polls vs. cash · updated daily

The Hope Index

The audacity to hope the polls are wrong. People are putting real money on races that are polled almost daily, and sometimes the money and the polls disagree. Polls are cheap talk. Bets cost something. Is the money real sentiment, or partisan cheerleading with a wallet? The Hope Index tracks the spread by race until November. Then we'll see where the smart money was.

Every race is read from the Democratic side. Amber: the money is more Democratic than the polls. Teal: more Republican.

The money updates daily. The poll side is an average updated weekly.

loading…
The board sorted by wedge
◂ money leans R money leans D ▸

How the wedge is computed

Market prices are from Polymarket, fetched daily (Kalshi's prices are also available as a toggle). The poll side is the RealClearPolling average for each race, refreshed weekly and converted to a win probability with a simple polling-error model. Senate control uses Decision Desk HQ's forecast instead, as there is no margin to convert. Wedge = market − polls, in percentage points. Every row is read from the Democratic side, so a positive wedge means the money leans more Democratic than the polls and a negative one more Republican.

The size of the wedge is the hope (someone is betting the polls are wrong). The sign is whose hope it is. Wedges near 0% or 100% should be read loosely, since markets rarely price above 90¢ or below 10¢.

Market history before launch (Oct 9) is backfilled from the prediction markets from mid-August. The poll line before launch is a 28-day rolling average of the individual RCP-listed polls (using at least three polls and not RCP's own average), shown as a dashed line. Full details can be found in the methods & assumptions. NOTE: this is a data project by Daniel A. N. Goldstein, not investment advice. Code & data on GitHub.