BitShares Forum
Main => General Discussion => Topic started by: luckybit on March 15, 2015, 05:33:46 am
-
http://www.augur.net/blog/augur-at-sxsw-s-startupdebut
http://www.augur.net/
Robin Hanson is involved and he's probably the thought leader in prediction markets.
Bitshares was supposed to be a prediction market. What happened?
-
http://www.augur.net/blog/augur-at-sxsw-s-startupdebut
http://www.augur.net/
Robin Hanson is involved and he's probably the thought leader in prediction markets.
Bitshares was supposed to be a prediction market. What happened?
you can't jump to the moon without a solid infrastructure .
Function is easy , stability and performance is hard .
You can add a lot of function in it . But then again , even the current load of function is buggy as it is , you don't want to rush for more and fix it later .
-
http://www.augur.link/augur.pdf :-\
-
http://www.augur.net/blog/augur-at-sxsw-s-startupdebut
http://www.augur.net/
Robin Hanson is involved and he's probably the thought leader in prediction markets.
Bitshares was supposed to be a prediction market. What happened?
Augur seems to have a creditable team to help it succeed.
Is BitShares into prediction market too? Are you referring to the Vote merger?
-
Sounds like an interesting one to keep an eye on, perhaps even have a discussion with on areas of cross-fertilisation. From my glancing through the whitepaper, sounds like they want to create a cash-like coin, dubbed USD coin, for betting. The whitepaper suggests this would be done by printing more of the token when demand is high (and price above $1), and ceasing to print when demand is low (price below $1). Of course this is only asymmetric flexibility in the supply, and does not peg to a USD when demand is falling. Maybe they could be approached on using bitUSD instead.
-
They are working with Ethereum...
http://www.augur.net/blog/why-ethereum
-
I dont think pm will work for mpa that bitshares solved... I think the prediction market capability should be solved and then ported into bitshares.. We prob would need turing complete scripts with smartncontracts also... Then we simply copy logic over.. Let them do the research work and testing first.
Under why decentralize whitepaper says that easiest way is to have oracles that update events... Then it becomes a trivial problem howver in bitcoin why would you trust some iracles over others? So they have to develop ability to update events on every node and inventisize updates... While with bitshares we simply updste events like price feeds... So we solve the first problem of trusted oracles that update events whom everyone trusts... It would really test integrity of the delegate system too
Because no delegates augur developed the reputation system which entitles ppl to post event results.. And obliged to checkin every 8 weeks.. To sync up blockchain to reality... This is also a pressure point that can break and if thr theory of delegate holds up then solved by bitshares... Essentially it boils down to craete a new asset type and allow ppl to buy those assets based on how pms work.. The ppl who guess right win money.. Should be as simple as that.. The only question is who would create the predictions and what kind of fee to create these assets?
-
Our competitive advantage is the delegates.. No other projects solved the problems like us so we should leverage this advantage as much as possible.. This is just another avenue to leverage and benefit from our delegates that will help us get network effect. Web services and other things should also be considered to be able to port centralized services which are cool and alot of ppl use.. Essentially decentralizing it.
-
While with bitshares we simply update events like price feeds...
Yes this.
Bitshares can do prediction markets quite well.
We elect judge delegates who we trust to run the prediction market, and they provide results in the way that delegates now provide price feeds.
The judges would be elected by shareholders, and if you do not trust them for some reason then you don't use our prediction market.
-
While with bitshares we simply update events like price feeds...
Yes this.
Bitshares can do prediction markets quite well.
We elect judge delegates who we trust to run the prediction market, and they provide results in the way that delegates now provide price feeds.
The judges would be elected by shareholders, and if you do not trust them for some reason then you don't use our prediction market.
I think we would have delegates checkin and update events... A delegate is a delegate..msoon as you get different delegates doing different things you break the decentralization concept and have issues with security and other attacks..mwhatever we do it shouldnbe generalized and easy to do for all delegates...
