Man vs. stats: How well did I predict the election?

I did glance at a few polls, but more importantly I read editorials in dozens of local newspapers of the battleground states, and put that all together with my own intuitions on voter behavior, election history, and sense of risk (I’ll admit, some of my calls were quite simply gambles – I’ll explain in depth down more).In the end, the four states that cannot be called right now + the one I really called wrong (North Dakota) were the exactsame five states I predicted would be tossups. I knew these were the risky bets: North Dakota especially, lol. That was more of a pot shot hoping-for-a-miracle-to-make-me-look-smart-bet.

So, in other words: I called it!

********** OBAMA: 365 **********Hawaii (4)
Washington (11)
Oregon (7)
California (55)
Nevada (5)
Colorado (9)
New Mexico (5)
Montana (3) – WRONG!
North Dakota (3) – WRONG!
Minnesota (10)
Iowa (7)
Nebraska (1/5) ***
Missouri (11) –> traded for Indiana (11) ***
Wisconsin (10)
Illinois (21)
Michigan (17)
Ohio (20)
Pennsylvania (21)
New York (31)
Virginia (13)
New Hampshire (4)
Maine (4)
Massachusetts (12)
Rode Island (4)
Vermont (3)
Connecticut (7)
New Jersey (15)
Delaware (3)
Maryland (10)
Washington D.C. (3)
Virginia (13)
North Carolina (15) ***
Florida (27)

********** MCCAIN: 173 **********

Alaska (3)
Idaho (4)
Utah (5)
Arizona (10)
Wyoming (3)
Montana (3) ***
North Dakota (3) ***
South Dakota (3)
Nebraska (4/5) ***
Kansas (6)
Oklahoma (7)
Texas (34)
Arkansas (6)
Louisiana (9)
Mississippi (6)
Alabama (9)
Georgia (15)
South Carolina (8)
Tennessee (11)
Kentucky (8)
Indiana (11) —> traded for Missouri (11) ***
West Virgina (5)

*** Note: North Carolina, Indiana, Missouri, and Montana are so close, it is probably too early to really say for sure who won those states. As I type this, major networks won’t call Montana, but I’m going to go out on a limb and say it will be a loss for McCain just from the early results. I’ll man up and call it! =o)

North Carolina (15) – Predicted for Sen. Obama – currently leaning Obama.

Sen. Obama has a 12,500 vote lead with 100% precincts reporting. Ouch! Bob Barr is going to get a few scowls for this one: He captured just over 25,000 votes! What is REALLY interesting is how many votes the Libertarian governor candidate received: 120,500. The Senate race saw 131,000 libertarian voters. If you compare the difference in total votes in between the two major parties in the Governor and Senate races: It becomes apparent that Libertarians actually tend to split their votes evenly between Sen. Obama and Sen. McCain. This assumes that basically everyone in N.C. who voted Republican in the senate & Governor races also voted Republican in the Presidential race. Same goes for the Democrats. This also assumes that the remaining 25,000 who voted Libertarian in the presidential race would have split the same way – or would have even voted at all. In other words, it is highly unlikely their votes would have affected the final outcome if they voted only for a Democrat or Republican. Sen. Obama still would have won. Still, this was uncomfortably close. It probably won’t be officially called for a while…

UPDATE: It looks like Sen. Obama will hold onto his lead in North Carolina, although it is dangerously close. 2,101,991 votes counted for Sen. Obama and 2,089,828 votes counted for Sen. McCain. 25,181 votes counted for Bob Barr – sure to earn Bob Barr the “spoiler” title for this race. There were 131,072 libertarian votes in the N.C. Senate race, and 120,608 in the governor’s race: It appears that only the die-hard libertarians stuck with Bob Barr on the presidential ticket, which makes me suspicious that they would have even voted for President at all if he was not on the ticket. Also, libertarians tend to pull support from both the left and right – and there is pretty good evidence within the demographic to indicate that they actually pull more from the left, so… I highly doubt those “spoiler” accusations. If I’ve said it once, I’ve said it a million times: votes don’t “belong” to a Democratic or Republican candidate by default and any deviation from that was “stolen” from one of the main two parties. Votes belong to individuals, and they will give them to the person they think is best. If Sen. McCain failed to convince enough people in North Carolina to vote for him to win the state, it is no one’s fault but his own.

