Headbashing statistics

In a recent article in New Scientist, there was a curious statistical inconsistency:

From 4000 to 3200 BC, Britons had a 1 in 14 chance of being bashed on the head, and a 1 in 50 chance of dying from their injuries...
Schulting and Wysocki have so far identified and studied the remains of about 350 skulls, mostly from southern England. The pair found healed depressed fractures in 4 to 5 per cent of the skulls, and unhealed injuries in about 2 per cent - suggesting the person died from their wounds, or at some point in the attack.

That should be approximately a 28% chance of dying from their injuries from or soon after the attack. The 1 in 50 statement implies a far lower lethality than is evidenced.