By Lindsay Toler
By Chad Garrison
By Brett Koshkin
By RFT Staff
By Lindsay Toler
By Riverfront Times
By Danny Wicentowski
By Pete Kotz
"The Obama campaign has set up the external conditions," says Quinn. "All they have to do is keep it close. In poker terminology, they have a lot more 'outs.'"
An "out" is a card that, when dealt, makes your hand a winning one. If any one of five cards gives you the win, you have five outs. For Obama and John McCain, the outs are the combinations of swing-state victories that will bring their Electoral College vote total to the magic number: 270.
"If Obama holds the states John Kerry won [in 2004] and wins Iowa and New Mexico, that's 264," Quinn says. "Now he can win any other state and be president. He has so many structural advantages."
Quinn returned to St. Louis last week to blog the vice-presidential debate for the increasingly popular and influential website FiveThirtyEight.com. Founded by Baseball Prospectus stat guru Nate Silver, FiveThirtyEight uses sophisticated statistical analysis to project the winner of the presidential race. Each day the site feeds the latest state and national polls into a computer model, which then simulates the election thousands of times. From the plethora of data these simulations generate, two key results are posted at the top of FiveThirtyEight's front page: the electoral spread and the win percentage (the percentage of the total number of simulations each candidate wins). As of the morning after the vice-presidential debate at Washington University, Obama, riding a surge in both national and swing-state polls, led the Electoral College race by a margin of 331-207 and wins more than 80 percent of the simulations.
"We see ourselves as people trying to change the perception of how politics is looked at and how we observe movement in polls," explains Quinn. "You see CNN or MSNBC — not only do they not necessarily understand polls, but they have to talk about their own polls. The public pollsters are different degrees of bad. Aggregating and putting a model over them, that's what we do."
Quinn likens FiveThirtyEight's project to baseball's new generation of sabermetric analysts, who evaluate players using complex formulas such as VORP or Nate Silver's own PECOTA rather than batting average and other traditional statistics. Though the site has its detractors, from those who object to its statistical model and/or to Silver's and Quinn's open Democratic bias — in whom they support, they emphasize, not in their computer modeling — it has become a must-read for political junkies. Silver has appeared on multiple editions of MSNBC's Countdown with Keith Olbermann and will serve as an analyst for Dan Rather on HDNet's election-night coverage.
FiveThirtyEight first attracted attention this spring. Silver, then writing under the Internet handle Poblano, made strikingly accurate predictions of Democratic primary results — most notably on the night when Obama, contrary to most polls and pundits' predictions, easily defeated Hillary Clinton in North Carolina and nearly caught her in Indiana. Poblano's work was so impressive that some fans speculated he was NBC political director and numbers guru Chuck Todd.
Quinn teamed up with Silver just before he outed himself as Poblano. The two men had in common an interest in Democratic politics, poker (Quinn knew Poblano from poker forums) and, of course, numbers.
"I was a math kid in high school," Quinn says. For the most part, though, he leaves the complex analysis to Silver: "I try to take his understanding and write things in a way my mom would understand."
Quinn also brings to the site experience in one of the election's key strategic issues, the ground game. In 2006 he worked in the field on Montana Democrat Jon Tester's successful U.S. Senate campaign. Then he went to Texas to support Democrat Ciro Rodriguez's campaign for the House of Representatives. Rodriguez, who had lost his House seat owing to redistricting a few years earlier, was now vying in a special election for a newly redrawn congressional district. Quinn points to this experience as an example of polling's limitations. An election-eve poll had Rodriguez down by five points. He ended up winning by nine.
Quinn explains that the Rodriguez campaign took advantage of early voting: "I spent over a week driving people to the polls. There were tons of canvassers. We just beat them on the ground."
Quinn is now reporting on both campaigns' ground games for FiveThirtyEight, traveling to battleground states, attending rallies and visiting local campaign offices. So far his overall impression is dominated by what he hasn't found. "I keep waiting to come across the warehouse full of thousands of McCain volunteers," he says. "I can't believe what I'm seeing. There must be a ground game going on somewhere, even if every office we go into is empty or closed." To his surprise, even in the conservative stronghold of Colorado Springs, Colorado, McCain's ground game didn't match Obama's.
What enthusiasm Quinn has found among McCain volunteers is clearly generated by the Arizona senator's running mate, Alaska governor Sarah Palin. At the Republican convention in St. Paul, those with whom Quinn spoke immediately after Palin's speech were "glowing" and "on cloud nine." And the top of the ticket? Quinn is blunt: "They are not excited about John McCain."