Holy smokes…
For the longest time I’ve been a polling nerd—both for investing and for teaching my students.
This is the first year I’ve finally started to buy into the science that betting markets are superior to polling.
Two weeks ago, based on the polls, Democrats had about a 1-in-3 chance of winning the Senate. They need to flip four seats.
From the polling landscape:
1 likely Dem: MN
1 lean Dem: NH
4 toss-ups: GA, ME, MI, NC
2 lean GOP: AK, OH
2 likely GOP: IA, TX
The lean-Dem race in New Hampshire has an outside chance of flipping Republican because Sununu is popular.
Alaska, Ohio, and Texas also have some outside chance of flipping blue, since several strong Democratic candidates are running.
Still, based purely on the polling, Democrats winning the Senate looks like a long shot.
But Kalshi currently has Democrats winning the Senate at 50%, with Republicans at 37%.
That pushes against my math/stats instincts from polling. But the theory behind betting markets is that when people with different backgrounds/information have real money on the line, their combined judgment can provide faster and sometimes better signals about the day-to-day situation than polls alone.
I'm juggling 2 different sets of contrasting theories however at this point, I believe Kalshi is the better metric.