I think I missed the deadline for the XIO accelerator…was in draft mode. Oh well, posting anyway:
Sports Wagering And Gamification (SWAG)
Summary Description
People who bet on sports would also likely wager on volume of sports betting. This could be total volume of betting in a given jurisdiction. This has value for several reasons: gamification; price predictions; and price discovery (the market will likely provide a better estimate that the regulators can predict). This could be done on a country by country basis, or state by state, for example in the US. I propose starting with a single US state where information is readily available (New Jersey), and if proves popular, expand out to other states, eventually covering the entire country.
Which metric will your synth track?
Monthly Sports Wagering Revenue for the state of New Jersey
How will you get data for your metric
A monthly report is generated and publicly available for free; for example here is February: https://www.nj.gov/oag/ge/docs/Financials/SWRTaxReturns/2021/February2021.pdf from the website: https://www.njoag.gov/about/divisions-and-offices/division-of-gaming-enforcement-home/financial-and-statistical-information/monthly-sports-wagering-revenue-reports/
These are state documents so are reliable. They may be some delay in the issuance of these reports (for example March 2021 is not yet available). This may be to our advantage however. Furthermore, this can be broken down into two more granular metrics:
- Casino/racetrack
- Sports wagering - internet
What collateral would you use for this synthetic
A stablecoin such as USDC. Ideally you want to use something that lowers the risk of liquidation for the user. People are very familiar with USDC and it has a very liquid market.
Describe how you would create this synthetic
This could be developed by the blockzero in-house team, or in collaboration with another team who have done similar work in this space such as degenerative.finance
Given it is US state government that issues the final report, there should be be very little if any room for any discrepancies or deviance. If so, a dispute can easily be put through the UMA voting system.
The synthetic would be designed so that people can go long or short the target total; that could based off the same month from the previous year, the mean of the previous 12 months or some other metric that would allow people take each side of the target, and ideally create a guide valuation of the monthly total (SWAG-NJ)
-
how you were tracking the most accurate measurement of the metric possible and how you would account for any discrepancies
There is no way to accurately track the ‘live’ valuation -
What issues might you encounter in the development of the synthetic
I don’t know how to code smart contracts and have no coding experience in solidity. I could help with the UMIP though. -
How you would make sure that people who would find the synthetic useful could access it.
Reddit, twitter, perhaps even take out adds on some New Jersey sports sites