OK, I'm officially suspicious. The latest deer-vehicle collision probability statistics, calculated by State Farm Insurance, came out Friday.

According to this year's projections, West Virginia's motorists have the nation's highest chance of colliding with Bambi or one of his cousins.

At first glance, the numbers make sense. West Virginia has a relatively high whitetail population, a dense network of roads, and forests that in many places extend right to road's edge.

The incidence of deer-vehicle collisions is high, too. In 2006, the last year for which complete data are available, nearly 19,000 deer were struck and killed along the state's highways.

Those statistics, compiled jointly by workers at the Division of Highways and the Division of Natural Resources, appear to be a cornerstone for State Farm's calculations. I say "appear to be" because the percentages match up and the numbers don't.

Here's what I mean. In last year's news release, State Farm officials placed West Virginians' chance of being in a deer-vehicle crash at 1 in 57. In this year's release, the chances leapt 19 percent, to 1 in 46.

Since the number of motor vehicle registrations stayed more or less the same at 1,441,099, it's easy to see where State Farm's actuaries arrived at their projected probability increase. They figured that if the number of road-killed deer increased 19 percent, the number of collisions probably increased by the same percentage. So they set their projected probability of collision 19 percent higher. Fair enough.

What caused my antenna to go up was the raw number of deer-vehicle collisions State Farm's computer wizards projected for this fall - 31,967. A 19-percent increase in road-killed deer would account for 22,000 accidents. State Farm's estimate is a full 45 percent higher.

According to this year's projections, West Virginia's motorists have the nation's highest chance of colliding with Bambi or one of his cousins.

At first glance, the numbers make sense. West Virginia has a relatively high whitetail population, a dense network of roads, and forests that in many places extend right to road's edge.

The incidence of deer-vehicle collisions is high, too. In 2006, the last year for which complete data are available, nearly 19,000 deer were struck and killed along the state's highways.

Those statistics, compiled jointly by workers at the Division of Highways and the Division of Natural Resources, appear to be a cornerstone for State Farm's calculations. I say "appear to be" because the percentages match up and the numbers don't.

Here's what I mean. In last year's news release, State Farm officials placed West Virginians' chance of being in a deer-vehicle crash at 1 in 57. In this year's release, the chances leapt 19 percent, to 1 in 46.

Since the number of motor vehicle registrations stayed more or less the same at 1,441,099, it's easy to see where State Farm's actuaries arrived at their projected probability increase. They figured that if the number of road-killed deer increased 19 percent, the number of collisions probably increased by the same percentage. So they set their projected probability of collision 19 percent higher. Fair enough.

What caused my antenna to go up was the raw number of deer-vehicle collisions State Farm's computer wizards projected for this fall - 31,967. A 19-percent increase in road-killed deer would account for 22,000 accidents. State Farm's estimate is a full 45 percent higher.

Granted, not all deer-vehicle collisions end up with deer corpses left beside the road. In many collisions, the deer survives. In others, it escapes out of sight but dies from its injuries.

But does that happen 45 percent of the time? I'm skeptical.

For one thing, some of the low-speed collisions that allow deer to escape relatively unharmed wouldn't cause enough damage to vehicles to warrant insurance claims. For another, not all motorists who incur deer damage claim it on their insurance.

That said, I suspect that more West Virginians claim deer-vehicle collisions than are actually involved in them. "I hit a deer" sounds a whole lot more honorable than "I was drunk as a skunk and clipped a car coming out of the beer joint's parking lot." People don't get dropped by insurance companies for deer-vehicle collisions.

Finally, because they appear to be based primarily on increases or decreases in numbers of road-killed deer, State Farm's statistics don't account for changes in some of the conditions that lead to deer-vehicle collisions.

When acorns and other high-value deer foods are abundant, deer tend to stay in the woods and away from the highways. Hickory and beech are abundant this year. Acorns are less so. Will the abundance of hickory and beech reduce the number of deer-vehicle collisions, or will the relative scarcity of oak increase it?

Until insurance folks start taking such biological factors into account, I'll remain skeptical of their deer-vehicle projections.