Indiana (11) – Predicted for Sen. McCain – went for Obama by a 1% margin

Sen. Obama leads by 23,500 votes with 99% precincts reporting. This is VERY close… I doubt it will be called anytime soon. Again, Bob Barr collected 29,000 votes – enough to put Sen. McCain over the top – assuming you buy into the malarkey that without Bob Barr in the race all 29,000 of those people would have voted for Sen. McCain. I don’t buy that. Many of those voters would have stayed home, and the remaining would have split between Obama & McCain. Let’s face it though: The tally there is still WAY too close, and there may be more surprises in store!

UPDATE: With 100% precincts reporting, Sen. Obama has officially been called as the winner of Indiana with a 23,000 vote lead.

Missouri (11) – Predicted for Sen. Obama – currently barely leaning McCain

Sen. McCain leads with 36,000 votes and 91% precincts reporting. The only tricky thing here is that St. Louis City has only reported 80% and Sen Obama has an 83%-16% advantage in this area. The rest of this district could give Sen. Obama slim the margin by 20,000 votes… but that still wouldn’t be enough to make up the total. This one will be too close to call for a while I think. Here’s one where people are going to call Ralph Nader a spoiler: Ralph Nader currently tallies about 16,000 votes… It is possible his tally combined with Sen. Obama’s might actually out weight Sen. McCain’s, but that’s tough to call at this moment. Very tough… This will probably be called in favor of Sen. McCain before North Carolina or Indiana are called.

NOTE: The only thing Sen. Obama has going in his favor is that Missouri has voted with the winner in every election since the civil war, save one. If Missouri breaks for Sen. McCain this election, it will be a MAJOR oddity compared to the historical trend (but perhaps not a statistical oddity).

UPDATE: This has played out EXACTLY as I predicted: St. Louis City finished reporting all its precincts, and now the spread is only 15,000 votes with Sen. McCain on top and 97% precincts reporting. I highly doubt Sen. Obama will pull it off. FiveThrityEight.com has called it for Sen. Obama, however. The guy did note that he “still hasn’t done the math” – which I did in about 30 seconds, so I’m guessing that’s why he’s making the bad call… Once he does the math, he’ll see this is super close, but highly likely to stay with Sen. McCain. I am going to have to correct something though: The Ralph Nader spoiler theory is kinda shot to hell by the fact that Bob Barr & Chuck Baldwin received a total of 19,000 votes to Nader’s 17,000.

UPDATE 2: Wow! 99% precincts reporting, this is now a dead heat! McCain only leads by 398 votes! Welcome to Missouri – the “2000 Florida” of 2008.

UPDATE 3: The current tally shows Sen. McCain in the lead with 1,442,613, and Sen. Obama with 1,436,745. This is a mere 5,868 vote lead for Sen. McCain. Interestingly enough, there are said to be a large number of provisional ballots in Missouri to be counted. Why does this matter? Most of the provisional ballots are in St. Louis City – strongly democratic, and it is also rumored that the demographic most likely to be forced to cast a provisional ballot are low income and new voters – who are also more likely to support Sen. Obama. This state could still go for Sen. Obama. I hope it does for two reasons: (1) This would put my electoral prediction only 5 off instead of 6. (2) Also, many in Missouri are going through a sort of identity crisis. Missouri has long taken pride in being a bellwether state in every election since 1904 (except 1956). I think it is kinda cool for Missouri to have that kind of status – and I’m sure plenty of Missourians who voted for John McCain are secretly hoping Sen. Obama wins the state as well, simply so they can retain their bellwether status!

Montana (3) – Predicted for Sen. Obama – went for McCain by a 3.5% margin

This was always a risky bet, and for a while I thought it was going to be a huge victory. 36% precincts reporting had Sen. Obama with a 6% lead. After 51% precincts reporting that was cut to a 1,000 vote lead. Now, with 64% precincts reporting the lead has been further cut to… oh my! He’s down by 4,300 points! This reversal is due to one fact: Metropolitan precincts were the first to report, and now all the rural precinct tallies are coming in. I think this will be reversed again as the night goes on for another simple fact: Lewis Clark county represents a significant portion of Sen. Obama’s votes and has only reported 36%. Still, Sen. Obama’s other strong counties: Missoula, & Cascade have already mostly reported all their votes. All eyes turn now to Yellowstone county. The race is in a dead heat and with only 55% precincts reporting (representing about 1/6 of Montana’s voters) a change in the race could push Sen. Obama up here as well. If it sticks with the current numbers, it will only give Sen. McCain a few hundred more votes.

