BitShares Forum
Main => General Discussion => Topic started by: Shentist on January 09, 2016, 12:44:21 am
-
i want to talk here about a FBA for decentralized betting!
the reason it should be run on the blockchain, is because in most countries it is illegal to run it, so how can we do it on bitshares?
so, my idea here is (and question is could we build it this on bitshares)
lets assume i want to bet on the outcome of the next presidential election. I am creating a new account/contract and i can define under which condition anyone can join this account/contract.
i am betting 10.000 BTS that the canidate of the democrates win and now the people can now join this betting account with betting also 10.000 BTS. they need to define what they are betting for,
in the end a lot of people can join this account/contract without anyones permission.
the problem is always in the end, how the contract knows who won the election. so the contract pays a portion of the amounts out to people who are reporting the result to the account. We could
setup accounts who are interested in earning some exctra cash.
step 1: contract/account is created from the initator with some specs he set
- possible answers to choose
step 2: interested people join this contract/account and bet too
step 3: the event occurs and interested people can now send to the account the winner
step 4: after x time the contract/account now sends the winning amounts out to the contract holders
step 5 : the account sends to the reporting people their share
step 6: account/contract is closed and will not be used anymore
this is just a rough idea, but the main point here is - how we can structure a decentralized betting event without the need that someone is legally responsible for it.
feedback :D
-
How would he contract be created? Would this be a new uia? Would the platform be completely decentralized?
-
How would he contract be created? Would this be a new uia? Would the platform be completely decentralized?
this is the point i want to talk.
could it be made possible with our account system or do we need new stuff to coded to create something like this.
-
How would he contract be created? Would this be a new uia? Would the platform be completely decentralized?
this is the point i want to talk.
could it be made possible with our account system or do we need new stuff to coded to create something like this.
I've been thinking about this topic for a few months now... I have a few solutions for the end of bet determination and payout, but I'm stuck on how the bet will actually be published. I would like to keep it fully decentralized, but it seems this part is where the centralization needs to be in order to keep costs contained.
-
How would he contract be created? Would this be a new uia? Would the platform be completely decentralized?
this is the point i want to talk.
could it be made possible with our account system or do we need new stuff to coded to create something like this.
Possibly one or a combination of those ideas combined with this:
I've been wondering if applications could be built on top of Bitshares that require off chain communication (like a troll box) by using WebRTC and making the original "handshake" on the blockchain. The way I understand it, WebRTC requires an initial handshake to be performed on by some centralized service. I posit it may be possible to do this handshake on the Bitshares blockchain. After the initial handshake, all communications between web browsers is decentralized. It would be a blockchain-less top level of Bitshares in which javascript applications or "plugins" could be built. It would allow people to easily build apps on top of Bitshares without stakeholder approval... truly opening the floodgates "ethereum-style"... allowing anyone to develop anything. You could also add the capability for the program/plugin/app write to the blockchain to add extra functionality, but that would be optional depending on each certain programs/plugins needs.
https://vimeo.com/108532169
-
Interestingly I had some thoughts about almost exactly the same idea yeaterday. Lets have a talk someday this week
Edit: everything needed for binary prediction markets is already there
-
Interestingly I had some thoughts about almost exactly the same idea yeaterday. Lets have a talk someday this week
Edit: everything needed for binary prediction markets is already there
That's funny you were thinking about this too! I think this avenue should be further explored- it could have wide reaching implications for Bitshares (in of course an extremely positive way.) There are so many things that could be built on top of Bitshares EASILY if this could be made to work.
Anyone that knows a little Javascript could make an addon or plugin for Bitshares.
-
We can already do this on blockchain all that is necessary is a little UI work in the wallet. Bet properties can be published on the blockchain via CUSTOM operation.
You would need to do LSMR prediction markets to justify a FBA.
-
We can already do this on blockchain all that is necessary is a little UI work in the wallet. Bet properties can be published on the blockchain via CUSTOM operation.
You would need to do LSMR prediction markets to justify a FBA.
nice to hear that!
i think it would be best if we are
"throwing as many stones as possible"
one stone will hit something huge :D
-
What I was thinking is this:
-I create a website where people can create there own bets.
