Milliman Inc., a Seattle-based consulting and actuarial firm, announced that it has created a driving “risk score,” with the help of tech startup Zendrive.

The risk score is up to six times more powerful than the leading predictive models, according to Milliman.

Milliman teamed up with Zendrive, a smartphone-powered driving analytics company, to study how distracted driving and other driving behaviors can lead to auto collisions. Using Zendrive data, Milliman verified the behaviors that were strong indicators of collision frequency and created a risk score to compare the “worst” drivers relative to the “best.”

Their findings revealed that the worst 10 percent of drivers were over 13 times more likely to be involved in a crash than the best 10 percent of drivers.

The results were based on one of the largest telematics data sets in the United States, according to Milliman. As of today, Zendrive has captured more than 40 billion miles of driving behavior via smartphone sensors.

“Smartphones can measure driving behaviors that traditional, first-generation telematics can’t, such as who is driving the vehicle and phone usage contributing to distracted driving,” said Sheri Scott, a principal at Milliman and co-author of the study. “These new-age predictors contributed to a risk score that is over six times more accurate than the current industry leader models, which use traditional hardware-based telematics devices. There’s an opportunity here for auto insurers, especially commercial auto fleet insurers, to be early-adopters of this technology, and improve their ability to measure and rate risk.”

To read more about the study, go to http://www.milliman.com/driving-risk-score/.

Editor’s note: This story originally appeared in TelematicsTalk, a sister publication of BodyShop Business.