The NTSB recently announced the probable cause of the Kobe Bryant fatal helicopter crash was the pilot’s flawed decision to continue visual flight into instrument conditions in high terrain, causing him to lose control of the S-76B helicopter. In this AVweb video, Paul Bertorelli dissects the NTSB hearing and presents the critical findings in context. For a complete analysis on the risk of Part 135 helicopter flying, see this AVweb video.


  1. Seems odd that a 135 charter company would only be allowed to fly VFR? Why was that – choice of the company or a restriction imposed by the FAA for some reason? The aircraft was presumably capable of IFR flight (with an autopilot?), the pilot was qualified, but rusty. Why have a VFR only charter operation, especially in California where the coastal low ceilings are very common?

    • “Seems odd that a 135 charter company would only be allowed to fly VFR?”

      Well, unironically this particular accident is a very good example of the FAA’s wisdom in this stance. To answer your presumably serious question, how about because helicopters crash often and aren’t good IMC platforms?

      If the numbskull at the controls had simply stayed in VMC conditions, (a colossally simple task) they’d all be alive. But don’t take my word for it…see for more fun-filled accident reports of idiots killing their passengers.

    • I think the discussion missed a couple of points leading up to the crash. The pilot did not follow the 101 Freeway after passing the Van Nuys airport. He went around VNY to the north and was directed to follow the 118 Freeway which somewhat parallels the 101 but about 8 miles north and at a higher elevation.

      He was sent on this route after having been held outside the Burbank airspace an extended wait for a special VFR clearance through Burbank and then through Van Nuys airspace.

      It is easy to second guess the pilot’s decision process. Rental cars were easily and quickly available at Burbank or VNY and would have gotten the passengers to their game on time.

      Tower communications Van Nuys indicated that he was flying lower than the assigned transit altitude and offsetting slightly to the south of the 118 Freeway. Subsequent multiple tower transmission indicated that he was flying lower and to the south of the 118 freeeway where terrain is lower.

      Following the falling terrain to the southwest he arrived into the “bowl” where the 101 freeway meets Malibu Canyon … There’s rising terrain in every direction including rapidly rising terrain south of the 101. Essentially he had flown into a box canyon

      Finally as I recall he had logged 85 hours of instrument time, about 1% of his flight hours . Probably most of that in sims and perhaps a few hours during checkrides.

      Sadly there were so many alternatives discarded including getting an IFR clearance to climb to VFR in the Burbank/Van Nuys area

  2. The final minute is the “money quote”–NTSB thinks mandating MORE training is better–and it’s hard to disagree with that. Mandating a “snitch” data/voice recorder on the aircraft crosses a line.

    Putting CVR/data recorders on jet charters didn’t directly affect those charter crashes. Why does NTSB think this would be different?

    As a former Corporate operator that also used to do charter, the odd thing was that EVEN WITH THE HIGHER REQUIREMENTS FOR MAINTENANCE, DISPATCH, AND PILOTS FOR CHARTER–CHARTER HAD A WORSE SAFETY RECORD THAN CORPORATE. The very same pilots, the very same airplanes–but the more restrictions, the WORSE the safety record. There WAS a time when a small corporation might decide to buy an airplane–they had the money, but they moved to a larger piston twin or turboprop instead–leasing the aircraft to an FBO charter operator for additional revenue. As the FAA enacted more and more regulations on the operator (maintenance, dispatch, check rides, and operations) the cost went out of sight. Within years, many charter operators decided it just wasn’t worth it. We fly the same airplanes–the same pilots, and the manufacturer maintenance procedures–and turn in a better record. The result? The manufacturers turn out FEWER airplanes (think light twins, for example–virtually disappeared). Not only did the FAA kill the market for those upgraded aircraft, but they killed the market for new entrants into business aviation–there are so few operators willing to do business today that the only way for many would-be aircraft owners is to go all the way to BUY the aircraft! With this program (or should I say POGROM)–the FAA will achieve perfect safety–WHEN THE LAST AIRCRAFT IS GROUNDED.

    Instead, the FAA/NTSB feels that MORE REGULATION is the key to improving safety–the aeronautical equivalent of the old maritime “The floggings will continue until morale improves!” THINK about that–do they really feel that pilots will be MORE AFRAID OF THE FAA THAN THEY ARE OF DYING?