-User A creates a bet with an expiration time (Packers vs Redskins for 2 bitUSD by 10pm 1/10/2016)
-This bet is displayed on the website for anyone to take (only 1 user can take the bet for 2 bitUSD)
-Once the bet expiration time expires, an email notification is sent to each player to settle there bet within 24hrs
-If both players agree on the winner, the money is released from escrow, and sent to the winning player. Each player can rate each other as well.
-If players disagree, the bet is randomly sent to a UIA holder to make the determination, The UIA holder receives 10% of the bet for work done. Basically 2 of 3 escrow.
-The UIA holder can also rate each player.
-The UIA would hold value because of it's ability to settle bets and make money. UIA holders would also be incentivised to make correct decisions in order to attract more players and settle more bets correctly.
I know there are some holes in this, and I believe it needs to be at least somewhat centralized. Please point out any improvements or show stoppers you see in this.
-
What I was thinking is this:
-I create a website where people can create there own bets.
-User A creates a bet with an expiration time (Packers vs Redskins for 2 bitUSD by 10pm 1/10/2016)
-This bet is displayed on the website for anyone to take (only 1 user can take the bet for 2 bitUSD)
-Once the bet expiration time expires, an email notification is sent to each player to settle there bet within 24hrs
-If both players agree on the winner, the money is released from escrow, and sent to the winning player. Each player can rate each other as well.
-If players disagree, the bet is randomly sent to a UIA holder to make the determination, The UIA holder receives 10% of the bet for work done. Basically 2 of 3 escrow.
-The UIA holder can also rate each player.
-The UIA would hold value because of it's ability to settle bets and make money. UIA holders would also be incentivised to make correct decisions in order to attract more players and settle more bets correctly.
I know there are some holes in this, and I believe it needs to be at least somewhat centralized. Please point out any improvements or show stoppers you see in this.
I really like this simple approach.
-
What I was thinking is this:
-I create a website where people can create there own bets.
-User A creates a bet with an expiration time (Packers vs Redskins for 2 bitUSD by 10pm 1/10/2016)
-This bet is displayed on the website for anyone to take (only 1 user can take the bet for 2 bitUSD)
-Once the bet expiration time expires, an email notification is sent to each player to settle there bet within 24hrs
-If both players agree on the winner, the money is released from escrow, and sent to the winning player. Each player can rate each other as well.
-If players disagree, the bet is randomly sent to a UIA holder to make the determination, The UIA holder receives 10% of the bet for work done. Basically 2 of 3 escrow.
-The UIA holder can also rate each player.
-The UIA would hold value because of it's ability to settle bets and make money. UIA holders would also be incentivised to make correct decisions in order to attract more players and settle more bets correctly.
I know there are some holes in this, and I believe it needs to be at least somewhat centralized. Please point out any improvements or show stoppers you see in this.
but the guy with the website could be charged responsible, this is the reason why i am looking into ways, that this can be run total without the need of a website etc.
- just a market on our client - a betting market
why do you limit it to 1 vs 1 ?
-
What I was thinking is this:
-I create a website where people can create there own bets.
-User A creates a bet with an expiration time (Packers vs Redskins for 2 bitUSD by 10pm 1/10/2016)
-This bet is displayed on the website for anyone to take (only 1 user can take the bet for 2 bitUSD)
-Once the bet expiration time expires, an email notification is sent to each player to settle there bet within 24hrs
-If both players agree on the winner, the money is released from escrow, and sent to the winning player. Each player can rate each other as well.
-If players disagree, the bet is randomly sent to a UIA holder to make the determination, The UIA holder receives 10% of the bet for work done. Basically 2 of 3 escrow.
-The UIA holder can also rate each player.
-The UIA would hold value because of it's ability to settle bets and make money. UIA holders would also be incentivised to make correct decisions in order to attract more players and settle more bets correctly.
I know there are some holes in this, and I believe it needs to be at least somewhat centralized. Please point out any improvements or show stoppers you see in this.
but the guy with the website could be charged responsible, this is the reason why i am looking into ways, that this can be run total without the need of a website etc.
- just a market on our client - a betting market
why do you limit it to 1 vs 1 ?
I agree with the part where the guy with the website could be charged with operating a gambling ring, but could this be hosted on a vps? Could multiple web pages in different countries with different gambling laws host their own versions of the webpage?
I limit it to 1 vs 1 because it allows for the 2 people to rate each other and create a reputation. People with higher reputations will be more likely to get the other side of their bets taken. It also would allow 2 of 3 escrow for winner determination, as each player would be a signer. If they don't agree, the bet description is sent to a random UIA holder to make the judgement (and collect 10% of the bet). The UIA holder would be the 3rd escrow agent.
