Trends
-
The only things I can think of is a rounding difference or editorial disputes…
-
Rounding would be my guess, but that indicates they are not careful with their code. They need to round before comparing to their defined ranges. Then there is no discrepancy between the number you see and the number used to get the category. I write scientific software and I know these things from painful experience.
-
Rounding would be my guess, but that indicates they are not careful with their code. They need to round before comparing to their defined ranges. Then there is no discrepancy between the number you see and the number used to get the category. I write scientific software and I know these things from painful experience.
-
@Horace I’m no expert but I doubt it’s rounding. These are Monte Carlo analyses after all. I’m guessing editorial disputes about what to do about the margin of error.
@Horace I’m no expert but I doubt it’s rounding. These are Monte Carlo analyses after all. I’m guessing editorial disputes about what to do about the margin of error.
They have a fresh discussion every time about what 54.95% means? The italics seem to indicate a formalized category. I looked on the website but didn’t find definitions for the italicized terms.
-
If you folks are talking about the FiveThirtyEight projections and terminology, note that they also have three different “models”, {Lite, Classic, Deluxe}, each with increasingly more human/expert “adjustments” added to polling data. Lots of potential to complicate their definitions for terms like “lean,” “toss up,” and “dead heat.”
-
It has to be a rounding thing…
-
Uh oh!
-
Speaking of trends, O’Dea in Colorado has just pulled into toss-up territory on RCP… https://www.realclearpolitics.com/epolls/2022/senate/co/colorado_senate_odea_vs_bennet-7773.html
Bennet was a double digit favorite a few weeks ago…