    If the FAA/NTSB REALLY feels that electronics and technology will improve safety, perhaps they should look inward. Why hasn’t anybody asked the question “WHY WAS THIS MULTI-MILLION DOLLAR AIRCRAFT AND HIGH TIME PILOT NOT FLYING IFR?” He had the flight time–yet like most helicopter pilots, not much IFR. The reason–the FAA has made it difficult to certify pilots and aircraft for IFR–instead, both civil and commercial operators attempt continued flight in VMC. This is the very same “reasoning” that led to the Buddy Holly crash–night departure in marginal weather conditions by a pilot not rated. Where do you suppose the accident record would be today without the ability of almost every commercial pilot to fly IFR? The multi-million dollar AIRCRAFT was equipped–right down to an exotic autopilot–why didn’t the pilot fly IFR? (This would be like allowing VFR-only in a corporate jet.) I don’t have that answer, but I’m guessing that the FAA-approved OPS SPECS DIDN’T INCLUDE IT.

    No–I lay the cause of this and similar accidents right at the feet of FAA/NTSB for creating an unsafe operating environment–failure to keep up with helicopter instrument capabilities–failure to come up with workable regulations–and failure to create an environment that not only allows but encourages instrument flight in helicopters (and fixed wings) that are properly equipped.

  3. For better or worse, a government agency with oversight on any operation will always think that more regulation and better procedures will solve the problem. Personally, I think that Paul’s comment on “get home itis” probably fits the situation pretty accurately. Bryant had a history of firing previous pilots who did not get him or his family to their destinations on time. That puts even more pressure on the pilot to go, especially when the charter company determined the forecast weather was okay to do so. Once airborne, the subtle trap slowly closed on the pilot as so often happens in these situations. He feels pressure to press on, even while the weather deteriorates around him. It doesn’t make sense that the FAA sees fit to allow VFR flights in aircraft that are fully equipped to fly IFR, rather than taking the opposite approach. If this was a private pilot flying a fixed wing plane, the expectation would certainly be that the flight be done under an IFR flight plan. In the end, individual human judgement is usually the trigger that starts the dominoes falling. Not sure there is a perfect system to prevent that.

  4. This video perfectly shows why FRAT scores and SMS have dubious value. It was mentioned that the score was not high enough at departure time to warrant consulting higher management in the company. Consider most line pilots would consult their chief pilot if a FRAT score is high enough. The accident pilot was the chief pilot. FRAT scores can be easily manipulated and require management to take them seriously. I worked for a company that has an FAA accepted SMS. Management (chief pilot and director of ops) took SMS seriously. Unfortunately the owner of that company did only if it did not cost him revenue or money making charters. In the end it just became a worthless paperwork exercise that did nothing but take extra time and delayed departures.

    • Matt I hear you about the need for top management, in your case your former company owner to buy into flight standards. Easier said than done I know, but ultimately it becomes incumbent on pilots to choose for whom they will and will not work. I know because early in my career I had to do that. As for FRATSs, the bottom line is that even if the owner has not bought in to your safety management system, the FRAT is what you make out of it. Even if the FRAT has not required a no go decision on your and management’s part, it should be the kind of FRAT which has heightened your awareness of the impending flight environment and caused you to think about it. That in my book is not “a worthless paperwork exercise that does nothing but take extra time and delayed departures”

  5. I’ll make a similar comment here as I did on the Youtube video. I don’t really agree that deciding to climb into IMC was the cause of this crash. To me, it was the pilot’s inability to fly the plane in IMC when it was equipped with a very good autopilot. All he had to do was turn it on, best done before he initiated the climb.

    • Really good comment.

      As I noted in a prior comment the discussion of the route taken was not what he actually flew. When he got to the south of Burbank airport it was IFR and he was held to the southeast of Burbank and then cleared to enter under special VFR and directed to fly north of Van Nuys Airport .

  6. The terms “Inadvertent IFR” and “Unintended IFR” are misnomers: there is no such thing as ‘unintended’ or ‘inadvertent’ IFR; most people who encountered it knew for minutes or miles that such a thing was possible. It should be called “Advertent IFR”…

    When it comes to pushing weather limits, the helicopter mantra is “If you go down, you slow down”.

    However, if you’re arsin’ around low level (say, 500 – 1000 feet) in poor viz at cruise speed and you “inadvertently” enter IMC conditions, you:
    1. Go on the dials.
    2. Note your heading.
    3. Note the reciprocal.
    4. Make a level, 180 degree turn toward the low ground.
    5. Fly back into better conditions.
    NOTE: if you’re in hilly/mountainous country that precludes a turn, then you’re committed to climb to a safe altitude on a heading that will keep you clear of the terrain: synthetic vision (as I have in my aircraft) is a great boon in this regard.

