Noted

Alex Eben Meyer

View full image

Online prediction markets, which allow people to bet small amounts on anything from election outcomes to Oscar winners, rival the accuracy of projections by financial professionals, according to a recent paper coauthored by a Yale researcher. The study showed that bettors in one prediction market forecast corporate earnings at the end of 2025 and early 2026 more accurately than experienced investment analysts. 

In prediction markets, individuals can bet “yes” or “no” on questions with a resolution date (e.g. whether a senator will resign before her term is up). With tens of million of dollars at stake, a correct bet can yield a big payout. To explain why bettors do so well against trained professionals, researchers suggest that professionals may have conflicts of interest—for instance, higher trading commissions for optimistic forecasts—or caution about predictions out of the norm. 

Cancerous tumors are adept at promoting their own growth, including by hijacking nerves from the peripheral nervous system that often embed themselves in tumors.  Studying a type of lung cancer in animals, Yale and other researchers experimented with inhibiting and activating various neurons in cancerous tissue to see the impact on tumor growth. They found that the tumors cause nerves to send signals to the brain, activating the sympathetic nervous system and triggering the release of the hormone norepinephrine at the cancer site. Norepinephrine activates specialized immune cells that then block another type of immune cell (T cells) from attacking the cancer, allowing tumors to maintain themselves through this “crosstalk” with the brain. Researchers say their findings could lead to therapies targeting the parts of the nervous system underlying brain-tumor communication.