Print Friendly, PDF & Email

MH370 Unknowns, Plausibilities, and Unlikelihoods

Assertions, interpretations, and inferences fall between the Scylla of MH370 facts (see preceding blog entry of 31 May) and the Charybdis of conspiracies (the next blog entry, “One Conspiracy to Rule Them All,” 14 June). As before, footnotes provide clickable sources and editorial comments appear in italics.

Let’s kick off with a quote that leaves many perplexed the first time they see it, but on re-reading makes sense (to me, at least):

“…as we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns—the ones we don’t know we don’t know.”

Donald Rumsfeld to reporters re the Second Iraq War (Feb, 2002)1.

In my business career, we referred to unknown unknowns as Level III ignorance: being too dumb to know what we didn’t know. Level III was a dangerous place to live. We would strive to move from Level I (“Dumb and we know we’re dumb”) to Level II (“Dumb, but we know what we need to do to improve”), while trying to avoid the hubris that typically nurtures Level III ignorance.

When I was raiding Wikipedia for the exact Rumsfeld quotation, I stumbled upon another category of unknown that neither Rummy nor I had seen before, complements of psychoanalytic philosopher Slavoj Žižek:

“There is a fourth, the unknown known, that which we intentionally refuse to acknowledge that we know. If Rumsfeld thinks that the main dangers in the confrontation with Iraq were the ‘unknown unknowns’, that is, the threats from Saddam whose nature we cannot even suspect, then the Abu Ghraib scandal shows that the main dangers lie in the ‘unknown knowns’ – the disavowed beliefs, suppositions and obscene practices we pretend not to know about . . . .”2 (Emphasis added)

Readers will appreciate why so many MH370 observers remain deeply unconvinced that we are anywhere other than in Level III (or worse, Level IV) ignorance. I’ve divided this blog entry into three sections, each simmering with potential controversy:




Next week we’ll step into deeper water by ranking the conspiracy theories by plausibility.

Remember, too, that none of is smarter than all of us. So please share your ideas or new sources of information in the Comments section, or email me at

Bradley West, Singapore, 7 June 2015



At the risk of spoiling the suspense, let’s start by describing the “filter,” the screening criteria for evaluating MH370 disappearance “conjectures” (subject to falsification) as distinct from “conspiracies” (which cannot be disproven, so by necessity are more open to debate).


The MH370 Conjectures Filter

  • The Inmarsat ping data were real, and the burst timing offset (“BTO”*) data were accurate. But we don’t know if the burst frequency offset (“BFO”*) data can be relied on for two reasons. First, very small errors have substantial implications for where the plane ended up (e.g. Δ1 Hz moves the flight path arc 100km). Second, the BFO transmissions could have been falsified, causing investigators to search along the southern arc* when, in fact, the plane had flown along the northern arc*, over land.
  • The plane landed or crashed along the BTO-defined arc, +50 nautical miles (58 statute miles or 93 kilometers). This is the distance adopted by the The Joint Investigation Team (“JIT”) [led by the US and UK (with Malaysia, China, Australia, and France also joining in)] and reported in the Australian Transport Safety Bureau report MH370 – Definition of Underwater Search Areas of June (revised August) 2014.3
  • If MH370 crashed into the ocean, it likely did so under expert control which in turn minimized the surface debris and helps explain why none has ever been found. As of 9-10 June, this topic heated up. Please see the follow-up post A Little More Meat for the latest thinking on what is a contentious issue.
  • There was a substantial gap between the loss of the transponder and ACARS systems and the commencement of what could have been autopilot (sans person) flying. Therefore, someone with expertise was flying the plane at or after 02:25, about 01:45 after takeoff (and almost six hours before the flight likely ended around 08:19).

 At least two people, one expert pilot and one other person, had to be involved if the sequence of flying and power-on/power-off events occurred as presently understood.

Considered in combination, the Conjecture Filter invalidates most popular explanations of how MH370 met its end.