    Now, if you’re hell bent on getting through, then you adopt another procedure: as the ceiling and viz decrease, so do you; you reduce both altitude and airspeed
    Your aim is to give yourself reaction time, so you gradually lose altitude and airspeed; you may be able to continue to destination at a couple of hundred feet at forty to sixty knots or it may turn out that you wind up at ground level in the hover. (Of course, you should have beat a hasty retreat long before this, but at least you’re in one piece and can make a decision to land or turn around to see what’s behind you.)
    The take-away is that unless you’re reasonably proficient on instruments, at the moment you start slowing down and losing altitude to maintain ground contact, you ‘get out of Dodge’ while the getting’s good.

    The pilot in question was not a neophyte; why he did not slow down is a mystery…

  7. Too low / too slow combinations are often in what’s called ’dead man’s curve’ for helos, due to the need for sufficient energy to safely auto. Possibly that was a factor?

  8. A wise man once staid “Stupid is expensive, stupid is painful and you can’t fix stupid”. Also Don’t fly VFR into IMC. That is all the “more training” you need. The FAA is clearly having a hard time trying to fix stupid. I don’t like the assumption that something must always be done for high profile events. That is not always the case. Nature is self correcting when we erase ourselves from the gene pool. I don’t like loss of innocents who never fully understood the risks. I don’t believe I will ever fully either, but I have at least accepted the possibilities which I can and cannot control as I attempt to be more than human. I know that is not very forgiving but neither is IMC. IMC simply means don’t go. We all need a reminder so I watch these videos, thx Paul. This is not neither an opportunity nor crisis in which to push narc boxes on operators. Flight recorders can help us understand what made a pilot stupid but technology doesn’t fix it or tell us what we didn’t already know. How many crashes have flight recorders prevented, I mean exactly how many?

  9. Given Mr. Bertorelli’s thorough presentation, plus all of the preceding comments, here is what I recommend:
    (1) All helicopters should be equipped with state of the art instrumentation (2) All helicopter pilots should
    be required to learn and master said instrumentation (3) Pilots should have at least as much experience w.
    instrument flying as w. visual flying, at every stage of professional certification (4) Wherever possible, two
    pilots or crew members should be on every flight, regardless of purpose or distance (5) Pilots should avoid
    conflicts of interest arising from personal or fiduciary relationships with passengers, or anyone on board.
    Pilots who have such relationships should recuse themselves from flights in which those individuals are
    on the manifest, or be forbidden to do so. These rules and regulations are both logical and imperative.
    They will make it far easier for pilots to perform their duties without fear of reprisal, while requiring
    both charter and commercial operators to demand rigorous compliance with all safety procedures.
    Pilots will be carefully vetted prior to each flight; all aircraft, including helicopters, will be properly
    outfitted and maintained in first-class condition. One last point: all risk factors, including weather,
    must be carefully reviewed within 10 minutes of scheduled departure. Flights without contingency
    plans will not be permitted. Conversely, pilots and crew members reserve the right to cancel or
    postpone any flight if in their judgment it is either too hazardous to fly, or that the aircraft is not
    entirely capable of doing its job, without being penalized by an employer for making that decision.
    If the FAA/NTSB adopted and enforced these simple rules, that would accomplish what Mr. Jim H.
    insists upon, without bloating their bureaucracy or hindering the growth of the aviation industry.
    Human error can never be eliminated; but it can and should be reduced. Not all accidents are
    preventable; but there are no reasons, no excuses and no forgiveness for gratuitous loss of life.

  10. Great video Paul. Sage comments and observations.

    I am not a helo pilot nor have I been in a helicopter. Never had the desire. From a purely mechanical perspective, too many moving parts, all moving at the same time, not necessarily in harmony with each other.

    Because of the high profile, celebrity status of Kobe Bryant, there has been much dissection of this accident chain by many aviation analysts. As with Paul’s videos, combined with several other reputable, knowledgeable video commentary, I have learned a few things that relate specifically to helo operations. Many times, us fixed wing fliers have assumed equipment such as an autopilot correlates equally when flying either a helicopter of fixed wing aircraft. Not exactly so, I am now finding out.

