03-17-2020, 11:16 AM
(03-16-2020, 10:31 PM)Eric the Green Wrote:(03-10-2020, 02:13 PM)JDG 66 Wrote: Sounds like you're copping out, Eric... Or perhaps that equates to a close election?
...Again, my Word doc doesn't translate well, but here are my list of states and districts that have an essentially 0% chance of going for Trump:
CA (55), DC (3), HW (3), IL (20), ME1 (1), MD (10), MA (11), NY (29), OR (7), RI (4), VT (3), WA (12).
These essentially have a 100% chance of going for Trump:
AL (9), AR (6), ID (4), KS (6), KY (8), LA (8), MS (6), MO (10), MT (3), NE(-) (2), NE1 (1), NE3 (1), ND (3), OK (7), SD (3), SC (9), TN (11), WV (5), WY (3).
That's right, a close election is predicted.
I would include Delaware (3) among the certain blue states, especially with Biden the likely nominee, CT (7), and NJ (14), but then I would expect you to shortchange the blues.
Utah (6) doesn't like Trump, but will vote Republican anyway. Indiana is likely safe Republican (11).
There are fewer swing states now, not more...
I'm not short-changing anyone, Eric, it's just the math goes.
Again, some Worddoc wackiness:
NM //40.04%//48.44%//0+1/45
CT //40.93%//49.40%//0+2/45
NJ //41.00%//49.47%//0+2/45
DE // 41.92% //49.38% //2+0/45
CO // 43.25%//48.86%//2+2/45
VA // 44.43% [w/o Kaine 45.94%?] // (49.23%)//4+0/45 [w/o Kaine 6+4/45?]
ME (-)//44.87%//49.03%//4+1/45
MN//44.92%//48.85%//4+2/45
NV//45.50%//49.13%//5+2/45
NH//46.61%//49.15%//10+10/45
UT // 45.54% //46.41%//8+28/45
NE 2//47.16%//48.95%//15+12/45
WI// 47.22%//49.10%//16+11/45
MI//47.50%//49.31%//21+6/45
PA//48.18%//49.42%//27+9/45
AZ//48.67%//48.35%//32+11/45
FL//49.02%//49.58%//36+6/45
NC//49.83%//49.47%//43+1/45
GA//50.77%//49.52%//44+0/45
IA//51.15%//49.06%//44+0/45
ME 2//51.26%//48.97%//44+0/45
AK//51.28%//48.38%//44+1/45
OH//51.69%//49.37%//44+0/45
TX//52.23%//49.40%//44+1/45
IN//56.82% [without Pence 52.70%?]//49.61%//45+0 [w/o Pence 45+0/45]
OK. 1st Column is the state// 2nd Column is R 2016 vote // 3rd Column is the estimated percentage by which the incumbent (i.e., Trump) would win the state by a plurality. I explained my reasoning for this figure above (2016 3rd party; divided by 3.75; divided by half, since I'm assuming that the miscellaneous vote was roughly 50-50) // Last Column is the Chance out of 45 that the incumbent (DJT) would win by a majority + Chance out of 45 that the incumbent would win by a plurality (see that chart I posted previously).
There are a lot of assumptions here. This assumes that the incumbent's (i.e., Trump's) improvement in the national vote, over his 2016 performance, is a multiplier of 1.059 (as explained above). I assume that Pence will be DJT's running mate (although, as you can see, the results in IN probably don't matter), that Kaine will not be the challenger's running mate (I estimated Kaine's impact on VA's 2016 results; I'll explain later), and does not take the challenger's home state (DE?) or his running mate's home state (?) into account. If Biden gets the nomination (likely), then, based on Biden's impact on the 2008 vote (I'll explain my calculations later), DE would go to 0+0/45 for DJT.
UT's results are admittedly an artifact of McMuffin's campaign in 2016. I agree with you that my method probably underrates the incumbent's chances, but I'm just going with the math. Also, remember that this is the probability that DJT wins the state; the remainder does not necessarily go to the "D" candidate. If PBrower2a's fantasies of a serious 3rd party challenger to DJT in 2020 have anything to them, according to my method, it would most likely show up in UT (6 EVs for Mitt Romney, I guess).
Note that when this method shows DJT having a chance of winning NM, CT, or NJ, that it's a small one, and then by a plurality. It also gives the "D" an essentially 0% of winning AK, TX, or IN once you allow for winning by plurality.
In 2004 (the pattern), the miscellaneous vote narrowed from 3.75% to 1.00%, as more people pick one side or the other than they did in the first election, but this does not necessarily narrow the number of swing states.
One other thing: I did most of these calculations in January or February 2017. The only change was in finding a way to calculate the miscellaneous vote, and thereby calculate a way of to figure out the chance of winning a state by a plurality, rather than just majority (a refinement the my system didn't have yet for my 2016 prediction). Again, these are the only things that might seriously the national calculation (multiplied by 1.059) or the state-by-state calculations:
Economic Downturns
Security Problems
3rd Party/Independent Candidate (symptom or cause?)
Loss of the incumbent POTUS/VPOTUS candidate
Home State of the POTUS/VPOTUS candidates