What Is A Prediction? What Is Crazy? Markets Can Help
By Ben Golden on October 27, 2015
Over at Grantland, Zach Lowe has published an article featuring 35 crazy prediction for the upcoming NBA season. Writes Lowe:
For this to be fun, we have to find the sweet spot between bat-crap crazy and probable. Let’s all be wrong together!
Lowe's forecasts aren't necessarily meant to be accurate, but I thought it would be fun to craft a series of questions on our new sports-themed prediction market SportsCast that mirror some of Lowe's predictions.
Note that some of Lowe's predictions are vague and not easily turned into prediction markets:
8. Jeremy Evans Becomes a Thing
9. More Teams Will Punt the Salary Floor in 2016-17
12. We Get the First Signs of Tension in Philly
13. The Spurs and Grizzlies Disappoint
23. By the End of the Finals, We Will See Significant Progress Toward a New CBA
25. At Least One Team Is in the Process of Being Sold
30. Tension Emerges Among the Owners in Sacramento
34. Tampering Becomes a More Public Issue
These claims aren't a great fit for prediction markets because it's not clear how to determine whether they've occurred or not. What kind of thing does Jeremy Evans need to become? This? What qualifies as a sign of tension? What constitutes the process of being sold? Etc. etc. It would be possible to turn these into verifiable prediction markets, but not without making a number of assumptions.
We've focused instead on verifiable questions like the following:
4. Chicago Gets the No. 1 Seed in the EastBen Golden/@BenGoldn is an Engineer at Cultivate Labs.
7. Indiana Makes the Playoffs
11. New Orleans Jumps at Least 10 Spots in Points Allowed per Possession
14. Matt Barnes Starts at Least 20 Games for Memphis
19. The Knicks Win the Battle of New York
26. The League Will Shatter the Record for 3-Point Attempts — Again
33. Detroit Cracks the Top 15 in Points per Possession