If people screw around with bad data they get voted out.. In both systems someone high on delegate list or with good reputation points can end up gaming the system one time and losing repuation or delegate... But if alot of money on line maybe it will be done so we need a way to try to fix thAt.
Perhaps one way is to limit markcap of these markets and grow them as a percentage of overall marketcap to reduce these oneoff manipulations.. Delegate wont give up his pay if he only wins say 1 month worth of salary by manipulation... These kinds of things we can do but others cant
-
What happen edwas that Bytemaster ignored the Truthcoin proposal. fyi augur's origins was truthcoin. I want you all to remember this could have been part of bitshares... And this wont be the only prediction market in development that will compete hard against bitshares.
And delegates for prediction markets are a core disadvantage. Its a disaster waiting to happen. I would not trust millions of dollars of trading to transparent delegates that can collude together.
However... if Bitshares can pivot quickly to scripting, and thats a big if they can do fast enough, bitshares can replicate and win out.
-
What happen was that Bytemaster ignored the Truthcoin proposal. fyi augur's origins was truthcoin. I want you all to remember this could have been part of bitshares... This also wont be the only prediction market in development that will compete hard against bitshares.
And delegates for prediction markets are a core disadvantage. Its a disaster waiting to happen. I would not trust millions of dollars of trading to transparent delegates that can collude together.
But then wouldnt the same arugment apply when your talking about signing blocks worth alot of money? Isnt it the card we are playing here? Otherwise whats the advantage here if any? I dont think we were at a point to implement truth coin... And its better research is done in another chain before we integrate... I dont think siegnorage coins should be done via prediction markets... With their model they have to...
-
What happen was that Bytemaster ignored the Truthcoin proposal. fyi augur's origins was truthcoin. I want you all to remember this could have been part of bitshares... This also wont be the only prediction market in development that will compete hard against bitshares.
And delegates for prediction markets are a core disadvantage. Its a disaster waiting to happen. I would not trust millions of dollars of trading to transparent delegates that can collude together.
But then wouldnt the same arugment apply when your talking about signing blocks worth alot of money? Isnt it the card we are playing here? Otherwise whats the advantage here if any? I dont think we were at a point to implement truth coin... And its better research is done in another chain before we integrate... I dont think siegnorage coins should be done via prediction markets... With their model they have to...
No because delegates only have the power to stunt the signed message from transmitting. It just makes the transaction not go through. Delegates with the power to decide resolutions, gives them the power to allocate winnings where they please, if they choose to collude.
Look, truthcoins implementation is much simplier than bitshares. So to get to a siegnorage coin won't be difficult.
-
What happen was that Bytemaster ignored the Truthcoin proposal. fyi augur's origins was truthcoin. I want you all to remember this could have been part of bitshares... This also wont be the only prediction market in development that will compete hard against bitshares.
And delegates for prediction markets are a core disadvantage. Its a disaster waiting to happen. I would not trust millions of dollars of trading to transparent delegates that can collude together.
But then wouldnt the same arugment apply when your talking about signing blocks worth alot of money? Isnt it the card we are playing here? Otherwise whats the advantage here if any? I dont think we were at a point to implement truth coin... And its better research is done in another chain before we integrate... I dont think siegnorage coins should be done via prediction markets... With their model they have to...
No because delegates only have the power to stunt the signed message from transmitting. It just makes the transaction not go through. Delegates with the power to decide resolutions, gives them the power to allocate winnings where they please, if they choose to collude.
Look, truthcoins implementation is much simplier than bitshares. So to get to a siegnorage coin won't be difficult.
No it isnt.. the reputation system is complex.
What if we vest the winnings over time so that would eliminate colluding. This way any manipulation wont be economical for delegates.
-
Truthcoin is so complicated-- yet their are several teams churning out versions by June. By June its possible 3 different teams will have the capability of replicating the entire public stock market. And after a year and a half, what do we have with Bitshares? We haven't yet received what we were sold on, the polymorphic digital asset.
Your effectively doing the same thing truthcoin is trying to do. Make the delegates have skin in the game. With Truthcoin the share ownership is their vested winnings. And with that your protecting identities as well making the system censorship resistent.
