Net firm makes anyone a prophet
Web application weighs everyone's predictions on baseball scores, Oscar wins and political races
DIFFERENT KIND OF FUTURES MARKET
By Eric Benderoff
Tribune staff reporter
February 24, 2007
If you believe Aramis Ramirez will hit more home runs this year than Paul Konerko, or perhaps that the Cubs will win more games than the White Sox, you can do more than make a bar bet.
Instead, create your own prediction market where you, your buddies and anyone else who fancies themselves a fan can speculate on how the coming baseball season will play out.
Setting up a market is old hat in the Windy City, the capital of the nation's futures industry. But in a twist, a Chicago Internet firm is offering everyday Web users and corporate giants a chance to predict outcomes.
It's an adaptation of the idea that a group of people should know more than an individual, and its use is growing. Companies, for example, have latched onto the concept because a boss who wants to know whether a project will finish on time may be more likely to get an honest answer by asking all the participants instead of relying on the word of a project manager.
"Running these markets can introduce a level of transparency that didn't exist at company before," said Adam Siegel, who with partner Nathan Kontny began Chicago-based Inkling in late 2005. "The messaging from the rank-and-file workers is often sugarcoated before it gets to the boss."
Inkling's clients include Abbott Labs and a San Francisco TV station, but the firm sees opportunities for individuals. It offers a simple Web application that allows anyone from a LaSalle Street trader to a zealous sport fan to set up a prediction market on a blog or Web site.
The firm also invites people to play along at InklingMarkets.com. Current markets include which film will win the best-picture Oscar on Sunday night, who will win the Republican nomination for the 2008 presidential election, if Barry Bonds will break Hank Aaron's home run record and if Google's stock price will hit $600 by year's end.
"People are creating markets each day," said Siegel. "Several thousand users are playing, trading for the fun of it."
Registration at the site is free, and new users receive $5,000 in imaginary money when they sign up. You can participate in an ongoing market or create a new one, which takes "about three minutes," Siegel said.
Prediction markets essentially work based on the wisdom of crowds, said Eric Zitzewitz, an assistant professor of economics at the Stanford graduate school of business. "You ask a bunch of people their opinions on something and it's aggregated. Many opinions are better than one."
In theory, those opinions should more accurately predict the outcome.
Consider one market that attempted to predict when Sen. Barack Obama of Illinois would announce his run for the presidency.
A time frame was established, giving dates in January, February March or that he wouldn't run. The consensus accurately predicted that he would announce between Feb. 1-14. He officially announced on Feb. 11.
One of the markets that closed recently accurately predicted that the Indianapolis Colts would win the Super Bowl. The market included every National Football League team since the market launched when the season started, but as the playoffs wound down, and as the two teams for the big game were determined, roughly 65 percent of the trades favored the Colts.
In the real world, the reason people use markets is to make money. If an outcome is accurately predicted, it is profitable. In virtual markets, "play money is not a bad predictor, but economists say when real money is at stake, it's a better indicator," Zitzewitz said.
Jimm Dispensa, a Chicago political junkie, is using Inkling's prediction tool on his blog to try to predict the outcome of Tuesday's aldermanic elections.
"I went looking for something in October," he said, "something that would specifically allow users to create and manage their own markets. But I wasn't about to build one on my own."
Dispensa wanted to emulate the University of Iowa's election market, which since 1988 has more accurately predicted the outcome of the presidential election than traditional polling methods.
He found the Inkling tool during a Web search. The software, known as a widget, easily attaches to a user's blog or Web site, much like adding a video from YouTube.com onto a blog by simply cutting and pasting the code.
At Dispensa's blog (http://aldertrack.typepad.com), you can use the prediction tool for just one of Chicago's 50 wards or 40 if you choose, since 10 aldermen are running unopposed.
So far, about 1,200 users have participated, but Dispensa suspects only 700 are unique users. Rather, many people are entering under multiple e-mail addresses in an effort to "game" the market, he said.
"Theoretically, every one can `game' the markets equally," Dispensa said, "but next time I'll monitor it more closely. It was an important lesson."
Hence, he said many of his results won't accurately reflect Tuesday's election. "They will more accurately reflect the candidate's fan base."
For Inkling, which provides the widget for free, the path to profitability is selling the prediction tool to corporations.
One client is North Chicago's Abbott Laboratories, which currently has a 12-month contract with Inkling.
Abbott declined to discuss how it is using prediction markets, but Zitzewitz said pharmaceutical companies are using prediction tools for sales forecasting and product development.
"Suppose you have a bunch of scientists who have some ideas on what chemical compounds might be effective in a new drug," he said. "But you don't know how a new drug would fare in clinical trials."
Hence, the company could create a market and pose the possibilities to a swath of people across the organization. "You could run a market on the potential for FDA approval" for instance, Zitzewitz said.
Web search giant Google is also experimenting with prediction markets, though it is not an Inkling client. In an official Google blog post, Bo Cowgill, a product manager, wrote that the company set up an internal predictive market.
"The markets were designed to forecast product launch dates, new office openings, and many other things of strategic importance to Google," he wrote. "Googlers from across the company contribute knowledge and opinions which are aggregated into a forecast by the market. Sometimes, just feeling lucky isn't enough, and these tools can help."
Perhaps the most interesting, and telling, market at Inkling concerns the company's future.
The question: "What is the probability each of these yCombinator start-ups will get bought out in the next two years?" (YCombinator is the venture firm backing Inkling, as well as the other 11 companies listed.)
Trading in two firms, Kiko and Reddit, closed because they have been purchased. In October, Conde Nast, which publishes Wired, purchased Reddit for an undisclosed price. And earlier in 2006, Kiko, which was struggling, was purchased on eBay for $258,100 by a Canadian technology firm.
Of the nine remaining firms left in the market, Inkling is listed as having second-highest probability of being purchased.
----------
ebenderoff@tribune.com
Copyright © 2007, Chicago Tribune
chicagotribune.com |