*Key Terms

Burst timing offset (“BTO”) values take the time delay between pings to plot arcs defining the distance between the Inmarsat satellite and the recipient (in this case, MH370). MH370 and the satellite exchanged seven handshakes (to-and-from pings), with the seventh arc at 08:11 defining the last measured distance between the plane and the satellite.

Burst frequency offset (“BFO”) measures the difference between the expected and actual frequencies of the pings received. The Doppler Effect causes the differences, and analyses should show the direction of MH370’s travel.

Southern arc describes a flight path west and south of MH370’s last known radar appearance 230km off Malaysia’s west coast island of Penang, heading south to the west of Bandar Aceh (Sumatra) and then off Western Australia over the Southern Indian Ocean (“SIO”). The official investigators and many unofficial experts (e.g., on the Independent Group, or “IG”—see below) believe that MH370 crashed along the southern arc.

Northern arc describes a flight path north and east through parts of Thailand, northern Burma, southwestern China, Azerbaijan, and Kazakhstan. Most observers discount the northern corridor as a viable path because of the BFO directional readings, plus the likelihood that one or more countries would have spotted the plane on radar had it overflown their territory.

The IG (Independent Group) is a collection of scientists, engineers and aviation experts who play ongoing roles in performing advanced analyses of the Inmarsat-sourced data. In 2014, their calculations convinced the official search team to shift the SIO search south. It’s a serious group of researchers.


Inmarsat data analyses and integrity are two crucially important issues

The JIT, the Australian Transport Safety Bureau, plus the intellectual heavyweight Independent Group (“IG”) led by Duncan Steel, agree that the key to finding MH370 lies in interpreting data sourced from Inmarsat satellite 3F1. If we take the numbers and their interpretations at face values, most experts concur that the flight crashed somewhere in the Southern Indian Ocean (“SIO”) far from Australia’s southwest coast. That may or may not be the case, however. Read on.

32. Flight Path from BTO and BFO
Source: “MH370 Story,” Richard Godfrey for the Independent Group (22 Feb 2015)

Inmarsat is a publicly-listed UK-based satellite communications company. Inmarsat began life providing international shipping companies with satellite-based communications to their vessels. Later the company branched into aviation. Today Inmarsat has five business verticals, one of which is solely US government (military and civil) focused and one which is non-US governments (military and civil) dedicated. So the US (and UK, and other) intelligence services are Inmarsat customers and, one would think, collaborators from time-to-time in the development and application of new satellite-oriented communications methods. There’s nothing sinister in the preceding, but it’s worth noting Inmarsat’s pre-existing close relationships with intelligence agencies, esp. the NSA and CIA.

One of Inmarsat’s services, Classic Aero, allowed MH370’s engines to provide status updates to MAS via satellite, VHF or high frequency transmissions. Inmarsat 3F1 is a satellite in high altitude (35,800km) geostationary orbit above the Indian Ocean off the coast of equatorial East Africa. Even after the ACARS was disabled (or malfunctioned), MH370 continued to send and receive “log-on interrogation” messages (a.k.a. LOI or pings) with satellite 3F1 from 02:25 until 08:19 Saturday 8 March.

Analysis of the pings provided meaningful information in respect of MH370’s location when additionally analyzed in one of two ways:

