Friday, October 7, 2016

Election Update 10/7/16

So, I've been hearing a lot about how Clinton is rebounding, and it's a Friday, so polls should be slower over the weekend. That being said, it's time for an election update since it's been over a week since my last one.

Aggregate polling - 4.5% Clinton (up from from 2.7%)

Yep, she's definitely rebounding back to where she was previously looking at the national polls at least.

Aggregate polling with third party candidates - 3.7% Clinton (up from 1.6%)

Same here.

Electoral College

Arizona - 0.7% Trump

Colorado - 5.0% Clinton

Florida - 3.2% Clinton

Georgia - 4.7% Trump

Iowa - 4.7% Trump

Maine - 5.0% Clinton

Maine CD2 - 5.4% Trump

Michigan - 6.8% Clinton

Minnesota - 4.0% Clinton

Nevada - 1.2% Clinton

New Hampshire - 5.2% Clinton

North Carolina - 2.6% Clinton

Ohio - 1.2% Trump

Pennsylvania - 7.5% Clinton

Virginia - 8.2% Clinton

Wisconsin - 4.7% Clinton

Clinton looks to have rebounded significantly and is back to where she was back in August, roughly. Let's see how it pans out in the electoral college.

Election Scenarios

I've decided to change a couple things about my methods here. Instead of having incremental changes for every point difference, I'm focusing on four levels, the most accurate, a one point lead, a three point lead, and a six point lead. I'm also going to explain where the tipping point is for the election, currently and try to estimate the percentage chance of that tipping point being reached. I will explain the methodology in another section.

Most likely scenario with no swing states <=6% lead) - 236-164 (no winner)

Most likely scenario with no swing states (<=3% lead) - 301-187 Clinton

Most likely scenario with no swing states (<=1% lead) - 322-205 Clinton

Most likely scenario (best guess) - 322-216 Clinton

Clinton + 1 - 333-205 Clinton

Clinton + 3 - 351-187 Clinton

Clinton + 6 - 374-164 Clinton

Trump + 1 - 322-216 Clinton

Trump + 3 - 301-237 Clinton

Trump + 6 - 236-302 Trump

Methods for new sections

 Okay, so I'm going to find the current tipping point of the election. I am also going to estimate the percentage chance that tipping point can occur. I will be assuming a 3 point margin of error (6 point spread), and am assuming a standard deviation is about 3 percentage points. This will allow me to estimate just how far outside of the polls given that an outcome can occur, assuming a standard bell curve. I will then use this chart to estimate the approximate standard deviation figure out what percentage chance a candidate has of winning. If there's a flat out tie, each candidate has a 50% chance of winning. If the tipping point is in my +1 model, the leading candidate has about a 62.9% chance of winning. +3 is about an 84.1% chance of winning. +6 is about a 97.7% chance of winning. One limitation of this approach is it ignores that one state may flip but others may not, which means that the candidate in the lead may have even a higher chance of winning than indicated here. As such, I am assuming that support rises and drops for a candidate universally across the country, which may or may not happen. We may see a concentrated push in a swing state but not elsewhere. As such, this isn't a perfect method, but it's something I'm going to be trying for now and perhaps refining. Without further ado, this is how this week's predictions compare to my last analysis'.

Current statistics

Current tipping point: Trump + 4 - 262-276 Trump

The state that tips it: Minnesota (+4.0 Clinton)

Standard deviation: 1.33

Current chance of Clinton winning: 90.8%

Current chance of Trump winning: 9.2%

Previous analysis' statistics

Previous tipping point (9/27): Trump + 0.6 - 263-275 Trump

The state that tipped it: Florida (+0.6% Clinton)

Standard deviation: 0.2

Previous chance of Clinton winning: 57.9%

Previous chance of Trump winning: 42.1%

Discussion

Clinton has rebounded significantly and once again has a good chance of winning the election. In my previous update it looked like it was very close where Clinton could have lost, and running my new analysis, I would have estimated Clinton's chance at a mere 58%. For people who like Trump, that's a good sign. For those who like Clinton, that's scary as heck. This time, Clinton has about a 91% chance of winning, which is pretty darned strong. Things can still turn around, there's a month until the election after all, and we saw how quickly things can change at the drop of a hat. Still, if the election were held today, you have a good chance Clinton is going to win.

No comments:

Post a Comment