So, polls keep shifting. I locked in my predictions last night, but just like this, the AZ race is now in the R column, and the margins on the PA and GA races have gone more in the red direction. As it stands, AZ is 0.3% in favor of Masters, indicating a 53% of republican victory, PA is 0.4% in favor of Oz, indicating a 54% chance of republican victory, and GA is 1.4% in favor of walker, indicating a 64% chance of republican victory. Given dems need all 3 races to get 50 senators, this is not looking good for dems, and I would now say that we're talking a 53-47 republican senate, with a 64% chance of republican victory.
How reliable is this? Well, it's still in coin flip territory, but I want to discuss two factors, one of which goes in favor of the GOP and one that might indicate polls OVERSTATE republican chances. First, the one in favor of the GOP. Polls are a lagging indicator of support, and polls might actually UNDERESTIMATE republican support here. This is because polls are a snapshot in time, and if we're seeing movement in one direction or another, we only see it at the time the poll was taken. This means polls can lag actual attitudes since by the time polls are published we see further movement in that direction. As such, we could potentially see any race within a point or two and going more republican over time actually lead to republican wins, even if the margin is still in the dem category. This happened in 2016. As last minute voters made up their mind, they swung trump, and while we saw some movement toward trump before hand, the results were a lot more pro trump than indicated.
To be fair, mail in voting could mitigate this and we've seen this since the election. I voted like back on like October 13 or something. I mean, I voted straight democrat, and my vote was locked in when the race was a lot more pro democratic. And while i wouldnt change my mind, some who voted early...might. To be fair, those who vote early tend to be more democratic in the first place (keep in mind republicans think mail in votes are "fake votes"), and since you need to request a ballot people might be more politically informed. What swings elections are casuals. The 10% of the population or so that makes up their mind last minute, and most indicators have that going republican. So how much will mail ins influence this? Not much probably.
The one factor that might work in the democrats' favor is the "Trafalgar Effect". It's a term I coined to show how blatantly biased right wing polls tend to skew polling averages. Some pollsters just end up adding points one way or another to account for biases they think are present, with many of them being republican. The term is named after the pollster Trafalgar, who always has these skewed off the wall numbers. Now, to be fair, in 2020, trafalgar actually ended up being the most accurate, although they did so with a bad and blatantly unscientific methodology. In 2020, I did try to correct for this effect, but given they ended up being correct, I decided this time, you know what, polling averages it is. Still, if we dont have the same red lean that we had in 2020, this could mean that polls underestimate republican chances and democrats will do better. We did see how many "swing" house districts are seemingly leaning blue, and even some races that lean red on RCP...ended up having blue results. SO maybe the results could be more blue.
I don't know. I just report the numbers and calculate probabilities based on them. If the numbers are wrong, my predictions are wrong. We'll see either way in the coming hours and days. As of writing this polls start closing in 3 hours, although for some races it might take days or even weeks for definitive results.
No comments:
Post a Comment