  • Burst timing offset (“BTO”) values take the time delay between pings to plot arcs defining the distance between the satellite and the recipient (in this case, MH370). The investigators used the pre-disappearance BTO readings when the location of the plane was known with precision to calibrate their BTO settings to estimate the distances once the plane’s transponder and ACARS had gone dark. The accuracy is thought to be good to approximately 1°, about 110km. The BTO data were used to generate the “northern corridor” (Malaysia-Thailand-Northern Myanmar-southwestern China, Azerbaijan, and ending as far as Kazakhstan by 08:19 Malaysia time) and the “southern corridor” (through the remote Southern Indian Ocean towards Antarctica). Indonesia-based aviation expert Gerry Soejatman writes, “. . . the BTO (pings) are not doubted whatsoever.4
  • Burst frequency offset (“BFO”) measures the difference between the expected and actual frequencies of the pings received. The Doppler Effect causes the differences, and analyses should show the direction of MH370’s travel. Running the numbers proved to be highly complex, with six sets of adjustments necessary just for the Doppler readings alone. Add to this a geostationary satellite that actually wobbled 1.5º either side of the equator, diurnal temperature differentials, errors in satellite and aircraft oscillators, and the lay reader begins to appreciate the degree of difficulty involved. Suspicions that the BFO figures either aren’t accurate or are being misinterpreted sit at the heart of doubts about Inmarsat-derived data.
16. All Possible End Points
Source: “MH370 Story,” Richard Godfrey for the Independent Group (22 Feb 2015)


Combining the BTO and BFO values led the authorities to announce that MH370 was headed south along the southern arc when it ran out of fuel, crashed and sank sometime around 08:19 on Saturday morning, 8 March. Game, set and match? Perhaps, but I have several problems with this packaged solution:


1. It’s too damned pat (I). In a case in which so little is known with certainty, the authorities ask the public to accept unchallenged the findings of what ultimately is an extrapolation of seven “handshakes” (ping pairs) and one missed handshake.

I’m not a mathematician, but I am numerate. The computations are so complex that, even as of early June 2015 academic heavyweights and talented amateurs are still re-running their analyses based on revised assumptions and ever more nuanced permutations (see, MH370 threads). Consider, too, that miniscule changes (e.g. 1 Hz in fixed frequency bias) would result in a 100 kilometer difference in aircraft location if flying along the 7th BTO arc.5

Yet in the history of air crash investigations, BTO and BFO analyses had never been done before MH370. Even with impeccable data, you would expect hiccups. Compounding matters, even collecting the ping data was something new to Inmarsat. Only in 2013 did the company upgrade the Perth Global Earth Station to record BFO and BTO log-on interrogations. Nevertheless, the Australian Transport Safety Bureau reported that “the [Perth] receiver is an old unit that is unable to accept negative latitude values to service southern hemisphere locations as described.”6

Inmarsat was slow to release information to the Independent Group (“IG”) from the outset, taking until 25 May 2014 to supply the ping data. More than a year later, the IG hasn’t received a full set of Inmarsat data. Yet the IG (with less information and unpaid volunteers) has consistently found errors in Inmarsat staff’s work, thereby helping narrow the search grid to higher probability areas. So even if we endorse the IG’s Popperian approach to hypothesis formulation and falsification, it could still be working with incomplete or faulty data. Meanwhile, the official investigators likely have the complete data set, but aren’t necessarily doing the most imaginative or insightful analyses.

Reading the ATSB’s June 2014 report reveals how much pressure the official investigators are under to show results of their expensive search. The authors devote pages to illustrating how large the MH370 search area is, how poor the data are, and how unlike Air France 447 was MH370. A multi-country, politicized investigation isn’t likely to produce the analytical breakthroughs that independent researchers generate working out of the limelight on a collaborative basis. Not surprisingly, in the space a few months in 2014 the target search area moved almost 2000 kilometers south as the IG found errors, Inmarsat released better data, and the IG conducted further analyses.

In April, 2015, aviation professor Jason Middleton at the University of New South Wales in Sydney said, “These calculations rely on a whole bunch of issues that are not easily verifiable by outside sources. The science is not demonstrably repeatable.”7

In May, 2015, Professor Middleton was back in the news. “And although I’m not suggesting [Inmarsat] have done anything improper, the search area relies very much on their calculations, and if they have made errors, we are not able to replicate their calculations. And there’s a chance they’ve stuffed up and the plane is not there at all.”8

42. Most Probable End Point
Source: “MH370 Story,” Richard Godfrey for the Independent Group (22 Feb 2015)


2. It’s too damned pat (II) . . . look at the time lags . . . the authorities seemingly didn’t start out believing the BTO/BFO data based on the slowness with which they acted:

  • 9 March Inmarsat did its initial BTO and BFO analyses on the ping data and internally concluded that MH370 was airborne for seven hours after the pilot signed off at 01:09.
  • 11 March Inmarsat alerted the US (e.g. FAA and NTSB) of its findings
  • 13 March the US shifted its search from the Gulf of Thailand to the Indian Ocean, redeploying the destroyer USS Kidd to the SIO. The JIT followed suit later the day.
  • 13 March two government sources told an ABC News reporter that, “We have an indication that the plane went down in the Indian Ocean”9, but then went on to say that they didn’t have a source or any further details.

If the BTO and BFO numbers were watertight, why the twin 48-hour delays at a time when every hour counted? The natural conclusion is that the senior officials believed that the numbers, the analyses, or both were speculative. Or maybe there was intentional foot-dragging for X [fill in your favorite conspiracy]. Or maybe the bureaucratic infighting was so ferocious that it was easier to let Malaysia (and other countries) keep searching the wrong areas until they faced up to the new information that MH370 had flown elsewhere.

The US search shifted 13 March from the Gulf of Thailand to the SIO. At this time the official version was that the JIT had the BTO data (but not the BFO figures). The BTO northern-southern arc hypothesis was sufficiently controversial that the Malaysia Defence Minister Hishammuddin Hussein would admit only to a “possible air turn back” even as the Americans sent the destroyer USS Kidd northwest out of the Malacca Straits and into the Indian Ocean to pursue a search along the southern arc.10

With the BTO data indicating either a northern or a southern arc with no further information to indicate direction, the Americans had no hesitation in asserting that the direction was south. The rationale cited was common sense. If the plane had taken the northern arc, then surely other countries’ commercial or military radars would have picked up a trace. But asking the public to swallow this when we’re also supposed to believe that the US Navy in Singapore, and militaries in Malaysia, Thailand, Indonesia and Australia in combination lost track of the plane at 02:22 and never saw it again despite it’s being in the air another six hours?

So is detecting-a-plane-with-radar harder than it sounds? No, apparently not, according to the US, because there’s no way a 777 could have flown over Thailand, Burma, southwestern China, Azerbaijan and Kazakhstan without radar operators detecting it. To my eye, the first line up of countries and military competencies is more impressive than the second. With the exception of China (and India, which lies to the south of the BTO northern arc), there isn’t any country on that list I’d expect to detect and act in real time to identify any strange aircraft flying through its airspace as long as it wasn’t approaching a military target.

18. Radar Traces
Source: “MH370 Story,” Richard Godfrey for the Independent Group (22 Feb 2015)


Just a few days later, Inmarsat surfaced with the BFO data. Now we had a further source of support for the southern arc hypothesis. If you’re not the least bit suspicious, then you are more trusting than I. Alternatively, could it be that the US (esp.) and ATSB had already made up their minds based on non-public information? The convenient arrival of BFO data may have provided the cover story that justified a decision that had already been taken for reasons having nothing to do with Inmarsat pings.


3. No plane and no debris. Despite fourteen months and over US$90m spent, searching along that southern arc hasn’t turned up any wreckage or other physical evidence. The ATSB report of June 2014 said the last five hours of the flight were most likely conducted on autopilot with an oxygen-depleted airplane full of dead passengers and crew. If so, the out-of-fuel plane would have crashed very close to the computed BTO path. Furthermore, an uncontrolled decent would have resulted in a disintegration as the plane plunged into a rough ocean. The resulting large field of floating wreckage would have been readily visible to searchers.