-
Truthcoin is so complicated-- yet their are several teams churning out versions by June. By June its possible 3 different teams will have the capability of replicating the entire public stock market. And after a year and a half, what do we have with Bitshares? We haven't yet received what we were sold on, the polymorphic digital asset.
Your effectively doing the same thing truthcoin is trying to do. Make the delegates have skin in the game. With Truthcoin the share ownership is their vested winnings. And with that your protecting identities as well making the system censorship resistent.
Our implemetation is simpler and maybe more effective with delegates.. All other implemtations need reputation system and incentives that may break it. We have a different approach that may or may not be better that is ourmarket advatnage
-
@Jsidhu
Delegates should not be the ones resolving markets. It would surprise me if Bytemaster was not on agreement with this. They should have a limited role, (think limited governance like how the founding fathers started the country) which is signing transactions. Giving them anything more would invite unchecked power. Information markets are for the lack of a better word going to be extremely powerful paradigm. Your turning the bitshares system into a democracy, which is dangerous when confronted with mob mentality.
Bytemaster has proposed a Judge slate which can be proportedly be voted in by people. But these judges are not to be confused with the role of a delegate. They are only to resolve markets, not sign transactions. I think judges are fine, and a good enough compromise in the short term. Wouldn't you agree with this?
-
@Jsidhu
Delegates should not be the ones resolving markets. It would surprise me if Bytemaster was not on agreement with this. They should have a limited role, (think limited governance like how the founding fathers started the country) which is signing transactions. Giving them anything more would invite unchecked power. Information markets are for the lack of a better word going to be extremely powerful paradigm. Your turning the bitshares system into a democracy, which is dangerous when confronted with mob mentality.
Bytemaster has proposed a Judge slate which can be proportedly be voted in by people. But these judges are not to be confused with the role of a delegate. They are only to resolve markets, not sign transactions. I think judges are fine, and a good enough compromise in the short term. Wouldn't you agree with this?
How would a judge slate work? Would it solve the issue of resolving events? It would need to be a decentralized solution as much as the delegates.. if so then ya i agree
-
How would a judge slate work? Would it solve the issue of resolving events? It would need to be a decentralized solution as much as the delegates.. if so then ya i agree
I believe a judge slate would be a set of up to 111 account IDs (not necessarily only delegate accounts). Then if a majority of the judges in that set agree to a particular outcome for a prediction market, the blockchain would treat it as the outcome for that prediction market to settle the rewards.
The decentralized would only be limited to 111 individuals in that case (not the potentially much larger group holding votecoins as in Truthcoin and Augur) and there wouldn't be any automatic stake redistribution based on how they voted (PCA / SVD), instead users would have to manually decide on which accounts were the most trustworthy based on their past record and use that to inform future decisions on which set of judges to pick in future prediction markets. But it is a much simpler system (compared to Truthcoin and Augur) to get quickly working. Unfortunately, even that simpler system has to a take a backseat until BitShares gets its core product in proper shape first.
-
How would a judge slate work? Would it solve the issue of resolving events? It would need to be a decentralized solution as much as the delegates.. if so then ya i agree
I believe a judge slate would be a set of up to 111 account IDs (not necessarily only delegate accounts). Then if a majority of the judges in that set agree to a particular outcome for a prediction market, the blockchain would treat it as the outcome for that prediction market to settle the rewards.
The decentralized would only be limited to 111 individuals in that case (not the potentially much larger group holding votecoins as in Truthcoin and Augur) and there wouldn't be any automatic stake redistribution based on how they voted (PCA / SVD), instead users would have to manually decide on which accounts were the most trustworthy based on their past record and use that to inform future decisions on which set of judges to pick in future prediction markets. But it is a much simpler system (compared to Truthcoin and Augur) to get quickly working. Unfortunately, even that simpler system has to a take a backseat until BitShares gets its core product in proper shape first.
ok so whats enforcing judges to participate? Truthcoin would pay ppl to particpiate but whats the right amount? Is it like a delegate that would tell community he woukd charge so and so to become a judge where ppl can choose that person to decide outcomes on events? Seems like its complicating it more than it should?
