Updated

The night of the Iowa caucuses Karl Rove and I got entrance poll information at the same time as we sat on the Fox News Election coverage set in New York. Entrance polls are for caucuses – exit polls are what they do in primaries.

Within a few minutes of the first caucus results we were both busy trying to figure out who was going to win the Iowa event that would start someone on their way to the GOP nomination. Karl and I had not been on air much together.

Truth be told I can’t think of a single thing we had ever agreed on. I had vehemently opposed the war in Iraq – he had been a senior advisor to the president that led us into it. I ran the Dean campaign in 2004 to end the war and lost – he had run the reelection campaign for George Bush and won. I had railed against the Bush preemptive doctrine – for all I knew Rove might as well have written it. To say I thought I was sitting on the set with Darth Vader would be an understatement. I am not sure he felt much kinder things about me.

Karl was turned slightly away from me going over precinct results for completed caucuses and scratching out the math by hand on multiple sheets of paper – I was turned slightly the other way doing the same. Within 5 minutes of receiving the entrance poll numbers and the first early precinct results we turned to each other to compare notes. I said “I can’t believe it but Rick Santorum is going to eke past Romney and win this thing.” Karl looked down at his sheets of equations and said he couldn’t believe it either – that Santorum was going to give Romney a dogfight no one had really expected – but he had Romney with a slight mathematical advantage. The two of us had calculated the race in minutes to within a tenth of a percent of each other.

We would work through the night – Karl would identify the two precincts out of all the precincts in all the 99 counties that night that would decide the outcome.

An amazing feat.

Chris Wallace started to call us the "Space Cowboys" on the air and held up reams of our calculations to show the audience why he had given us the name.

Through it all our calculations would swing slightly back and forth. At one point my math showed that Karl was right and Romney would win. Minutes later Rove would run some new precinct results and come to the conclusion that I had been right and Santorum would win. Later we would realize that we were both right. Iowa in the end would be decided by fewer than a 100 votes -- both of us had been arguing narrowly on each side of a 100 vote margin of error. I learned then that no matter how different Karl Rove and I were or how wide our differences on the war, tax breaks for the wealthy or superpacs (I hated them) we had a very similar way we looked at numbers and calculated election results -- so similar we often came up with the same numbers. It was eerie how close we mirrored each other’s projections and we didn’t BS each other or spin.

As the primaries moved forward it got to the point where we ran through the counties together. One of us calling out a county we wanted to look at – the other calling out results – then one of us pounding the numbers into a calculator or computer and offering an opinion of what they meant – “that’s not good for Newt”, “Romney is doing better with catholics” – in state after state we got it right and usually ahead of when we could say it on the air. We are not the Fox News Decision Desk and we can’t call a state on the air before they call it – but often we would have if we could. The Decision Desk had to be certain – had to be careful – The Space Cowboys? We often muttered into our mics, “why haven’t you guys called this?” Frustrated that we knew what was going to happen but knowing they were right to wait.

Fast forward.

Sometime before the conventions I was doing some focus groups in Ohio when I saw the dam break against Mitt Romney among undecided voters. They were undecided but when you asked them why they were not voting for Romney they would volunteer that they were worried about Bain Capital or didn’t like him hiding his tax returns, or that Romney wanted the auto industry to go bankrupt – fair or not – the unanswered negatives on Romney had taken hold – and I knew in my gut Ohio was over – Romney would never win it.

A few days later I was in Chicago. David Axelrod was in Iowa traveling with the President but I got an email from him saying he would be back the next day and if I stayed we could get together. When I mentioned to Ax that I thought Obama was going to win Ohio and that Romney was a terrible candidate for the Midwest, he told me he had always thought the same thing and that their polling for weeks had had the President ahead by 4 to 5 points always outside the margin of error in Ohio. I knew they were building a massive early vote and get out the vote effort in Ohio and other key states. Armed with my focus group experience, Axelrod’s comments reinforced my own confidence that Ohio was moving solidly in Obama’s direction.

By now Karl Rove and I were doing an electoral map projection every other Monday or so with Bret Baier who anchored Special Report on Fox News. We tended to agree on what the polling was saying about most states – but Ohio was among those we rarely agreed. I almost always thought it should be lean Obama – Karl almost always thought it should be a toss up. To be fair though there were weeks when Karl made it a lean Obama state on his map. But at some point it dawned on me that not once had there been a single week where anyone made a case for making Ohio lean Romney.

Then one day at the Democratic Convention in Charlotte, North Carolina the Obama team invited the Fox News contingent in for a briefing on what to expect from the convention line-up. When we walked into the briefing Jim Messina, the President’s campaign manager, and I started to talk as others found their chairs and he whispered to me that he had seen two new internal Democratic polls in Ohio that showed 9 point leads for Obama in the state and that the President’s lead had expanded since I had seen Axelrod a few weeks earlier.

I was certain whoever won Ohio was going to be the next President of the United States and I had now become pretty sure Ohio and the Presidency would be won by Barack Obama.

For four years the Obama campaign had been working on their Ohio firewall and it was becoming clear it was going to hold. Even after the first debate when polls everywhere showed the race tighten Ohio proved stubborn for Romney. Things got closer in Ohio too – but Obama always held his lead there.

Back in April Karl Rove had outlined the 3-2-1 path of states Romney needed to put together to win 270 electoral votes and the presidency. Romney had to win all the states John McCain had won in 2008 against Obama. Then he had to win three states the GOP had traditionally won that Obama had taken from them in 2008 – Indiana, North Carolina and Virginia. If he won those three Romney had to win both Ohio and Florida and if he won those two he had to win one out of the other 5 or so remaining swing states. The Obama campaign had set out to block that path by mounting an all out campaign in all the 3-2-1 states except for Indiana. Winning just one of the critical 3-2 states would make it not impossible but very tough for Romney to get to 270 electoral votes – winning Ohio would be a huge stop for Obama.

