I agree that most pax are not interested in safety stats (the ones that read PPRuNe are probably not a representative sample
), but even for those of us who are interested, it's very difficult to draw meaningful comparisons.
This is because air accidents are very rare - in large part due to enormous efforts from all those involved in aviation to make it (and keep it) so safe. Because they are rare events, you need to go back a very long way to gather enough data for a statistically valid analysis, probably more than 25 years. The trouble is that over the course of 25 years, everything changes. The aircraft fleets, the pilots flying them and the safety systems associated with them are all very different.
But what we want to know is: will my flight next week be safer if I travel with operator A or operator B. The 'fact' that operator A has a better record over 25 years, doesn't necessarily tell us about how much safer their operation is today.
In an earlier post, I gave an example of choosing to fly Lufthansa as opposed to Garuda. I'm too lazy to analyse the stats, but I wouldn't be surprised if they suggested that Lufthansa is the safer choice. But I've no way of knowing whether that's because they have more modern aircraft, better or better trained pilots - or if it's because Lufthansa operate the majority of their flights in Europe to airports equipped with modern navigational aids and in relatively benign weather. If it's the potential for tropical storms that is the problem, then the 'dangerous' choice I've made is to travel to Jakarta, but I may not have much control over that!