-
ok so whats enforcing judges to participate? Truthcoin would pay ppl to particpiate but whats the right amount? Is it like a delegate that would tell community he woukd charge so and so to become a judge where ppl can choose that person to decide outcomes on events? Seems like its complicating it more than it should?
Yeah, once you start adding in details it gets more complicated. I think the creator of the prediction market should specify the parameters like how much the payment for the judges that provide the consensus outcome should be, what percentage fee to charge per market order, and what percentage of that goes towards increasing the liquidity of the LS-LMSR market maker while the rest goes into a pool from which to pay the judges and the initial liquidity providers. Also, the prediction market should likely not even open until a super majority of the selected judges agree to judge the prediction market (then they would be judged negatively by the public if they committed to provide outcomes but failed to do so).
-
ok so whats enforcing judges to participate? Truthcoin would pay ppl to particpiate but whats the right amount? Is it like a delegate that would tell community he woukd charge so and so to become a judge where ppl can choose that person to decide outcomes on events? Seems like its complicating it more than it should?
Yeah, once you start adding in details it gets more complicated. I think the creator of the prediction market should specify the parameters like how much the payment for the judges that provide the consensus outcome should be, what percentage fee to charge per market order, and what percentage of that goes towards increasing the liquidity of the LS-LMSR market maker while the rest goes into a pool from which to pay the judges and the initial liquidity providers. Also, the prediction market should likely not even open until a super majority of the selected judges agree to judge the prediction market (then they would be judged negatively by the public if they committed to provide outcomes but failed to do so).
So why not make it part of the delegates job and really simplify things? Make it vested over time to avoid collusion? Once the mail system is in place itreally would be simple to notify delegates to update events.. they should be up to date with current affairs anyway.
-
ok so whats enforcing judges to participate? Truthcoin would pay ppl to particpiate but whats the right amount? Is it like a delegate that would tell community he woukd charge so and so to become a judge where ppl can choose that person to decide outcomes on events? Seems like its complicating it more than it should?
Yeah, once you start adding in details it gets more complicated. I think the creator of the prediction market should specify the parameters like how much the payment for the judges that provide the consensus outcome should be, what percentage fee to charge per market order, and what percentage of that goes towards increasing the liquidity of the LS-LMSR market maker while the rest goes into a pool from which to pay the judges and the initial liquidity providers. Also, the prediction market should likely not even open until a super majority of the selected judges agree to judge the prediction market (then they would be judged negatively by the public if they committed to provide outcomes but failed to do so).
So why not make it part of the delegates job and really simplify things? Make it vested over time to avoid collusion? Once the mail system is in place itreally would be simple to notify delegates to update events.. they should be up to date with current affairs anyway.
Won't there come a time when we can't have the delegates doing every trusted human function? Currently we have them signing blocks, developing and marketing, providing price feeds and setting up voting slates. We could potentially set them multiple other tasks such as judging PM outcomes, scoring reputations, arbitrating, determining claims of different sorts etc etc. But at some point we need specialists in each of these areas so that they are chosen on that particular skill (instead of a conglomerate of skills) and incentivised to perform accordingly. It seems that may be a tough thing to achieve technically, but isn't that where things must ultimately go?
-
How would a judge slate work? Would it solve the issue of resolving events? It would need to be a decentralized solution as much as the delegates.. if so then ya i agree
I believe a judge slate would be a set of up to 111 account IDs (not necessarily only delegate accounts). Then if a majority of the judges in that set agree to a particular outcome for a prediction market, the blockchain would treat it as the outcome for that prediction market to settle the rewards.
The decentralized would only be limited to 111 individuals in that case (not the potentially much larger group holding votecoins as in Truthcoin and Augur) and there wouldn't be any automatic stake redistribution based on how they voted (PCA / SVD), instead users would have to manually decide on which accounts were the most trustworthy based on their past record and use that to inform future decisions on which set of judges to pick in future prediction markets. But it is a much simpler system (compared to Truthcoin and Augur) to get quickly working. Unfortunately, even that simpler system has to a take a backseat until BitShares gets its core product in proper shape first.
