Election Update 10/30

Okay, polling seems to be slowing down and I'm seeing stuff saying that many of the polls being released this week are the "final" from that polling company, so I wonder how much the data will change from here. I still plan on doing at least one more update on Monday-Tuesday, but whether I do one over the weekend remains to be seen. Polls are scant on the weekend and if most pollsters already dumped their stuff this week, well, the data might not change much. The data hasn't even changed much from yesterday, but I do think Friday is a good time to get an idea of what the race is. I planned on doing one around now regardless of the results. That said, as potentially the second to last forecast, let's look at the state of the race.

2 way: Biden +7.8%

4 way: Biden +7.8%

Well, there's almost no spoiler effect to be noted. As I've stated many times the spoiler effect doesn't exist on a statistically significant level. Compared to 2016, Biden is almost 6 points ahead of Hillary. On 11/4, 4 days before the election, Clinton was down to a 1.5% popular vote lead. So whatever prediction I make now is going to be fairly accurate. I don't expect that much of a shift from this baseline. Trump's window to make a miraculous comeback is closing fast, and unless the polls are off or he pulls some major bullcrap, I don't expect much to change.

States	Winner/Margin	Z score	% D Win	% R Win	EV D win	EV R win
Michigan	Biden +6.5%	-1.625	94.8%	5.2%	232	322
Nebraska CD2	Biden +6.5%	-1.625	94.8%	5.2%	233	306
Wisconsin	Biden +6.4%	-1.600	94.5%	5.5%	243	305
Minnesota	Biden +4.7%	-1.175	88.0%	12.0%	253	295
Nevada	Biden +4.0%	-1.000	84.1%	15.9%	259	285
Pennsylvania	Biden +3.6%	-0.900	81.6%	18.4%	279	279
Maine CD2	Biden +2.3%	-0.575	71.7%	28.3%	280	259
Florida	Biden +1.2%	-0.300	61.8%	38.2%	309	258
Iowa	Biden +1.2%	-0.300	61.8%	38.2%	315	229
North Carolina	Biden +1.2%	-0.300	61.8%	38.2%	330	223
Georgia	Biden +0.8%	-0.200	57.9%	42.1%	346	208
Arizona	Tie +0.0%	0.000	50.0%	50.0%	357	192
Ohio	Tie +0.0%	0.000	50.0%	50.0%	375	181
Texas	Trump +2.3%	+0.575	28.3%	71.7%	413	163
Alaska	Trump +6.0%	+1.500	6.7%	93.3%	416	125
South Carolina	Trump +7.5%	+1.875	3.0%	97.0%	425	122
Missouri	Trump +7.5%	+1.875	3.0%	97.0%	435	113
Montana	Trump +8.0%	+2.000	2.3%	97.7%	438	103

So one change I'm making in terms of my color scheme. I'm going to make anything under 1.5% whited out like they were under 1%. This will be represented with a pale pink or blue color in the electoral college map. I'm doing this because I keep seeing these polls show up as like 1.2%, 1.4%, and then the next day they're 0.7%. I don't feel very confident in these averages and the races in that range bounce around way too much to be reliable locks either way. This isn't a huge change. Basically it's a shift from 60% being the barrier of confidence there to 65% or so, and I think that's fair. I mean, Florida was, 3 days ago, a soft red state. Iowa and North Carolina have been trending blue, but I really don't feel very confident in them at all. They're also more or less coin flips, especially given their trends. 

I'll go more into my predictions in these states on/right before election day, but Ohio, Arizona, Georgia, North Carolina, Iowa, and Florida are all on my very close watch list. I really don't feel confident in them either way and they have a strong propensity to flip on a dime. States that are more in the 2+ range I'm more confident on, and think they will ultimately go as predicted, but there is a chance they don't. States above 4 or so I don't think will flip at all. 

Now, you might be wondering, why did the polling averages go down in a lot of states? Well, some of this is just restructuring of how averages were calculated on RCP, or in how I interpret the averages in cases in which no clear average is given, but some of it is what I'm calling the "Trafalgar effect", after the Trafalgar Group polling organization. I've talked about them before. They seem to think there's a hidden Trump vote in the polls and they're skewing a lot of averages to be less pro Biden by releasing polls that are like Trump +2 when the average is like Biden +6. That's why Nevada and Michigan are down so much. They haven't really changed otherwise.

That said:

Chance of Biden win: 81.6%

Chance of Trump win: 18.4%

Deciding state: Pennsylvania at 3.6% Biden

Most likely Electoral College map: 346-163 Biden (outcomes range between 346-192 Biden and 375-163 Biden)

I know, this seems worse, but again, this is primarily the Trafalgar effect at work, if not for some really skewed PA polls, Biden would be up 5.2, and he would have a 90% chance of winning there. Regardless, even this skewed result is much better than the 60% I gave Clinton in 2016. Biden's lead is actually comfortable. And while there is almost a 20% chance he can still win, I would consider it a narrow possibility. Biden is doing much better than Clinton was at this point in 2016 and Trump's path of victory is fairly narrow. Everything has to go right for Trump to win, and I am not going to expect that to happen again.

Senate forecast

States	Winner/Margin	Z score	% D Win	% R Win	Seats D win	Seats R win
Michigan	Peters 7.6%	-1.900	97.1%	2.9%	46	55
Minnesota	Smith +6.3%	-1.575	94.2%	5.8%	47	54
Maine	Gideon +4.0%	-1.000	84.1%	15.9%	48	53
Arizona	Kelly +3.5%	-0.875	80.9%	19.1%	49	52
North Carolina	Cunningham +3.2%	-0.800	78.8%	21.2%	50	51
Iowa	Greenfield +1.5%	-0.375	64.6%	35.4%	51	50
Georgia 1	Perdue +0.2%	+0.050	48.0%	52.0%	52	49
South Carolina	Graham +2.5%	+0.625	26.6%	73.4%	53	48
Montana	Daines +3.3%	+0.825	20.5%	79.5%	54	47
Kansas	Marshall +4.0%	+1.000	15.9%	84.1%	55	46
Georgia 2	GOP +4.5%*	+1.125	13.0%	87.0%	56	45
Texas	Cornyn +7.5%	+1.875	3.0%	97.0%	57	44

Chance of democratic control: 64.6%

Chance of republican control: 21.2%

Chance of a 50-50 split: 14.2%

Most likely outcome: 51-49 DEM

Not much has changed. Due to Cunningham going up in North Carolina, the chance of GOP control declined to a mere 21%, whereas the chance of a tie improved to 14%. Other than that the race is fundamentally the same. 

As I said I will likely make at least one more forecast on Monday-Tuesday, and I might also do something over the weekend if anything justifies it.