What specifically are you looking for? They show the studies that are used an input into the model, but the model that weights the input is private IP.
Like with 538, you see the input and percentages out, but not how the sausage is made.
Thanks but these are the inputs to the model. The output would generally be a probabilistic prediction of the outcomes, like 60% chance of Harris victory, 40% Trump, etc.
That’s the part you have to pay for I believe. Although the polling averages are interesting.
What specifically are you looking for? They show the studies that are used an input into the model, but the model that weights the input is private IP.
Like with 538, you see the input and percentages out, but not how the sausage is made.
No just the results. It’s paywalled so I can’t access it.
Boom!
Thanks but these are the inputs to the model. The output would generally be a probabilistic prediction of the outcomes, like 60% chance of Harris victory, 40% Trump, etc.
That’s the part you have to pay for I believe. Although the polling averages are interesting.
Here ya go
Thanks! Surprised they still have Trump favored but I guess the EC is a big factor.
The old version of his model at 538 has it flipped. Harris wins 57 out of 100 times. Still basically coin toss.
I wonder if his new model is accounting for some of the recent voting access changes.