In April 2015, aviation professor Jason Middleton said, “The fact that nothing has been found by the way of debris suggests to me that they’re looking in the wrong place.”11


4. The BFO data may have been “spoofed” (i.e., falsified). On 16 May 2015 independent researcher Victor Iannello (member of the IG, but publishing in an individual capacity) dropped a bombshell by publishing a paper12 which built on work of the IG, especially New York-based aviation consultant Jeff Wise, to extrapolate from a previously-identified unexplained event. At 02:22 the airplane vanished from radar. We know from the ping data that at 02:22 someone on board also manually re-booted the electrical systems. Three minutes later at 02:25 (the boot process takes 2:40) the satellite data unit (“SDU”) pinged Inmarsat 3F1. An incidental message transmitted simultaneously reported that the airline entertainment system was rebooting as well.

Jeff Wise, Victor Iannello and Gerry Soejatman provided separate descriptions of how hijackers could have cut the power to either the SDU or most of the cabin (including the SDU). Next, hijackers may have plugged in a preconfigured laptop to act as a proxy SDU which would then disguise the BFO readings once power was restored. More specifically, the altered BFO data would appear to indicate travel to the south when the plane actually was following the northern arc. Raw meat for the conspiracists among us.

Wise, Iannello and Soejatman have shown that at best the official investigation may be mired in Level III ignorance: they still don’t know what they don’t know. And it could be Level IV ignorance, even worse: “. . . the disavowed . . . obscene practices we pretend not to know about” (see page one of this blog entry for Slavoj Žižek’s full quote). There’s been no official reply to Wise’s work (which appeared in New York Magazine 23 February 2015) or Soejatman’s blog of 1 March 2015, and it’s too early to expect a response to Iannello’s paper which was published on Duncan Steel’s ( website only on 18 May 2015.


Were there other ways of tracking MH370?

In summary, it’s as if the US (and Australia, and maybe the UK) already knew where the plane had crashed, and needed to produce something that gave the same directional results without compromising intelligence sources. The resulting PR campaign seemingly placed disproportionately high confidence in small sets of numbers that, on further investigation, could have been completely faked with only a moderate amount of software programming and no specialist equipment. The key units were accessible to anyone who could access economy class, basically anyone on board.

Potential covert sources of information regarding MH370’s direction of travel and final resting place include:

  • The US’s IUSS underwater submarine tracking network in the Indian Ocean
  • Data collected by US and/or allied submarines patrolling in the Indian Ocean
  • US and Australia military radars around the region
  • US spy satellites

A suspicious mind could conclude that the US exaggerated the reliability of the BTO and BFO data, fed it to the public and scientific communities, and then waited for them to reach the conclusions the US wanted. Furthermore, a conspiratorial mind might even conclude that the US didn’t want searchers looking anywhere but the Southern Indian Ocean where several years could be harmlessly (but expensively) spent towing probes and mapping undulations on the pristine seabed . . . perhaps to be used to chart submarine routes through one of the least-explored places on the planet.


But could all the Inmarsat satellite data be fake?

Via emails exchanged in late May 2015, I posed the following question to two MH370 investigators of renown, and neither one thought this likely. (Neither did I, but I had to ask).

Bradley West,Could Inmarsat have faked the BTO and BFO ping data at the Perth Ground Earth Station? Here is where I’m out of my depth and need your help. Presumably the Inmarsat equipment automatically captures the pings and saves the handshake data as it’s received. Presumably Inmarsat has enough redundancy in their system that there’s another server located elsewhere backing up this information in real time/hourly/daily (?). If that’s true, then it would be difficult or impossible to introduce false data without a very complicated conspiracy.

But consider the alternative. If you had a senior enough person in Perth, could he have altered the only computer records to show different BTO and/or BFO values? (In my mind’s eye I see a senior exec coming in at night and sitting at a terminal to alter and then save this raw data.) If that were the case, relatively few people would have been required to perpetrate this fraud.

The experts were kind enough to respond promptly:

“In principle, yes. However, the complexity is so high that the possibility of anyone doing this is extremely remote. The idea of such ‘spoofing’ has been aired by many, but I think that it’s more appropriate for a novel than any realistic analysis. The IG* has chewed over this concept many, many times. Of much more interest is the interpretation of the data in hand, and possible alternatives which adhere to those data.”

