Trump Job Approval

Pollsters wrong in 2024:42.0 / 55.1-13.0
Pollsters right in 2024:46.2 / 51.0-4.8

2026 House Forecast

GOPDEM
Democrats +3216219

2026 Senate Forecast

TXMEOHMIGANHNCMN
Gap+4.5+2.0+1.4+1.4+1.4+3.6+6.2
Count5051525347464544

Polling Aggregator Methodology

The Hedgehog Report Polling Aggregator includes the most recent polls of each state with weighting applied in four ways:

  1. Age. The newest poll receives the highest share of the average. Then as time passes, the same poll keeps being reduced until its advanced age drops it out completely.
  2. Outliers. When a new poll is received into the average that stands out as distinctively different from other polls in the same sampling dates, then a smaller share value will be assigned to that poll throughout it’s inclusion in the aggregator. If such a poll turns out to be confirmed by subsequent polls, then normal share value is restored.
  3. Tiny Samples. Sometimes polls are released with a very tiny sample size, resulting in a very large margin of error. These will be assigned a smaller share of the average, proportional to the severity of the undersized sample.
  4. Partisan poll releases. Often campaigns release internal polling. These polls, along with releases from little known pollsters with no track record are assigned a smaller share of the average.

Independent candidate Robert Kennedy has qualified to appear on the ballot in a growing numbers of states, with the expectation that he will appear on the ballot in most states. The polling average used for the Aggregator will include the poll numbers with Kennedy included, if sufficient polling is available.

Chase Oliver (L), Jill Stein (G), and Cornel West (I) have not polled sufficiently in any state to warrant inclusion in the tracking or averaging.

One response to “Polling Aggregator Methodology”