Each of us was born with-or we develop-our own herd of personal demons. Some pilots sweat bullets at the mere thought of actual IMC while others fear the bolt-from-the-blue mechanical failure. Thats another way of saying we tend to inflate the risk of things that scare us while discounting things that don’t, thus we sometimes protect ourselves expensively against small risks while ignoring larger ones.

The only rational way of judging risk in aviation (or anything else) is to look at hard numbers-overall accident totals and accident rates. Some categories loom large-landing and takeoff accidents, VFR into IMC, fuel exhaustion-while others rank low enough to be all but flukes, such as pilot incapacitation or even mid-air collisions. The best snapshot view of the accident big picture is found in AOPA Air Safety Foundations Nall Report, which you can download from AOPAs Web site. The Nall report is a no-nonsense summary of accident causes thats as good a source as any of data by which to make risk assessments. Although its conclusions apply to the pilot world at large, the Nall report does beg this question for anyone who reads it: How do I fit in? Am I typical of the pilot population at large or do I have more or less risk in some areas of flight operations?

Were utterly unable to answer this, but our view is that the pilot population is probably more homogenous than most of us would admit. In other words, if you think youre too smart to be susceptible to a fuel exhaustion accident-a big percentage of total accidents-you may be kidding yourself.Herewith are our views on the best safety enhancing investments.

The Nall report grimly notes that 75 percent of all GA accidents are due to pilot error and more than 80 percent of fatals are due to pilot mistakes or misjudgments. These are lack-of-proficiency accidents that can, one way or another, be traced to lack of or poor training or anemic proficiency due to limited flight activity. Either way, the results are the same.

In our view, ongoing training addresses this shortcoming across the board, improving flight skills from basic stick and rudder work to instrument scan to radio skills, all of which contribute to a less accident-prone pilot. Typically, an instrument rating costs $5000 to $7500, depending on the aircraft used and how much sim time is included. But in our view, no other single investment impacts proficiency and safety to the degree that adding an instrument rating does.

Or a real instrument proficiency check on a regular basis. Cost: under $500, including the airplane time. Were talking here about a challenging, skill-enhancing flight review or IPC from an instructor who gets the picture, not a pencil-whipped love fest that barely warms the engine oil. Do it on a gusty day with a crosswind and you may avoid being among the several hundred pilots a year who suffer runway loss-of-control embarrassments.

See numer 1. The accident record suggests that you cant be over-trained. Typically, a commercial rating improves stick-and-rudder skills for the sum total of about $1500, including the airplane time. If you already have an instrument rating and a commercial certificate, adding an ATP for under $1000 is chump change and sharpens instrument skills.

Of all the stupid pilot tricks that pilots pull, one of the dumbest is not just running out of fuel but, as a group, doing it about twice a week, on average. This accounted for more than 100 accidents in 2005, according to the Nall numbers. We think there are multiple reasons for this, but our theory is that two stand out: Pilots don’t really know how much fuel theyre burning and they rely too heavily on inaccurate panel gauges.

The Nall study finds that weather is a factor in 4.6 percent of all of the pilot-caused accidents, but 13.6 percent of the fatals. That amounts to about 50 accidents a year, 33 of them fatal.

Its not that weather reached out and smote the airplane; the pilot simply blundered into weather he couldnt handle. The overwhelming majority of weather-related accidents are VFR-into-IMC poor judgment calls but-and this significant- 9.1 percent are thunderstorm related. For the past five years, weve had the technology to avoid the thunderstorm threat strategically and tactically in inexpensive datalinked weather. Our first choice is Garmins GPSmap 396, which, at $2195, is capable, inexpensive and relatively foolproof. If you want more capability, spend more for the GPSmap 496 and get terrain avoidance thrown into the deal.

Why not a sferics device, which have been available for three decades? A fair question. But theyre more expensive, require installation and interpretation and they provide little strategic avoidance capability. Furthermore, weatherlink can also provide real-time METARs and forecasts, which further inform weather judgment calls.

Yes, maintenance. More than 200 accidents a year-thats about 16 percent of the total-are due to mechanical or maintenance-related causes, although these tend to have a low lethality rate. About half are related to engine and propeller malfunctions of some kind. Its anyones guess how many of these accidents wouldnt have occurred if a maintenance technician spotted and fixed a fault before something serious broke. With most airplanes flying less than ever, proactive inspection and maintenance is probably more important than ever. Our notion of this is off-annual inspections of the engine and prop for emerging faults, battery capacity tests and regular oil analysis. A thousand bucks will go a long way in this arena.

Late-model wonderplanes like the Cirrus and the Diamonds are notably

crashworthy, but its no secret that older airplanes arent. Seatbelts are a player in this grim calculus and even though many older models have shoulder harnesses, they arent necessarily good shoulder harnesses.

ELTs have been mandated in light aircraft since 1974 and in the intervening three decades, they havent compiled an impressive performance record, in our opinion. Thats because 121.5 MHz ELTs havent proven reliable, often failing to activate and are sometimes difficult to locate even when they do. Further, false alarms have been numerous and a nuisance.

The new 406 MHz ELTs are just becoming market ready and not many owners seem to understand that 121.5 MHz wont be satellite monitored after February 2009. That means if you crash in the boonies and you want to be found, a 406 MHz will actually make that possible, especially if its equipped with GPS, which the emerging models will have as an option. For about $1000, 406 ELTs look to be meaningful, high-value survival enhancers to us and the technology that may finally deliver on the failed promise of ELTs. If a 406 beacon isn’t in the cards, consider a personal locater beacon instead.

Conspicuously absent from our list is one of the most popular avionics upgrades: traffic awareness/collision avoidance equipment. Our reasoning is simple. At prices ranging from $10,000 to $20,000, this equipment is simply not a cost-effective safety enhancer. There are typically about 10 midairs a year and, surprisingly, the lethality of these accidents is 50 percent, lower than in VFR-into-IMC or weather-related accidents.

Its not that we think collision avoidance gear isn’t a good thing or that it doesnt perform. It is and it does. Its just that it requires spending a pile of money to address a relatively small risk, probably at the expense of deferring investment in things that mitigate larger and well-documented risk, such as weather-related accidents or runway loss of control. You could argue that 10 midairs represent a bigger risk than five thunderstorm accidents, but weather is a much broader risk than just thunderstorms.

Similarly, we don’t have backup gyros or vacuum systems or terrain awareness on our must-have list for the same reason. There are more immediate concerns, in our view. On the other hand, if you fly serious IFR and have invested in all the obvious safety enhancements, a backup gyro might rise to the top of your list, even though it isn’t on ours. If approaches in real weather are your thing, terrain awareness equipment may be a priority due to your exposure profile. But the big picture numbers don’t identify it as a big risk area.

The point of this exercise is to adopt a hierarchical approach to safety-related aircraft investments based on a cold analysis of where the real risks lie and not to succumb to gadget fever to address smaller risk areas by spending a lot of money. The latter results in “between-the-ears” safety enhancement; you may feel safer, but the reality is that youve left large risk segments unaddressed. Mundane training and upgrades like seatbelts will never be as sexy as a new GPS or a traffic minder. But they may be more likely to save your life.