There is definitely something not right with the outcome of the Kentucky governor's race, though I'm still not ready to say it was definitely the voting machines.

I had a long conversation with Brad Friedman Wednesday on and off the air about the differences between Conway's totals and other Democrats down the ticket. The on-air conversation is at the top of this post.

Significant anomaly

The vote distributions on the downticket races are what's considered a "significant anomaly" by Bev Harris of BlackBoxVoting.org, as Brad explains:

Bev Harris, of BlackBoxVoting.org, who I spoke with earlier today, described the higher vote totals in the down ballot races as a "significant anomaly". She tells me that, at least until more records are requested and examined, the KY-Gov's race "has to be looked at as a questionable outcome, particularly because of the discrepancies in the down ballot races. More votes in those races and not at the top...that just doesn't happen."

(Here is a link to a helpful Public Records Request toolkit [PDF] from Black Box Voting for those of you who may be interested in helping to try and obtain some transparency in this race, as we also discussed on today's program.)

Now, here's the thing. It's entirely possible that the voting went exactly the way it was reported. It's also possible that it didn't. Brad says that at least in the more densely populated portions of Kentucky, they've switched from touchscreen machines to optical scanners, at least have a paper trail accompanying the marked ballots.

This is a good thing. However, it doesn't amount to a hill of beans if any of the machines simply read the ballots wrong, or there was a more complicated but still real hack called "man in the middle" between the ballot scans and vote tallies.

The only way to have confidence in these results is to count the ballots by hand.

Kentucky isn't going to do that. In fact, Brad told me that if they did a recount for some reason, it would simply be a process of feeding the ballots back through the same machines again, which doesn't lend itself to an accurate result.

Lots of spin, little factual basis

Brad entitled his show "Primal Scream: The Kentucky Governor's Race," and for good reason. Everyone has a reason why Jack Conway lost the race to Matt Bevin. Chris Hayes went on and on about low turnout Wednesday night. But just to set some perspective here, the current governor, Steve Beshear, won in 2011 with lower turnout than this year. The turnout this year was 30.7 percent of registered voters. In 2011, it was 28 percent, and Beshear won.

E.J. Dionne talked to Kentucky Rep. John Yarmuth who suggests it's an insider/outsider issue, so Donald Trump, Ben Carson and Ted Cruz have reason for hope. Also, turnout.

But the biggest surprise of the evening was Bevin’s sweep, and Rep. John Yarmuth (D-Ky.) saw it as instructive. “There is something to this outsider theory,” he said in an interview. “People are looking for something new and fresh.” Ben Carson, Donald Trump and Ted Cruz all have reason to take heart.

Yarmuth and other Democrats also pointed to low turnout among their supporters in nonpresidential years, aggravated by a Catch-22 in Republican-leaning states: Getting too close to Obama and the national party can hurt Democratic candidates in the broader electorate, but creating too much distance can dampen turnout among loyalists they badly need to show up.

Oh yes, the Obama/race thing again too. Except that one is harder to get my head around from a race perspective, given that Bevin's running-mate was African-American. In fact, I'd argue that having an African-American woman as his running mate might be an argument for why there could have been legitimate ticket-splitting.

By the numbers

I took a look at the numbers on a county-by-county basis, comparing the Conway/Bevin race to the Grimes/Knipper Race. Alison Lundergan Grimes received 66,656 more votes than Jack Conway. She won in nearly 40 counties; Conway won less than 10. The margins weren't even close in the counties he lost and she won.

I've uploaded the data in PDF format if you want to take a gander at it and number-crunch or make suggestions for how it might best be analyzed. These are unofficial results at this time, but they're what the press is using to craft their reasoning for how this race differed so greatly from the polls.

What's next?

To me, the fact that we're even having this discussion means that records requests should be made to get the tapes and any other publicly available information. Kentuckians deserve to have confidence in their elections. We all deserve to have confidence in our elections.

This isn't intended to be a sour-grapes effort. Matt Bevin should want to have confidence in the result as much as Jack Conway does, because if we don't believe our ballots are being counted with integrity, confidence in democracy declines, which erodes the democracy itself.

If you were to toss all of Grimes' votes into Conway's column, he still wouldn't have won. But at least it would have been within the margin of error. That's a result I would have more confidence in than a rout without any kind of reasonable explanation.

Yes, turnout matters, and in every election. I preach this every time we come to a midterm or off-cycle election. Republicans are far more reliable voters in midterm elections than Democrats are. That has to change if we want to make any progress at all. Every election, whether for dogcatcher, the local school board, city council, or president matters. Every one.

But on the other side of that, how those votes are counted matters at least as much as casting one in the first place.

So which among you would like to start putting together records requests? Let's make it a group project.