out on a regular basis so it's too late to freeze the funds anyway. If a very small rake is charged (as sad as that would be) maybe it could be used to reimburse players when these things happen, but I don't know if that is even possible being that it is decentralized.

Even if your anti-collusion tech works pretty well and you catch players colluding what is to stop them from making new usernames and using a new IP address (usually you just have to call your ISP to have that changed).

Nobody wants to be colluded against but IMO it will always be going on whether it's decentralized or not.

The best way I can think of to avoid collusion on a decentralized type of site is to do away with the table name & hand number, no chat, anonymize the player names like Bovada does, and shuffle the players between tables fast fold style every hand. If there is a history of the hand in the blockchain then everything (table name, hand #, players) could be recorded there for review. Even then collusion could still happen on a limited scale by 2 colluders being on the phone and confirming they face the same action / board runout and sharing cards.

Personally I go opposite way and I've put a lot of thought into this since 2008 -- when the poker rooms started getting greedy.

You basically just have to gamify collusion, botting, staking, stables, railtarding and build it into the system as a game model imo. And let game theory do what she does best without any humans crapping on the table.

If you can't beat them join them, if you can't beat them fairly, then outnumber them. Basically the same principles as building a guild of players for whatever game. numbers always win.

Regards,
Tops

One way to avoid anonymous tables I thought of would be to allow other players to collectively mark a player as a cheater/colluder, but there would have to be some kind of player association that reviews the reports to prevent abuse (such as players deciding they just don't like someone and making new accounts to report them with). People who constantly make bad reports would have their reports ignored. Still doesn't solve the problem of people making a new account if they get banned. Interested to hear how that could be prevented.

First, thanks for the feedback. The approach I'd like to take for this d-app regarding collusion is self-policing (in the beginning).
Tables/Tournaments are private and require knowledge of their name/access code.
The app "rents" tables for an advance on the rakes collected: A portion is paid back to the table creator who is expected to prevent collusion on his tables. For example, he could be a private game host or a club owner. (The d-app privacy is in the d-app not having any identifiable data about its users).
In the meantime, I am looking into automatic collusion detection based on patterns of actions though I'm not sure how reliable these algorithms are at this point.

I like the idea of shuffling players around frequently. This could be added as an option at table creation: "Move players after N hands". It will be fairly easy to add because the app has a player rebalancing logic for multi table tournaments.

Thanks,

Hey! Nice project.

In the vid I saw that you are planning to make it open source.
How soon?

I want the code cos I'm curious if it is possible to speed up it by porting to TON.
Although I can't promise I'll have the time to put in the work.
And I'm a noob at TON/C++/Rust/WebApps/DApps so will need some time to catch up.

CoinPoker uses smart contracts as well and it is platform, which has been around a few years already.

Personally i do not play there but this review https://thepokeragent.com/is-coinpoker-legit/ , got me interested in this concept. Especially the part abouT tieing RNG to blockchain.

Arriving late to this conversation, but the advantages of smart contracts versus sending your money to a cardroom are not yet appreciated. People still don't understand that a smart contract means you don't actually send your money to a cardroom that can then randomly decide to keep it. If your're still working on this project, what people want is protection from cheaters, loggers, and colluders. Anonymous identities and random table placement will go a long way on that front. And I for one can't wait to find a cardroom with no KYC, no avatar identity, no sending bitcoin somewhere and wondering if I will ever see it again, and no opportunity for collusion because it's functionally impossible. It's coming. Someday...