Every Monday on Special Report I would argue that Romney was not going to win Ohio. But other after the first debate other paths to 270 for Romney began to open up. Karl would argue that Romney could win Colorado and Wisconsin or another combination of swing states that included Wisconsin and that would make up for an Ohio loss (plan B). And the Romney campaign late in the game started talking about taking Pennsylvania (plan C).

The thing that was so different about this election was that no matter how confident Axelrod and Messina were about all the swing states or I was about Ohio – Karl Rove and Rich Beeson or whoever you talked to on the GOP side was just as confident about Ohio or Virginia or other key swing states for Romney. Both sides were convinced their model was right. About three weeks out from election day it hit me whoever was right was going to win over 300 electoral votes. The race was too close – if Romney’s model was right he wasn’t tied in all those states he was slightly ahead in them and would win them all. But I was sure Obama’s model was right. I knew something about the new metrics from the Dean campaign I managed in 2004 – politics had changed – undecided voters don’t break automatically to the challenger anymore, targeting is better, social networks have changed organizing and information flow and the demographics of the nation were moving away from the Republican party. If the Obama model was right he would sweep the swing states. Someone was wrong and I was pretty sure it was the guys in the Grand Old Party.

On the Monday before the election when Bret Baier asked me for my final electoral map showing what I thought would happen the next day I projected 303 electoral votes for Obama – projecting every swing state to Obama except for North Carolina and Florida. Saying “if Obama’s model is right I think he will take Florida too”. I am still kicking myself for being such a chicken – I knew they would all go one way but I had chickened out on Florida. Axelrod had even sent me an email urging me to take a flyer on Florida – he thought they were going to surprise there – but I had chickened out of calling it. No excuses. Can’t explain it . Would anyone believe 332?

Election Night

So what happened on election night? Where is all of this going?

On election night for some reason neither Rove nor I can get into the Ohio Secretary of State’s website to get any county data – or any data for that matter. We both tore into the Florida and Virginia election data the second their polls close.

Karl started calling out GOP leaning counties in Florida and I started to call out the percentages Obama and Romney were getting in them. Rove quickly pointed out that Romney was performing two percentage points ahead of the 2008 McCain percentage in these counties. I called out Palm Beach, Broward and Miami-Dade and the Obama and Romney percentages in those huge Democratic leaning counties – and early on Obama was performing 2.8 points ahead of where he did in 2008. Obama was winning Hillsborough county by 6 points! It was going to be close – really close. It's was almost like that first night in Iowa we had together at the beginning of the whole thing – Karl convinced Romney would pull it out by a few votes – and I was convinced Obama had Florida. Florida! What a chicken – I started kicking myself some more.

We started to look at Virginia. Same thing -- it’s a dead heat everywhere and everything is coming in except for Fairfax county and Virginia Beach. Virginia Beach is leaning Romney hard and Karl thinks there might be enough there for Romney to pull it out. He has to pull it out – it’s a 3-2-1 state – hell they are all 3-2-1 states. Fairfax county is moving hard for Obama and I think there are more votes still out there than there are in Virginia Beach. I think maybe just maybe Obama can win Virginia too and start to run up his electoral victory.

The Fox Decision Desk calls Pennsylvania for Obama – Romney’s plan C is gone.

A little later plan B would fall too as they called Wisconsin for Obama.

Plan A is all that is left for Romney – he has to win Ohio, Virginia and Florida. All are 3-2 states and when one goes so too goes the Presidency. Win them all or Romney never gets to Colorado or any of the 1 states.

Chris Wallace off air is begging us to get into the Ohio Secretary of State website but Karl and I get the same error messages we have gotten all night long.

At that moment two emails come in at the exact same time.

Chris Wallace received an email that he would read on the air from someone high up in the Romney campaign – that email says that Romney will win Florida by 10,000 votes. Virginia is tight and Ohio is moving their way. At that same moment I receive an email report from an Obama contact insisting that Obama will win Florida and with it the Presidency, Virginia is close but they will win it – but (and this is a direct quote) “Ohio is scary”. Chris reads his email to Karl and me off air – I read my email to the two of them. And then I turn to Chris Wallace and said – “wow wouldn’t it be the ironies of ironies if we ended up calling Ohio for Romney and everyone thought he won the presidency and then it turned out that Obama won Florida and won reelection that way instead?”

Karl and I scrambled to see if we could get into the Ohio Secretary of State’s website again. Karl yelled out that he had gotten in. And just at that moment we hear in our ear Bret Baier declare that the Fox News Decision Desk has called the state of Ohio for Barack Obama.

We were separated from the main anchor desk that Bret Baier and Megyn Kelly hosted the election coverage from by a good 30 feet or so – but it felt like miles.

I blurted out to Chris Wallace and Karl Rove, “Do they understand the implications of that call?” I felt calling Ohio was tantamount to calling the presidency and thought we would be seeing blazing graphics declaring Obama reelected. Just as I asked the question I realized immediately why we were not seeing it – I knew Florida was gone to Obama – the Decision Desk may have been pretty sure it was going that way but from their perch Obama was not over 270 electoral votes yet – neither Virginia nor Florida had been called. Chris looked at me like I was on Mars and said “Of course they understand what this means – but we have to wait for those other states to call it”.

Just then Karl Rove’s phone rang it was someone in the Romney campaign….you know the rest.