The first graph comes from the 1964 book Recount, by Ronald F. Stinnett and Charles H. Backstrom, an authoritative account of the legendarily close election two years earlier for governor of Minnesota. That contest ended — after 139 days of counting and recounting — with a 91-vote margin of victory for Democrat-Farmer-Labor Party challenger Karl Rolvaag over the Republican incumbent Elmer L. Andersen. But it started with an electoral see-saw of first thousands and then dozens of votes that saw the lead change six times in four days.

The graph is a snapshot of a little more than a day of that upheaval, from 9 p.m. election night until midnight the next day. Two lines show the size of the gap favoring first Rolvaag and then Andersen, as tracked by the rival news services United Press and The Associated Press, who were competing for the latest counts from counties across the state. At first, the lines rise sharply into territory indicating a Rolvaag plurality, then dive steeply down as press tallies indicated an Andersen lead, the first of a pattern of plurality shifts that continued for days after the time period depicted in the graph.

The second graph shows a different 28 hours in last week’s counting, based on tabulations courtesy of the St. Paul Pioneer Press’s Political Animal Twitter feed, from nearly 5 p.m. Thursday, Nov. 6, to past 8 p.m. on Friday, Nov. 7. Over that time, the vote margin favoring U.S. Sen. Norm Coleman dropped from 590 to 221 votes, with a few sudden jolts but no swings in the lead.

How the gap in the Minnesota gubernatorial vote count changed over 27 hours, Nov. 6–7, 1962, in thousands of votes. Lead for Rolvaag in the upper half of the graph, lead for Andersen in the lower half.

How the lead for U.S. Sen. Norm Coleman changed in the Minnesota senatorial vote count over 28 hours, Nov. 6–7, 2008, in hundreds of votes.

If, over the long run, we expect Franken to win 51% of corrected ballots, his odds of winning the recount may be quite strong — in fact, he may be the prohibitive favorite depending on the number of recounted ballots.

UPDATE: Silver has already revisited this in a new post. His adjusted take on the situation: 

I hesitate to say this, but I think the evidence points on balance toward Franken being a slight favorite to win the recount.

UPDATE: In yet another installment, Silver says what we already knew: It’s a tie. 

 It is very, very close. 

A lot rides on the margin between Franken and Coleman staying small when votes are certified today, and on the belief, buttressed by exit polls, that under-counted ballots will break for Franken. Yet when he appeared Nov. 3 on MSNBC’s “Countdown with Keith Olbermann,” Silver didn’t put much stock in exit poll data:

Olbermann: Nate, when the exit polls leak out tomorrow … is there anything that you actually want to look at, anything that’s an actual valid indicator, or should you just throw them out as they come in?
Silver: No, totally throw them out. These things are not anything they’re cracked up to be. They’ve had a Democratic lean for years and years.

Silver may have been dissing exit polls’ utility in predicting election results, however, not in conducting after-the-fact demographic analysis.

In any case, it’s the kind of in-depth analysis Silver lavished only on select candidate preference polls during the heat of the campaign season. Now that it’s all over but the counting in a few places like Minnesota, the former baseball stats geek is apparently able and willing to turn his full mental powers on an individual non-presidential race to produce an at-length interpretation that leaves his readers’ heads spinning. “Swear to God, Nate, you’ve broken my brain with this one,” writes one. ”My goodness Nate — do you ever sleep? You need to get out more,” advises another. Adds someone named Bob: “Your sadistical analsis [sic] leaves me stunned.”