COLUMBUS – As traffic begins to increase for the Thanksgiving holiday travel weekend, Ohio State Highway Patrol troopers are closer to predicting where serious injury and fatal crashes occur before they happen thanks to a probability forecasting model developed by The Ohio State University’s Statistical Consulting Service (SCS).

The forecasting model, believed to be the first of its kind in the nation, provides quantitative predictions of the likelihood of future crashes. Additionally, the model provides some insight into the causes of crashes and will be used to enhance Patrol dispatching and crash prevention strategies.

Built on historical crash data, the forecasting model quantifies the likelihood of where, when and why a serious or fatal crash will occur. The predictions produced by this model will be used along with Geomapping statistics to dispatch troopers to likely crash locations at the appropriate times.

In addition, the model provides insight into key crash-causing factors leading to a higher likelihood of crashes. Information from the model will be used to reduce the number of serious and fatal crashes on Ohio roadways.

The SCS in the Department of Statistics at Ohio State provides professional statistical research and consulting services to the university's student and faculty community, as well as clients outside the university. This project with the Patrol utilized five years of crash data to develop models that may keep the people of Ohio safer on the roadways.

This model was first used by the Patrol in June of this year to predict the likelihood of crashes during the Fourth of July holiday. The Thanksgiving holiday models are available on the Patrol’s Web site at www.statepatrol.ohio.gov/osu.htm.

# # #

Note to editors: Satellite images of all 2006 Ohio fatal crash sites are available through Google Earth on the Patrol’s Web site at www.statepatrol.ohio.gov/google.htm. Fatal crash map from 2006 is available on the Patrol’s Web site at www.statepatrol.ohio.gov/fatals/statemap_06.htm.

06-228