By David Weigel

Tomorrow most of Washington will be proven completely, hilariously wrong, unless they had the sense to say nothing. A selected few will have guessed correctly, but most of the pundits who modeled and wargamed the elections across the media will have their predictions blown to bits by actual voters. When the polls close, they'll be exposed for the hacks they are.

And on Wednesday? They'll go back to work.

Political experts don't suffer a bit when they make a prediction that goes wildly wrong. Think James Carville calling Democratic victories in 2002 and 2004 and being retained by CNN. Think Fred Beltway Boys Barnes, forseeing the incredible Rick Santorum comeback behind his crossed arms. Even pollsters can brush off their bad calls; John Zogby called the 2004 elections for Kerry on election day, and was punished with a lucrative contract polling for the Wall Street Journal.

Since the pundit class doesn't stand to lose anything from the elections, online election pools -- organizations like InTrade and the Iowa Electronic Markets -- are coming back into focus as a reliable way -- maybe the most reliable way -- of judging who's going to win. Unlike pundits, traders (around 50,000 of them on InTrade and its TradeSports site alone) have to pony up when they've got an election hunch. Unlike pundits, traders suffer for their mistakes.

There's not much to the election trading sites. InTrade, launched in 2000, offers contracts on a 100-point (and $100) scale; the higher a bid (and price), the more likely bettors say that candidate's going to win the election. Take the example of a race that was wrapped up months ago. Hillary Clinton's chances of retaining her Senate seat are a 99 bid. Former Yonkers Mayor John Spencer (the New York GOP's in great shape, isn't it?) is given practically no chance of upsetting her -- he has a 0.4 bid. It's less evocative than the ways Dan Rather used to handicap elections, but it's a rock-solid call for a race no one is following.

That's why election bettors are taken seriously, even as they fly under the radar this cycle. In 2004, the conservatives at National Review watched the numbers on TradeSports to cut through the murk of October polls showing John Kerry regaining momentum in the presidential race. On the afternoon of October 15, NR economics contributor Donald Luskin let out a Janet Leigh shriek when Bush futures plummeted from 54 bid to 10. Luskin saw the gnarled hand of George Soros at work; InTrade only admitted that a prank-prone bettor had decided to sell off a record amount of futures.

But the online bettors got the rest of the 2004 campaign spot-on. When Bush and Kerry met for their first debate, pundits didn't immediately agree that the president lost. On Tradesports, bettors were dropping Bush like an arsenic ice cream cone. It soon came out that Karl Rove had known Kerry won the debate; the Bush campaign spinners were actively lying, and half the pundits were lapping it up. Advantage: Guys on the Internet.

Of course, election betting predates InTrade by more than a century. Bookies in New York's Curb Exchange (before they moved to Wall Street) were taking bids on the Benjamin Harrison-Grover Cleveland deathmatch back in 1888. The 1916 Wilson-Hughes election (the narrowest re-election of an incumbent president until Bush-Kerry) set a record for election betting that won't be matched this year: $158 million, adjusted for inflation. InTrade saw $15 million change hands in 2004, and a bit less is going to be spent this year. But the traders of 2006, in addition to being smarter than our pundits, might be the smartest class of election predictors in the sport's long history.

The current group of traders have the best information of any bettors in the history of the game; they're geographically spread out, they have access to reams of data, and they can see the media campaigns of candidates working their magic (or falling completely flat) when they flip on their TVs. They go online and see poll aggregators like Electoral-Vote.com and Election Projection, which put up each new survey from House and Senate races. And because there are so many traders betting on each race, one or two hardcore partisans can't move the numbers very far -- 2004's Sorosgate was the exception that proves the rule.

Since the two parties have gotten better at spinning their campaigns (well, at least the Republicans have), the "hivemind" effect of the trading pits has been the best guide of which breaking news and scandals actually have an effect, and which are hot air. Before the Mark Foley scandal broke, traders took stock of increased Republican optimism around the country and decent polls for most GOP candidates and priced the "Republican Party 2006 Mid Term Election Control" futures up over 80 (for Senate) and around 58 (for House). When the scandal leaked, Republicans and conservative pundits (and bloggers) did their best to explain it away, even pushing a conspiracy theory that had the notoriously fiendish, well-organized Democrats coordinating the scandal and hoping their slime wouldn't be exposed. But the news sent traders scrambling, driving futures for the House under 40 bid. After Speaker Dennis Hastert's inspiring, Saint Crispin's Day-like statements on the scandal, some pundits pronounced the scandal over. TradeSports introduced a new pool: "Democratic Pickups in the House of Representatives." Listen to talk radio today and you'll be informed that Republicans are mounting a comeback. But traders, unlike radio hosts, haven't been personally spun by the White House. And they're not changing their bets.

Online election pools aren't infallible. On Election Day 2004, when leaked network exit polling hinted at a big Kerry victory, traders rushed online to short Bush futures. But it made sense for bettors to soak up data like that -- it wasn't campaign spin or favorable polling, but surveys of people who'd just left the polling booths. The traders aren't only following the news that helps the Pelosi Party or the party of Bush. They just want to make some easy money. And that's why you can trust them.

David Weigel is an assistant editor of Reason magazine. He recently wrote about George Allen for Political Bite.