Very excited about the prospects of futarchy for XTZ and appreciate the context murbard is taking here to not let the perfect be the enemy of the good for an initial implementation. We can optimize this mechanism over time, which is inevitable in any case. While others are busy with their endless research and roadmaps, let's deploy a working solution into the wild and show the world once again Tezos has tangible ability to progress. Let the debate begin!
In my view, having futarchy as a part of the Tezos protocol will not only allow for better identification and prioritization of on-chain improvements, it can be leveraged and deployed into a massive amount of use cases in outside markets and governance challenges. This should result in an exponential fundamental value increase beyond its on-chain governance utility.
I think it would be useful to use a real life example for helping make the article more comprehensible. Say a team proposes that tezos incorporates their implementation of ZK snarks (or any other example you wish). How would the prediction market look like for that? What metric would be used to tell if the update was a success?
One thing I'm not clear on is that if the no votes win on any proposal, we'll never be able to see how effective the the proposed upgrade would be. Isn't that a problem?
Hi. I am working at a university (economics), in particular on game theory and mechanism design. By chance, I happen to have published a couple of articles on laboratory experiments related to prediction markets.
I would like to get some clarification on the medium article. My understanding is that attached to each proposal comes an invoice stating the amount of XTZ the devs are receiving if the proposal is implemented. If I am not mistaken, the corresponding XTZ are introduced into the system through inflation. Naturally, devs have incentives to get as many XTZ as possible for their work, while the token holders would like to get hit by inflation as little as possible.
Since the added value (call it X) of a proposal is unknown (and there are important information asymmetries), the objective of the futarchy markets is to discover the added value X of the project that got selected in its cohort. Markets are likely to be rather imperfect, but this looks still much better than not doing anything. Clever idea.
What I do not understand is how the discovered added value X relates to the amount of XTZ the devs receives. That is, how is the added value shared between the devs and the token holders? For example, if the whole surplus goes to the devs, token holders will likely reject the proposal as there is nothing to be gained (taking into account that there is also some substantial risk that the estimate of X turned out too high). On the other hand, if devs do not get anything, they will dedicate their valuable time to something else. Consequently, it seems that there is a bargaining stage missing (between devs and token holders). To say it differently, once a proposal has passed, it should be possible to renegotiate the amount devs receive.
Thanks a lot for any answer that clarifies my doubts.
It is highly interesting to see how this project develops.
Sorry for the quick write-up, it is already quite late here.
PS: Obviously, a contract between devs and the network owners is not fully specified by only a price. What about timelines and deliverables? For example, it should be possible that funds for an approved projects are released step by step and that intermediate elections are held in order to decide whether the project should be pursued further (the quality of the product fulfils expectations). Etc, etc.
Such an invoice (in whatever forms there is) would be a part of the proposal. The proposals have a lot of latitude (at least initially) in what they kind modify, so it's not like there's a structured field called "invoice". In all cases, the prediction market would take into consideration the *net* benefit of the proposal.
If prediction markets did end up providing usable information on the counterfactual value of XTZ\_P, I would expect the actual realisable 'invoice' would be a quite heavily discounted fraction of X.
In the proposal discussed, there is no direct mechanism for the prediction market result to alter the protocol proposal, which must be submitted in its complete form, including any proposed fixed invoice, but the protocol proposal could be re-submitted using information gained from the prediction market.
I would say the optimal move would be for something like the original submitter to propose it with an 'aspirational' price reflecting a discounted version of a positive reception by the market, and then re-submit if necessary if the market does not tend to support such a valuation.