At this time the recount situation in MN is being gamed with more than 3600 challenges to ballots. The rate is accelerating and could eventually exceed 5000. 82% of the vote has been recounted appearing to show Coleman up by 231. It is clear that Coleman’s lead could easily be swallowed by the results of the challenged vote canvass. But how to tell what is in those X thousand challenges.
Like a lot of youse guys, I got hooked on 538 recently --so I’m gonna take a KraK at it here on the Franken-Coleman dustup.
We have results from 80 MN cities, with no challenges from either side (43,894R, 27,114D). They have reported 100% returns for pcts. & voters. Franken gained 6 votes in those 80 cities for a rate of 1 per 11,834 voters. If the recount is of 2.42M votes, then he will gain 204 total at that rate and lose by 11 to Coleman who led by 215 before the recount began.
However one of the cities, Oak Park Heights, has very odd results @ +8 for Coleman with only 2034 votes cast. Franken had 10 of his 994 votes taken away by election officials without a single challenge. Somewhere from my old life I recall it’s OK to toss outliers, so I ditched Oak Park Heights & just to be fair I also ditched Shorewood @+5 for Franken – even though it is the largest city of the 80 and its results were normal looking. Gotta be fair & balanced.
From the remaining 78 cities in the sample, (40,190R, 24,514D) -- Franken gained 9 votes for a rate of 1 per 7,189. If the recount is of 2.42M votes for senator then he will gain 336 total on Coleman and win by 121.
If you recalculate with Shorewood back in the mix, Franken gains 14 actual votes from 79 cities and he goes to a 491 votes projected gain.
If the places where Franken polled higher than in these republican cities, give him higher percentages for the recount, he wins in a veritable landslide.
Why do I like these projections? Because they are based on raw vote tallies that are final and were produced exclusively by election officials, it’s the closest thing that we have ATM to what an official end-product will look like. There are a lot of cities & voters in the sample & it Just seems like ~70K is a good size sample, even for a blogger.
Major caveat is Minneapolis looks ass backwards @ +51 for Coleman. But Coleman has 150 challenges to only 98 for Franken. There are ~181,000 votes that need recounting & it’s 59% complete. Franken will gain ~15 to 37 here if the above ratios hold up. He will gain more if the ratio improves for him with the demographics. Anyway, that’s the point – to look beyond the challenges & guess what is in the ballot boxes.
Too bad we’re not getting the recount data on the 3rd party guy Barkley. That info could shed some light on the plight of some ballots.
Nonetheless, I like Franken by 121.
Now about those absentee ballots.....,,,