From here I found the data interesting
I think this chart speaks as much about aircraft design as it does about utilization, both in quantity and type of use.
For example their all-weather travelling usage and difference in quantity of utilization may explain the PA-46 and P210 data, the DA-42’s predominant training use surely plays a role in those numbers…
If you share some data on other types you are interested in I can amend the spreadsheet to include them.
I was always amazed by the safety record of the Eclipse 500 fleet. It would be interesting to know if there is any statistical difference between aircraft with a simple class rating, type rating and those with mandatory multi day recurrent training. And if there is such a difference maybe a similar training approach could be recreated for smaller aircraft.
The very good rates of the DA40 in terms of fatality may be influenced by the fact that this is mostly a flying school plane, and there is an FI on the right seat most of its flight time. This is less the case of DA42, because it is flown less and there are fewer of them.
The fatality rate remains good – and could also be influenced by it, as well as (I suppose) a good cell.
To me fatality rate looks mostly explained by speed, which is quite logical.
Slow aircrafts: DA40, DA42 (slow for a twin), C172 and 177RG have low fatality rates.
On the other hand, all 6-cylinders have high fatality rates (save maybe for the SR22 which remains average, potentially thanks to chute).
It’s not great overall. It means if you own and fly a SR22 or DA40 for 25 years (roughly the time since they’ve been in production, doubling the rate because most have been built more recently), there’s about a 6% chance it will kill you.
Damn, I need to be careful with my Mooney. Those numbers don’t look too great.
there’s about a 6% chance it will kill you
Yes, “it” will kill you. The pilot is no factor in this 
LeSving wrote:
The pilot is no factor in this
Hehe
It’s the “flying” and “owning” that kills you (which of the two is more likely, I’m not really sure )
Am I missing something here?
Let’s say we are comparing 10 C172s to 10 SR22s. Let’s say the rate of accidents in the SR22 is double that of the C172.
This only tells us that the rate was double per 10 planes. But what if none of the C172 flew one single hour?
This table doesn’t mention how many hours each model flew. That is key, no?
Fernando wrote:
This table doesn’t mention how many hours each model flew. That is key, no?
Agree. However, it’s notoriously difficult to get hard numbers for the GA fleet, especially for types that are not typically flown in schools. IMHO, statistics like the above are largely meaningless for exactly that reason.
You are right: in order to have a meaningful measure of risk we would need utilization.
I have made up a proxy for utilization (total hours flown accounted for in stats) by using years since the middle of the total production run, and that is the bottom row.
I think a lot of DA-42’s are used in a training environment with high utilization which results in the above numbers. Actually Mooneys are quite safe if you account for that. You just don’t want to crash a Mooney because fatality is high, better to crash a Diamond. Passive safety in Diamonds is second to none so this results in the fatality rate being the best.
I would not say this data is meaningless, I would say it is to be used judiciously. For example, if you compare a C172 vs a PA28, you can fathom utilization is going to be similar. Same goes for TB20 vs Mooney and perhaps P210 vs PA-46
As to the six-cylinders …well, you have to factor that the number of fatalities is what drives the fatality rate. If an aircraft has more average POB then for equivalent level of safety fatality is going to be higher. Hence fatality corrected for seats count (this assumes there will be 50% more POB on average on a 6-seater than on a 4-seater, possibly 3 avg POB vs 2 avg POB, yet another proxy). When you factor seat count you actually see the “six-cylinder” aircraft are quite reasonable in fatality. They just have a high accident rate, possibly connected to the all-weather travel mission.
Eclipse 500 data added