They are called oracles.
-
I'm sure it will be judges will be selected by the pm creator. You could also have judges elected in by the bitshares community but this voting will be completely separate from delegate voting. just an option that would serve a need and doesn't have to be the best.
-
It is a shame that I have to bite my tongue on forward looking features.
Overall I am in favor of providing tools that allow people to create their own PM with their own judges and profit from it. The delegates should stick to the role of block production and price feeds. There are entirely too many different prediction markets out there for the delegates to be the judge.
-
It is a shame that I have to bite my tongue on forward looking features.
Overall I am in favor of providing tools that allow people to create their own PM with their own judges and profit from it. The delegates should stick to the role of block production and price feeds. There are entirely too many different prediction markets out there for the delegates to be the judge.
Isnt the whole point to create a secure decentralized environment for prediction markets functioning honestly? Creating judges seems like more complexity without actually levereging our edge.. might aswell keep pms outside the chain and somehow just reference them.. let some other system worry about the issues since one crack and it would ruin confidence in the whole goal of
the project.
-
It is a shame that I have to bite my tongue on forward looking features.
Overall I am in favor of providing tools that allow people to create their own PM with their own judges and profit from it. The delegates should stick to the role of block production and price feeds. There are entirely too many different prediction markets out there for the delegates to be the judge.
Isnt the whole point to create a secure decentralized environment for prediction markets functioning honestly? Creating judges seems like more complexity without actually levereging our edge.. might aswell keep pms outside the chain and somehow just reference them.. let some other system worry about the issues since one crack and it would ruin confidence in the whole goal of
the project.
If you remember the prediction markets were the whole point of Bitshares. I remember early on that the main feature after a decentralized exchange that I was excited about was the prediction markets.
I don't know what happened to that or what the roadmap is anymore. I think the price is suffering now compared to last summer because the roadmap is less clear now.
-
It is a shame that I have to bite my tongue on forward looking features.
Overall I am in favor of providing tools that allow people to create their own PM with their own judges and profit from it. The delegates should stick to the role of block production and price feeds. There are entirely too many different prediction markets out there for the delegates to be the judge.
Isnt the whole point to create a secure decentralized environment for prediction markets functioning honestly? Creating judges seems like more complexity without actually levereging our edge.. might aswell keep pms outside the chain and somehow just reference them.. let some other system worry about the issues since one crack and it would ruin confidence in the whole goal of
the project.
If you remember the prediction markets were the whole point of Bitshares. I remember early on that the main feature after a decentralized exchange that I was excited about was the prediction markets.
I don't know what happened to that or what the roadmap is anymore. I think the price is suffering now compared to last summer because the roadmap is less clear now.
Seriously, roadmap is so badly needed it's crazy that it hasn't happened yet. We need less tongue-biting and more concrete details. Just avoid saying anything obviously controversial (negative attacks against competitors, proposals without discussing the ramifications with fellow devs first, more inflation [hah]) and the BTS community won't panic sell every time you open your mouth. Stick to facts, tell us what is legit going on. Honestly, it really shouldn't be that difficult.
-
It is a shame that I have to bite my tongue on forward looking features.
Overall I am in favor of providing tools that allow people to create their own PM with their own judges and profit from it. The delegates should stick to the role of block production and price feeds. There are entirely too many different prediction markets out there for the delegates to be the judge.
Isnt the whole point to create a secure decentralized environment for prediction markets functioning honestly? Creating judges seems like more complexity without actually levereging our edge.. might aswell keep pms outside the chain and somehow just reference them.. let some other system worry about the issues since one crack and it would ruin confidence in the whole goal of
the project.
@jidshu
how secure would it be if delegates were to rule on already publicly traded stocks? This opens the door for the SEC to attack delegates with federal lawsuits. They have no leniency policy against this. Why would you want delegates to run a business always in fear of suspicion?
