2 deaths per hundred thousand hours is really a lot.
So if you're a hobbyist, and you fly a light sport aircraft six hours a week for thirty years, you've got a 20% chance of it killing you. (Right? I haven't had coffee yet.)
So if you're a hobbyist, and you fly a light sport aircraft six hours a week for thirty years, you've got a 20% chance of it killing you. (Right? I haven't had coffee yet.)
Several issues though:
The risk is far from linear. Infrequent pilots have much higher risk per hour.
The number of hours for part 91 (most private flying) flights cannot be reliably determined.
About 75% to 95% of risk is under the control or strong influence of the pilot. Weather, fuel, crew readiness, pre-flight planning and related issues are much more under pilot control (if you're willing to cancel or delay flights) than in comparable car trips. It's extremely rare for another pilot to kill you.
I came up with 9% if you assume two fatals per accident.
Light sport is restricted to day VFR, which is lower risk, other things being equal. Unfortunately, that also saps some utility and pushes you towards taking a late day flight rather than waiting for better night weather. I don't think LSA is viable for transportation, which makes it all the more unusual for someone to fly 300 hours per year.
(Context: I fly about 150-200 hours per year, which is 80+% of our vacation travel. I also read every NTSB fatal report.)
So if you remove the "accident waiting to happen" category of pilots, the fatal accident rate will likely be much lower. I've heard it said that GA has a similar accident rate to riding motorbikes, which is another activity where the risk mostly depends on the rider.
The type of pilot who is most dangerous is the type-A business owner who feels they have to get to their meeting no matter what the weather. Unfortunately these are the people who will have $250k in their pocket and want to use a plane for transportation. Perhaps the only solution is to have self-flying planes, similar to google's cars. (And of course, they would need to be hacker-proof!)
Just this prior weekend, we had to be in Columbus, OH for my inlaw grandparents' 70th wedding anniversary and then I had to be back to head out to Re-invent in Vegas. With the hurricane possibility, we cancelled our plans to use our airplane and booked 3 day advance purchase trips on one carrier to Cleveland (to avoid connections on the east coast), then drive to Columbus, then fly out of Columbus on a different carrier.
Was expensive on its face (4 people, late notice direct flights), but cheaper than risking killing my whole family tangling with weather. Of course, the bad weather didn't end up materializing, but you have to be willing to make those calls frequently to use general aviation for travel to take the "get-there-itis" pressure off.
If you're going to sweat the money for commercial tix/hotels/cabs or are not willing to sit in the FBO weather room and pilots' lounge if you're caught in weather in the middle of a trip, your risk is going to be way higher than it needs to be.
The way I usually explain it is to say that in a car you are much more at the mercy of other drivers. A huge percentage of people killed driving are killed by drivers of other cars. In an airplane nearly all accidents involve only one aircraft. With that, and the fact that so many accidents are pilot indiced the stats for a well trained conservative pilot are far far better than "average." Good pilots constantly re-train and do things like read NTSB reports to minimize this risk as far as possible.
There's still a "crap happens" factor (eg catastrophic mechanical failure) that can take down even the best of pilots but nothing in life is completely devoid of "crap happens" risk.