    I did not know the S-76 autopilot requires a minimum of 50kts forward speed to work. Before this crash, I asked the question, why didn’t the pilot simply engage the autopilot, take a break to reassess his situation? I have an S-Tec 30 with altitude hold in my old airplane. While I enjoy hand flying the airplane more than just sitting back, I admit it is really nice to have the AP when occupied with navigating, taking pictures, or rummaging thru my flight bag. I have experimented with putting the airplane in some different attitudes with AP off, and then engage the AP to see essentially how fast the AP can put the greasy side down and shiney side up. Since the Bonanza stalls around 55mph clean, and that airplanes are always moving forward outside of a hammerhead or whip stall, even when stalled, the AP has all the cues available to work.

    The helicopter AP’s need the same input…forward motion to work. Until this accident investigation, I assumed a four axis autopilot in a helo would allow for the AP to maintain a hover in IMC. The AP will not allow for controlled hover flight. As I understand it, just like most if not all fixed wing AP, it requires forward motion to work. This is why, for the AP to work in the accident S-76, the pilot would have to maintain a minimum of 50 kts forward speed. That explains to me, why he did not engage the AP.

    The accident pilot was a CFII. He had taught IMC flight. He had 75 hours TT “under the hood”. But only 1.5 hours of actual IMC experience. I have no idea of his TT flight instruction experience, nor how many students he might have had during his instructing career. But it begs the question to a VFR pilot like me, how do you become a CFII, teaching others to fly IMC, when your personal TT in actual IMC is practically nil?

    The pilot had developed a friendship with Bryant that appears to be deeper than Kobe simply preferring this pilot because he always got him to the destination no matter what. Another part of the accident chain of destinationitis.

    He knew the route extremely well plus understood the normal ATC traffic flow, expectations, and communications, including the likelihood of holds as ATC directed the normal conga line of helo traffic. He knew pretty much when and where radar coverage would be lost or compromised.

    He had signed off on the risk assessment as reasonably low. I believe he thought this was simply going to be another day at the office. I believe he was so confidant of being able to complete this flight, his only “plan B” was a zoom climb thru the marine layer if he lost sight of the ground, to pop out on top. From my understanding, the cloud base was approximately 1000-1100 ft with the tops at 2400 ft. At 140 kts cruising speed he was flying at, he would be in soup for less than 30 seconds, if that. Look for a hole over or near the destination to come back down. Since there was a celebrity onboard, there would be a pretty good excuse to justify his “Plan B” should ATC question the zoom climb on top with the added bonus, the pilot would get Koby to his game even faster.

    I agree with the NTSB suggesting onboard cameras and CVR capabilities for part 121/135 ops. I have heard several airline captains comment that these onboard visual and listening devices enforce procedures because, in most cases, while no accident or incidence may result, out of parameter flying gets flagged with someone in the employment chain making a call asking why the pilot(s) may have deviated from established norms. In other words, knowing “big brother” employment is watching helps to discourage too much independent decision-making when an abnormality of deviation from established protocol will be clearly seen by someone outside the cockpit that has something to say about your paycheck. And if found in the wreckage, it helps the accident investigating agency come to a logical conclusion much faster.

    Obviously, when he zoomed into the murk, he lost control, spatially disoriented, with the inevitable left hand descending , high speed corkscrew dive into the side of one of the many hills in the area. Another 90 feet, another few seconds, and we would have never known about his zoom climb to get on top.

    I wonder how many times previous to this flight, there was similar behavior by him or if other helo pilots might habitually do likewise in similar situations. In many cases of accidents by both fixed wing pilots as well as helo pilots, there has been an established pattern of behavior that pushes the boundaries of safety somewhat regularly. As other noted, this pilot often flew under SVFR. That appears a normal flight for most professional VFR charter pilots in that area. I am not buying inadvertent VFR flight into IMC. I now wonder, how often he performed “Plan B” prior to the fatal flight.

  11. I’m just a VFR pilot with experience in nothing more complex than a C 172…about VFR
    I am under the impression that VFR over the California coast would require…
    1. maintain 2000 feet from human structures. (both horz and vertical sep)
    2. maintain 1000 feet below ceiling.
    3. Stay out of Class D airspace without permission.
    4. 3 miles vis min.
    If the ceiling is 2000 feet…that means I don’t fly..period.
    I assume that staying at least 2000 feet from human structures is partially based on the need to
    minimize noise impact on people from my loud aircraft.
    So..don’t the same rules apply to VFR in a copter?
    And don’t complex, expensive copters require a co pilot? Insane if FAA doesn’t require Co pilot..right?