Even in a more mild case of unregistered IPO's, the SEC has administered material penalties.
http://www.crowdfundinsider.com/2014/12/58808-sec-fines-bitcoin-virtual-stock-exchange-conducting-unregistered-offerings/
-
It is a shame that I have to bite my tongue on forward looking features.
Overall I am in favor of providing tools that allow people to create their own PM with their own judges and profit from it. The delegates should stick to the role of block production and price feeds. There are entirely too many different prediction markets out there for the delegates to be the judge.
Isnt the whole point to create a secure decentralized environment for prediction markets functioning honestly? Creating judges seems like more complexity without actually levereging our edge.. might aswell keep pms outside the chain and somehow just reference them.. let some other system worry about the issues since one crack and it would ruin confidence in the whole goal of
the project.
@jidshu
how secure would it be if delegates were to rule on already publicly traded stocks? This opens the door for the SEC to attack delegates with federal lawsuits. They have no leniency policy against this. Why would you want delegates to run a business always in fear of suspicion?
Even in a more mild case of unregistered IPO's, the SEC has administered material penalties.
http://www.crowdfundinsider.com/2014/12/58808-sec-fines-bitcoin-virtual-stock-exchange-conducting-unregistered-offerings/
What does that even mean? If they publish price feeds on stocks and we have MPAs for them.. then we are not trading the actual stock just a token like bitUSD. So I don't see how that has anything to do with prediction markets. My issue is that this judge slate thing won't have security like DPOS and delegates signing blocks. We are creating a whole new system to deal with something that delegates can do. Imagine when price is 200x higher... delegates will be expected to do such things because otherwise people will feel they are not doing enough. Most delegates will be at 1% pay but some may want more and will offer to do more..
The real issue is surfacing what needs to be done for delegates as a checklist through a mechanism such as the mail system.. if that is the bottleneck we shoudl focus on building it and getting it based on specs and then workign with it (the mail system) to deliever tasks to delegates and others who have something "due". Delegates should be "checking in" every day or so anyway or they should be voted out. I have already said we can find ways to work around the "power" issue with delegates by making vested payments... these issues are easier to solve then hacking together something to work specifically for prediction markets.
We just need an integrated application that takes into account all vectors and starts to polish it to become usable.. the core code doesn't have to change just the apps that use the API.
Going forward i was hoping that we would be able to add web services to delegate nodes to do even more things.. ofcourse not through the HTTP server on the wallet but perhaps a hosted service either on the same node or somehow connected, maybe making a mesh network b etween all 101 delegates to allow a "decentralized" web app experience to do social logins, single sign on apps like getclef.com and other things not possible with any other coin (use our edge). If we get away from this paradigm we will end up with a few delegates hosting the services and being the same as any other central service provider... so not any better.
-
Going forward i was hoping that we would be able to add web services to delegate nodes to do even more things.. ofcourse not through the HTTP server on the wallet but perhaps a hosted service either on the same node or somehow connected, maybe making a mesh network b etween all 101 delegates to allow a "decentralized" web app experience to do social logins, single sign on apps like getclef.com and other things not possible with any other coin (use our edge). If we get away from this paradigm we will end up with a few delegates hosting the services and being the same as any other central service provider... so not any better.
Say there was a prediction market for what will MSFT stock price be in 1Q 2016. Your answer that it is a simple token is inconsistent with BITUSD, which is a IOU. The American dollar is an IOU. However securities are fungible certification. Its a legalized and protected ownership in a publicily traded company. Now if you started offering IOU's for registered stocked, your walking into a legal minefield.The SEC will say it has the right to govern because you are stepping in their regulated court. Not to mention the federal financial bodies will see it as an attempt to avoid financial laws through these alternative payment pathways and will follow up with hefty suits. Look it up. Just google SEC fines for unregistered securities. Who would they fine? It would be the delegates since they are managing and administering the PM's.
Beware legislative and regulatory risk is the most detrimental of risks. It will put you out of business if you don't take it seriously.
Your vision for delegates is a whole another conversation.