Resolves YES if Swift Centre has a better Brier score than Polymarket, across their five forecasts. Resolves NO otherwise.
Context: Swift Centre claims that Polymarket is not the ultimate source of truth and that their forecasters can beat Polymarket.
https://www.swiftcentre.org/publicforecasts/yk77y2zct07ttcl00lmuk2x2s629fh
Here were their five forecasts:
1) The Swift Centre forecasters saw a ~4% chance of the United States acquiring Greenland in 2026, distinctly lower than the 12% that was on the relevant Polymarket question.
2) The likelihood of Russia and Ukraine agreeing to an indefinite ceasefire of peace agreement this year was forecasted at 22%, considerably lower than the ~43% odds on Polymarket.
3) The forecasters saw a ~6.5% chance of the US taking control of the Panama Canal in 2026, lower than Polymarket’s odds of ~15%.
4) The likelihood of Iranian Supreme Leader Khamenei being out of power by the end of March 2026 was pegged by the Swift Centre forecasters at 27%, compared with Polymarket’s 26%...
5) ...while the likelihood of the US undertaking military strikes on Iran by the same date was estimated at 68%, notably higher than Polymarket’s ~55%.
If there's controversy as to the outcomes of the Polymarket markets (Swift Centre disputes the resolution or something), then I will side with Polymarket's interpretation; after all, the Swift Centre are forecasting Polymarket questions, not their own analogues.
People are also trading
@AlexBen
Swift is the favorite based on ensamble starting probabilities - and turning a coinflip when Polymarket knows the truth.
I had similar results. I don’t know Swift’s current numbers, but I’m going to take another look at Polymarket’s now. Your post made me curious again.
But what makes it a little bit boring indeed: Khamenei can never be the tiebreaker, so in practice we’re really down to four. He’s also not independent from regime change - and I assume he had to be included and mentioned in the blog post because of the disagreement about regime fall.
I still appreciate the 50:50 outcome for a claim against Polymarket, although with more markets and more than just 5 (4) questions it would have been more competitive.
-
Yes, it’s only five questions - but it’s still a claim against Polymarket.
Even if Polymarket knew the “true” probabilities perfectly, the outcome is basically a coin flip and doesn’t favor Swift by much - using the probabilities at the time the market opened.
So why not? Kind of fun - if my math is correct :)
-
@SemioticRivalry the funny thing is that Polymarket still has a pretty decent chance to beat them anyway
@bens yeah I'll give them credit because you could very trivially rig this game such that you will essentially never lose, and they haven't done that. but it's still quite silly