Duncan Steel (email to Bradley West, 28 May 2015)


“If you had a senior exec in on it, then all kinds of kooky things could transpire. I’ve never been a fan of this kind of scenario, though. Seems like bad hacking, to have to have a confederate on the inside!”

Jeff Wise (email to Bradley West, 28 May 2015)


Disabling the ACARS, SDU and Entertainment Systems is a Game-Changer

Recall from the first blog entry The Facts of the Matter (31 May) that specialist knowledge was required to disable the main ACARS navigational system. Furthermore, there remains an active debate as to whether a pilot could switch off the VHF and SATCOM panels in the cockpit to indirectly disable the ACARS. Based on the work done in a B777 simulator in November 2014 by aviation consultant and IG member Michael Exner, it seems that the ACARS can be switched off from the cockpit but “. . . only by those with specialized knowledge.”13 The alternative method would require someone to access the E/E bay via a trapdoor which in turn was located under the carpet in the first class galley. Others have challenged whether the B777 simulator and the actual plane (i.e., 9M-MRO) flying as MH370 were configured identically. For now, let’s consider the issue as unresolved. But if it wasn’t possible to turn off the ACARS from the cockpit, the case for a single pilot suicide falls apart.

Next, consider the unrelated action. Either by design or accident, power also was cut to the SDU and entertainment system sometime between 01:07 (last normal ACARS transmission) and 02:22 (when power restarted), and at 02:25 (signaled the Inmarsat satellite). By coincidence or design, 02:22 was when MH370 disappeared off Malaysia military radar, never to appear again in records released to the public.

SATCOM expert Gerry Soejatman described the two ways power could be cut to the SDU. One was the loss of power only to the SDU-SATCOM antenna, and the other the cessation of power supplying half the cabin (see “The Facts of the Matter” 31 May) along with most of the cabin lights and the in-flight entertainment system.

The reason for bringing up again this curious incident is that, at the time the power was switched back on at 02:22, the plane was (at least) turning sharply under active pilot control and (quite possibly) also climbing rapidly. There’s no way that a pilot could fly the plane and restore satellite data unit (and probably cabin) power simultaneously. The flying maneuver couldn’t be done by the autopilot. The alternative “lone gunman” theory would have required the cabin’s electrical systems to have partially shut down—perhaps due to a fire in the avionics bay—and then come back on again. All the while, neither pilot nor co-pilot got on the radio to issue a mayday call.

Someone was flipping switches and/or pulling (and reinserting) plugs. With at least one pilot at the controls, then there had to be a second participant in whatever transpired on MH370.

We also know that whoever was in control of the plane wanted to make it invisible, resulting in the transponder and ACARS being disconnected early on. If the culprits had taken the trouble to disable the SATCOM, then they would have left it off unless by restarting power this would serve their purpose. Any explanation of what happened on MH370 needs to take into account the powering down and then restarting of the SDU (and in-flight entertainment) systems by a second person.



Within ten days of the disappearance, the Malaysians were hinting at pilot suicide:


“We are now looking at one of four possible scenarios. These are sabotage, terrorism, hijacking and personal problems.”

Malaysian Chief of Police, Khalid Abu Bakar (17 Mar 2014)14


To my eye, that list reduces down to two: pilot suicide and terrorism, with a failed or successful hijacking being the only two plausible terrorist-related events. Below, we address this first hypothesis, plus review a list of less plausible conjectures (that nonetheless fall short of full-blown conspiracies).


Pilot suicide is plausible only if two people were involved

Occam’s Razor says use the simplest explanation that fits the facts. On the KISS (keep-it-simple, stupid) principle, pilot Zaharie Ahmad Shah (“ZAS”) would have duped his co-pilot into leaving the cockpit before locking him out, then killed the passengers and crew by depressurizing the plane, flew the plane evasively to elude detection (while having a last look at his home island of Penang), and then put the plane on autopilot until it crashed deep in the Southern Indian Ocean. It’s an explanation that’s abhorrent to many of ZAS’s family, colleagues, and friends. It may have legal liability implications for MAS if his employer had ignored signs that ZAS wasn’t fit to fly. But it more or less explains how the plane ended up so far from anywhere in the SIO and then vanished without a trace.

