A team of mathematicians and computer scientists at the Los Alamos National Laboratory in New Mexico say it’s probable that this season’s flu cases will spike in February. That’s in contrast to the past three years, when the flu peaked in mid-to-late-December.

Scientist Sara Del Valle, who leads a group tracking flu cases for the federal government, said her team looks at two types of data to make their predictions: the number of ‘flu-like’ virus symptoms being reported at doctor’s offices, and actual lab samples of the flu. (Both are collected by the Centers for Disease Control and Prevention.) Putting the two pieces together, Del Valle’s team can get a more accurate picture of how the flu is spreading around the country.

Now the team’s system is getting faster and smarter with the help of a third, and more unusual, data source: Wikipedia searches. Del Valle’s found that as the number of flu cases increase, so do searches of about a dozen ‘influenza’-related Wikipedia pages. And, because people search online before they ever step foot in their doctor’s office, the team thinks the addition will help them track flu patterns more quickly. Wikipedia search data is available hour by hour, as opposed to the two weeks it can take to get lab test results back from the CDC.

Current models are also predicting this will be a mild flu year. But that doesn’t mean it’s time to start licking your co-worker’s coffee cup just yet. “Just because we’re predicting a mild season doesn’t mean that people shouldn’t get vaccinated,” Del Valle says. “Because you never know if you might be the one to get the flu.”

The team plans to update their flu forecasts every two weeks this season, and hope to have the Wikipedia data integrated into their predictions within the month. Until then, just keep your distance, please.

Make sure to subscribe to Data Sheet, Fortune’s daily newsletter on the business of technology.

For more on America’s healthcare system, check out the following Fortune video: