In point of fact, I took half a dozen A to C level statistics and data modeling courses at a top 20 university.
And - to your question?
I suggest you learn concepts like Berkson’s Paradox (collider bias) and Simpson’s paradox.
This is statistics 101 stuff.
Good lord man, that’s how Trump won. The electorate wasn’t static and you can’t assume static turnout generally, much less static turnout in smaller datasets.
Statistically impossible? Poppycock. ESPECIALLY in a midterm.
It IS politics. You don’t know math or statistics.
This is like Biden’s 81M votes. Obozo gets 68 and 65. Hill gets 65. Kammy gets 68.
That means there were 13M new voters for Joe pantload and then 13M Biden voters stayed home in 2024? NFW
This is all the evidence anyone needs to begin a data dive into the details.
Deal with Mssrs. Simpson and Berkson in our ornithology research.
Had a good math professor who told me to “not get out over your skies on what you think your data shows”
Well said.
bttt