A tied poll in FL-9 leaves our forecast where it was

One new survey in Florida's 9th District shows a close race. SEA Polling, fielded August 24-30 among 400 likely voters and published by floridapolitics.com, has Republican Dan Green and Democratic incumbent Darren Soto tied at 41 percent each.

Our forecast did not move on it. We give Soto a 99 in 100 chance of winning the seat and a mean margin of D+21.3, with a 5th-to-95th percentile range running from D+5.3 to D+37.2. The previous run, from 10:32 AM ET on September 4, had Soto at 99 percent with a mean margin of D+21.4 and a range of D+5.4 to D+37.5. That is a shift of about a tenth of a point in the expected margin and none at all in the win probability.

The gap between the poll and the model is worth stating plainly rather than resolving. A single 400-respondent sample is a thin input, and our estimate of the margin here is built mostly from other information. A tied result sits outside our central range but not outside the distribution entirely; the low end of that range is D+5.3, so the model already allows for a considerably tighter race than D+21.3. One survey at 41-41 is not enough to pull the mean.

Nationally, our average of 25 generic-ballot polls over the past 45 days stands at D+6.3. We give Democrats a 93 percent chance of winning a House majority, with a mean of 239.4 seats against the 218 needed. One caveat on that figure is ours: our simulation correlates error nationally, by census division, by state and by district, but we have no data to correlate demographically similar districts across the map, so the control probability is marginally firmer than a fuller error model would leave it.