If you check the roller coaster boards and statistics, the vast majority of people that get hurt, do so because they weren't following direction. They break the rules, ignore the size requirements, unstrap themselves, etc.
Here's an example from Markey's site:
http://www.house.gov/markey/iss_parkrides_97safety.pdfIt states that roller coasters are surpassed by only automobiles on the likelihood of injury. If you look at the details, it states that 3 people in 1997 were killed.
Just one year! Hardly a statistically significant sample size. In addition, it compares fatalities to miles traveled as if distance has any meaning whatsoever on a coaster.
