Prepare to be shocked. You do say a lot of smart things about Arizona, which I appreciate. I love good intel.
Certainly I would never argue that markets are always rational, azcat. The problem is that nothing and no one are always rational. Like you point out, individuals all have their biases, so the nice part about knowledge aggregated from a group is that these biases at least have the potential to cancel out. Your evidence with the bots is interesting. Perhaps a few bots can outpace all other humans, but to what end? Some real human(s) is actually putting up these funds, and their money is on the line. Meanwhile, anyone else in the world can take advantage if the pricing is off. And the article makes a good point that PredictIt has pretty similar odds, even though it has a very small max bet. And as I pointed out above, Metaculus is moving toward Trump, and it doesn’t even use money at all. It could be that polymarket’s price is biased toward Trump, but I think that people will bring their own biases toward assessing that question, so we can’t have perfect knowledge either way.
Trump won Arizona in 2016 by only 3.5%. Since then, the state has voted for two D Senators and a D Governor. Really think Trump will outperform his 2016 total? Again, just looking at political party of the voter does not determine whom they voted for. Especially in Arizona. This article outlines four R former Trump voters who refuse to vote for him again. How many Arizona voters do they represent?
Polymarket is an easy target for the haters, because of Thiel, but they never talk about the other major gambling sites which all show the odds of the race in the same general range:
I think you are underestimating the R for Harris vote in Arizona. In Maricopa County, of those registered R or D, Rs have a 5% advantage. The difference was even greater in 2020. Biden won Maricopa County. I say whoever wins Arizona will be lucky to win by more than 2%.
Bloomberg/Morning Consult Polls Arizona Harris +0.4% 49.1%-48.8% Georgia Trump +1.5% 49.9%-48.4% Michigan Harris +3.1% 49.6%-46.5% Nevada Harris +0.5% 48.8%-48.3% North Carolina Trump +1.2% 49.6%-48.5% Pennsylvania Harris +1.7% 50.0%-48.2% Wisconsin Trump +0.3% 48.3%-48.0%
How can you see anything in early voting? 2020 trends were so affected by covid. Pubs looking like they aren’t going to be afraid to vote early. Too much shifting that there is no good comparable.
Because every professional political campaign reporter is seeing the early voting trends so far and saying if these trends continue, there is a good chance the Networks make the calls for Trump early in the night. There is plenty of time to turn that around but as of now, it is not good for Democrats. The Democratic campaign consultant on Mark Halperin Show said he flipped his Mets hat upside down and inside out today because Democratic voters need to rally in the early voting. The one thing everyone agrees on is Democrats are not hitting their numbers in early voting. I posted the 10 minute video clip in the in the early voting thread. It is worth a watch.
About those RCP polling averages ... here's what "flooding the zone" looks like ... On RCP Trump has 0.4% lead in North Carolina ... when 6 of the 9 polls used are either straight republican or aligned with republicans ... Note that all 3 of the non-republican polls have Harris winning. Being republican aligned doesn't necessary mean bad, as Silver points out, Emerson is not a bad poll, but take out just two of the known really bad ones -- Rassmussen and Trafalgar -- and it's Harris +3.... Could probably do the same for all the swing states. NC was just the first I looked at.
538 rates Morning Consult 2 stars, and 107th best. Around average. For comparison Trafalgar is 1 star, and 270.
Penn is the same ... 6 of the 10 are "republican aligned" pollsters ... all 6 have Harris losing. 3 of the 4 The non partisan polls have Harris winning ...
There is no way to tell if Republicans are voting earlier and are in no way incremental. There is no way to see if a few more republicans cross over. There is no way to project whether fewer democrats are voting early or by mail since there is no covid. There are just too many variables. I looked at the raw data upthread that I linked in that UF site and you just don’t have a baseline. It looks very similar to 2020 in every way to me which looks like a razor thin election.
538 is an absolute joke. They fired Rasmussen for being too accurate earlier this year. ABC owns 538 and they sent Rasmussen a letter demanding Rasmussen reveal proprietary info that would explain their polling success. I posted the ABC letter as pr9vided by Rasmussen ~2 days ago. Rasmussen would not give up their info and 538 dropped them. In the betting community we know 538 is a fraud. Trafalgar is one of the most accurate polls, that''s why I read their polls. Rich Baris was one of the most accurate if not the most accurate pollsters in 2020 but his poll was not a part of 538 that year. Look at the most accurate polls in 2020. Note where RASMUSSEN and TRAFALGAR are. Look where MORNING CONSULT IS. Morning Consult often does much worse in the ratings than this .THEY SUCK. Most bettors know this.