-
The probability of obtaining a bet to settle could be based on your percentage holding of the asset.
This could be 'gamed' where a large stake holder colludes with others who dispute the bet.
Perhaps some formula that takes both stake holding and reputation into account.
-
Just do it.
-
I'm curious how this approach compares to Augur:
http://augur.link/augur.pdf
I'm working on a business around MUSE which provides incentives for people to get involved in Artist Coins as well as highlighting the top trending artists on the system. Think of it as the place to get the scoop before digging into the Peertracks system.
If there was a BitShares Prediction Market, that would be a service that my potential customers would see as a natural fit.
Example: Will the band SuperCool reach the top 20 trending chart by March 1st, 2016?
If I was able to take a small referral fee for that traffic, even better.
-
http://cryptofresh.com/a/REPGENSIXTEEN (http://cryptofresh.com/a/REPGENSIXTEEN)
GUI will not allow you to short. Please fix that. Can short in CLI
borrow_asset account-name 1000 REPGENSIXTEEN 1000 true
collateral has to be 1 for 1. I am willing to update asset owner to account that is controlled by multi-sig trusted parties.
-
The probability of obtaining a bet to settle could be based on your percentage holding of the asset.
This could be 'gamed' where a large stake holder colludes with others who dispute the bet.
Perhaps some formula that takes both stake holding and reputation into account.
UIA holders colluding with other players is definitely possible, but this wouldn't be in the best interest of the asset holder. It would damage the platform and make the associated UIA worth less, therefore hurting the holder.
I do like the idea of taking reputation into account, but it may add a layer of unnecessary complexity. If the better was ever taken advantage of in your situation, he could always give negative feedback to both the player and UIA holder. I would like to see an algorithm that randomly assigns the disputed bet judgement to a random UIA. This would make the system very difficult to game.
-
Interestingly I had some thoughts about almost exactly the same idea yeaterday. Lets have a talk someday this week
Edit: everything needed for binary prediction markets is already there
In the CLI only?
http://cryptofresh.com/a/REPGENSIXTEEN (http://cryptofresh.com/a/REPGENSIXTEEN)
GUI will not allow you to short. Please fix that. Can short in CLI
borrow_asset account-name 1000 REPGENSIXTEEN 1000 true
collateral has to be 1 for 1. I am willing to update asset owner to account that is controlled by multi-sig trusted parties.
Hmm. Can't short cause there are no feeds. BM said there shouldn't be any feeds though. Is this just a GUI error? I'll try to short in the CLI when I get a chance. Unless we can get the GUI working I don't think this will catch on at all though.
-
The probability of obtaining a bet to settle could be based on your percentage holding of the asset.
This could be 'gamed' where a large stake holder colludes with others who dispute the bet.
Perhaps some formula that takes both stake holding and reputation into account.
UIA holders colluding with other players is definitely possible, but this wouldn't be in the best interest of the asset holder. It would damage the platform and make the associated UIA worth less, therefore hurting the holder.
I do like the idea of taking reputation into account, but it may add a layer of unnecessary complexity. If the better was ever taken advantage of in your situation, he could always give negative feedback to both the player and UIA holder. I would like to see an algorithm that randomly assigns the disputed bet judgement to a random UIA. This would make the system very difficult to game.
The user would need to somehow indicate that they are available to settle (unlocked wallet?). And have a time period within which to act otherwise the bet moves to someone else to settle.
I find this interesting because it would be really nice to have a game that builds on bitshares. Any game. Something to make Bitshares fun. I think Bitshares is too way ahead of its time to be useful to many people.
-
If we're gonna go the FBA route why don't we just have the top five hodlers of the FBA responsible for providing the correct result. If they cheat then no one will use the system and their revenue stream will end.
-
If we're gonna go the FBA route why don't we just have the top five hodlers of the FBA responsible for providing the correct result. If they cheat then no one will use the system and their revenue stream will end.
That's along the lines I was thinking... but I would prefer every FBA or UIA holder to have a chance at settling the disputed outcome. This would hopefully entice smaller people to buy into the system and would provide further decentralization.
Lets say I call my UIA IBET. I would like all IBET holders to have a chance to settle disputed outcomes. Can I give each IBET a unique identifier that could randomly be selected by an algorithm?
