Illustration by Michelle Barbera
Perhaps you’ve used the forecasting feature on the Bing search engine: After entering your flight info, Bing’s price predictor calculates if it’s the best time to buy a ticket or if you should wait for fares to drop.
Using this type of predictive analysis, all kinds of Washington state businesses are taking a cue from Bing and adding forecasting features of their own. Here are a few upcoming programs to help navigate your future:
UrbanSpoon is adding a Food Forecast feature to its informative restaurant reviews. Factoring in the catch of the day, the distance ingredients have traveled, chef’s sick days, antibacterial soap orders and historic Department of Public Health inspection grades, the Food Forecaster will give customers the odds of having a great meal or getting food poisoning. Bon appétit.
Amazon is not only using predictive analysis to suggest products for customers (“If you like Shakespeare, you might also like Sarah Palin’s autobiography”), but also indicating the likelihood you will spend more or less than your entire paycheck on these additional items. As you get more comfortable with Amazon’s customized predictive searches, you may sign up for NoClick, an automated program that constantly finds items you will like and immediately sends them by FedEx.
Microsoft recently applied for a patent for an image-based dating search, allowing users to enter the physical characteristics they want in a mate, and let the web work its magic. Not to be outdone, Match.com is adding several predictive elements to its dating website. A few hours into a date, members enter vital stats, including food and alcohol consumed and the number of times ex-lovers were mentioned, to get the odds on when/if the new couple will consummate the relationship, chances of STDs and potential future restraining orders.
Match’s Marriage Forecaster predicts with 89 percent accuracy whether you should marry the guy or gal based on an analysis of cell phone pics of the interior of his or her car. The ever-exact Divorce Forecaster is no longer being featured on the site.
LawnMow is a great new app that uses predictive analysis to give individuals the best chance of mowing the lawn without rain, snow, lightning or excessive heat. Customers simply log into their LawnMow account, prop open a window with their hard drive and look up into the sky.
Starbucks’ new predictor combines traffic flow with customer-preference data. The Barista Barometer shows the wait time at your favorite Starbucks location, and contrasts it with the number of minutes it would take to cross the street to the other Starbucks location. (The program also alerts you which jazz or folk classic you’ll have to endure while in line.)
Future versions of Microsoft’s Farecast will not only tell travelers the best time to buy airline tickets, but will also access medical records to help them select a seat where they will be less likely to be placed next to a compulsive talker, a person with the flu or a conspiracy theorist. The Terrorist-Estimator is still working out some kinks (namely, the profiler program).
ValueAppeal uses data mining to let you know if your property-tax assessment is too high, and displays comparable property values in the ’hood. New graphing tools also evaluate if your home is financially underwater, the chance you’ll go into foreclosure in the next 30 days and the best paint colors for resale.
Even the Seattle Public School District is getting into the act of forecasting. Based on a variety of aggregated test scores as well as an analysis of a child’s early stick figures, the SPS Braincaster lets parents know if it’s worth paying the big bucks to place the child in private school.
Finally, the Washington State Lotto has gotten into the predictive analysis game with its handy-dandy Billionaire-by-Tomorrow forecaster. The program shows the chance of winning your favorite lotto game, as well as the numbers that are most likely to come up for SuperLotto. The most helpful predictor, “You are less likely to win if you don’t play,” has not actually been statistically proven.