UPDATE: With 72% precincts reporting, Sen. McCain now has an 8000 vote lead. Grrrrrrr… I sure hope there’s a last minute turnaround! I’d hate to be wrong on this one! (FiveThirtyEight.com actually called it for Sen. Obama earlier tonight as well).

UPDATE 2: It looks like Sen. McCain has captured this state by 16,000 votes. What a tight race. Even though Sen. Obama didn’t win, this certainly shows how poor Sen. McCain was performing. President George W. Bush won Montana with a 20% margin and didn’t even campaign here. Sen. McCain was forced to campaign here and barely pulled out a win. Incredible.

North Dakota (3) – Predicted for Sen. Obama – went for McCain by an 8% margin

As I said before: This was always my pie-in-the-sky bet. I took a gamble and I lost. I should have played it safe with those predictions and downgraded this to a “surprise state” category. I realized this later on in the week, but I had already made my predictions final! Oh well. You live, you learn.

Nebraska (1/5) (Omaha) – Predicted for Sen. McCain – currently too close to call

Nebraska and Main do not use a winner take all system. Instead, each congressional district awards an elector to the winner in those districts. The winner of the state gets the remaining two electors. This is why, anticipating a close race, Sen. Obama did some early campaigning in the congressional district occupied by Omaha (and not the whole state). Sen. McCain tried to do the same thing in Maine, but with little success. Now, it looks like this may have paid off! Although the district is currently leaning towards Sen. McCain, there is such a narrow margin – and so many provisional ballots left to be counted, it could go either way! If you read my post earlier about how demographically Democratic voters are more likely to get a provisional ballot, you’ll understand why many pundits are beginning to think this one might end up going for Sen. Obama.

******** FINAL THOUGHTS ********

So, it looks like I called Missouri & Indiana wrong… but Missouri is still SO close… It could still go either way… wow. Who would have thought? I sort of saw this scenario coming: they are each 11 electorates, so I knew I was making a “safe” swap bet… In other words, as 50/50 bets, I figured it was more likely for the candidates to each take one rather than one candidate take both.

The real risky “bets” I made were on Montana and North Dakota. It looks like I lost the bet on North Dakota – by a big margin, but I either won my bet on Montana or will have lost by just a few votes. An interesting note: most pundits were saying that McCain would lose because of the conservative support that should go to him would go to Ron Paul. With about 7,000 votes between Ron Paul and Bob Barr (the lion’s share being Rep. Paul’s) it looks like they will certainly be saying “told you so!” tomorrow morning if Sen. Obama does indeed win by a few thousand votes. Again, I don’t really agree with this crap: I’ve found that Ron Paul supporters prefer Sen. Obama over Sen. McCain – even the Ron Paul campaign has verified this – and they worked so hard to get him on the ballot – despite his requests to be removed – that I doubt any of those people would have voted for anyone else anyway…

*** UPDATED ***

McCain: Montana (3)
McCain: Missouri (11)
Obama: North Carolina (15)
Obama: Indiana (11)

365 – Obama – winner
173 – McCain

Keep in mind that this doesn’t take into account late (absentee and provisional) votes to be counted – and Missouri & Nebraska/Omaha are just SO close… I suppose it could still swing the other way. We’ll just have to wait and see.

With all the remaining uncertainty, I’m going to guess the range will end up something like this:

364-376 – Obama – winner
162-174 – McCain

So take THAT FiveThirtyEight.com! We both got Indiana wrong, I got North Dakota & Montana (barely) wrong, you and you wouldn’t call Missouri in your final prediction – but when you did, you called it for Sen. Obama! This turns out to be my same call, so we’re either going to both be right OR wrong on this one, but at least I had the balls to actually call it. Basically, I’d say we’re about even…

…except I didn’t have all those models and polling data you used to make your predictions! All I had was a couple of polls, some local paper editorials (I’ve been reading editorials in the local papers of most of the close states) and my intuition. That’s right, my intuition.

Heck yeah!

I think I see a bright future career as a campaign strategist! (*wink*) j/k

Or maybe I should think about becoming a political journalist, lol…

If you enjoyed this post, please consider leaving a comment or subscribing to the RSS feed to have future articles delivered to your feed reader.