-
If we're gonna go the FBA route why don't we just have the top five hodlers of the FBA responsible for providing the correct result. If they cheat then no one will use the system and their revenue stream will end.
No need for an FBA since the feature is already available and can be used any time .. so to speak BTS is the FBA .. except that you could either have the committe, the witnesses or an arbitrary set of accounts 'resolve' the bet .. all of this is already available ..
-
I really like gambling so if you guys create different prediction market assets in order not to lose track it would be nice if @cryptofresh they could seperate these assets in a category of their own "market prediction assets" and these assets have a description of the bet
-
I think first we should define who has the final saying on the outcome of an event. Witnesses, committee? Or does that add a work overload they they shouldn't have? Maybe if we create a new type of "status" or account type, like Prediction Witnesses or Oracles.
All of them should report the same outcome, the correct one.
People will argue against that saying with Augur lots of people decide it (if I got it right) and it's more decentralized.
We can counter that argument that shareholders can vote in and out Oracles who provide the correct outcomes.
I think the problem here is the incentive. In Augur or any other system, Oracles should have an incentive to provide the correct outcome. With BitShares we don't have that atm I think.
Possible solutions:
- They earn bts like witnesses for their efforts
- They earn a % of all the winnings of a certain event.
Say there's a prediction market about US Presidentials. Person P wins and there's an X amount to be distributed to all who participated on the market. a % of that could be divided through all the Oracles maybe?
-
I noticed this message on Telegram from @Xypher . I wanted to pass it along:
"Crypted Xypher
If you are working on prediction markets. PLEASE contact me as am working on a betting project and I can really use help there. I'd be glad to plugnin your product to what we are building"
-
I noticed this message on Telegram from @Xypher . I wanted to pass it along:
"Crypted Xypher
If you are working on prediction markets. PLEASE contact me as am working on a betting project and I can really use help there. I'd be glad to plugnin your product to what we are building"
Indeed, I'd like to confirm this.
I've been "hunting' for documentation on prediction markets for a while. Since one of our projects does interact with "probabilities" of a certain event happening a lot, I believe PM's can be built into the product in a very efficient fashion. If there are individuals working on this, I'd be glad to work with them to see possibilities of a collaboration.
Regards
-
I have created a second prediction market. This tracks the winner of SuperBowl 50. http://cryptofresh.com/a/SPORTS.SBLAFC (http://cryptofresh.com/a/SPORTS.SBLAFC) If the AFC wins SuperBowl 50 then this asset will be worth 1BTS if the NFC wins SuperBowl 50 then this asset will be worth 0BTS. Short and buy accordingly.
Have Fun.
-
I have created a second prediction market. This tracks the winner of SuperBowl 50. http://cryptofresh.com/a/SPORTS.SBLAFC (http://cryptofresh.com/a/SPORTS.SBLAFC) If the AFC wins SuperBowl 50 then this asset will be worth 1BTS if the NFC wins SuperBowl 50 then this asset will be worth 0BTS. Short and buy accordingly.
Have Fun.
Shouldn't you close the market some days before the outcome in order to avoid manipulation? Does augur do this? If they do, it would be an interesting approach if we didn't, we could get preference from some because of that
Really excited about this. Hope everything goes well with these first prediction markets so we also get more exposure and hopefully someone to develop a service and interface for PMs
-
I have created a second prediction market. This tracks the winner of SuperBowl 50. http://cryptofresh.com/a/SPORTS.SBLAFC (http://cryptofresh.com/a/SPORTS.SBLAFC) If the AFC wins SuperBowl 50 then this asset will be worth 1BTS if the NFC wins SuperBowl 50 then this asset will be worth 0BTS. Short and buy accordingly.
Have Fun.
Shouldn't you close the market some days before the outcome in order to avoid manipulation? Does augur do this? If they do, it would be an interesting approach if we didn't, we could get preference from some because of that
Really excited about this. Hope everything goes well with these first prediction markets so we also get more exposure and hopefully someone to develop a service and interface for PMs
I don't think its really needed. The odds will be adjusting the price up until the start of the game. Anyone that has a losing bet still on the books after the the result is obvious, but before the market can be closed deserves what they get.
-
I think I picked the wrong thing to predict in my test prediction market. Btc is down $50 since I posted it Wednesday.