Once again, “Case closed” is the tempting conclusion. But it’s also wrong. At least until new facts emerge or errors of interpretation were made, two people had to be involved, one to fly a 777 and one to perform complex routines to disable and then re-enable electrical systems. So if the chief pilot Zaharie Ahmad Shah was the main bad guy, then either his co-pilot Fariq Abdul Hamid (“FAH”) or someone else was in on the suicide plan, as one man had to fly while the other toggled switches.

First Officer FAH was just 27 years old, recently married, and a father of three. He was on his maiden unsupervised flight as the second man in the two-man crews that fly the 777. FAH didn’t have a suicidal profile. ZAS was a senior pilot with twenty-six years and an 18,500-hour experience advantage on the young man. I think it’s reasonable that FAH would obey a command to “Go get me a cup of coffee” [so I can lock you out of the cockpit] versus something of a different order of magnitude: “Go into the E/E bay and flip [this particular circuit breaker] and in [20-75] minutes later, reverse everything.” Even deferential new co-pilots could not be relied upon to always do as told by their senior pilot, especially when the steps were complicated and the implications potentially ominous. Jeff Wise emphasized the complexity in a February, 2015 article for New York Magazine:

The only way [to turn the satcom off and on], apart from turning off half the entire electrical system, would be to go into the E/E bay and pull three particular circuit breakers. It is a maneuver that only a sophisticated operator would know how to execute . . .”15

So either FAH was in on the plot from the outset or he wasn’t. He wasn’t duped on the day by ZAS. And if it wasn’t FAH, then it had to be another crew member or a passenger. No one has suggested a pilot suicide scenario that doesn’t envisage ZAS taking the sole lead role. So rather than speculate as to whom the other person(s) might have been, let’s start by assuming ZAS acted alone. If so, here’s what must have happened more or less in order:

  • ZAS told FAH to leave the cockpit on one pretext or another, and locked the door
  • From the cockpit, ZAS turned off the transponder (easy) and the ACARS (harder, if not impossible).
  • Turned the plane around in the dead space over the Malaysia and border to maximize the likely amount of time he’d have before someone came looking for them.
  • ZAS flew the plane high and depressurized it to kill everyone else on board. (Malaysia’s military radar put the plane at 43,000’-45,000’ feet for twenty-three minutes before it dropped down to fly 31,000’-33,000’.) ZAS’s backup cockpit oxygen supply (three hours) would far outlast that of the cabin crew (seventy minutes) or passengers (twelve-to-twenty minutes). Once everyone outside the cockpit was dead from hypoxia (lack of oxygen), then ZAS re-pressurized the cabin and breathed normally without a mask.What is less well-known is that the 777 allows the pilot(s) to control the emergency oxygen supply in the commercial cabin. This gives pilots the option to cut off oxygen if they believe there’s an electrical fire that would be exacerbated if the masks deployed.16 While this fact doesn’t change any outcomes, it shows how much damage a rogue pilot can do even if he doesn’t fly the aircraft into the ground.
  • Once the plane was re-pressurized, ZAS put it on autopilot and either disabled the SDU (supplying the SATCOM) directly, or pulled three circuit breakers in the avionics bay to kill power to the SDU and several other electrical systems (including in-flight entertainment, which later showed up in the Inmarsat ping records).
  • In addition to altitude changes, MH370 also made three distinct turns to the left (west, then south). This path couldn’t have been the result of mechanical malfunction (e.g. cockpit fire which also knocked out the communications), or even pre-programmed once the flight was in the air. Someone who knew how to fly a 777 was at the controls. The plane did not fly navigational beacon-to-beacon as originally reported. It skirted all land, except perhaps the extreme western tip of Sumatra (Banda Aceh). Malaysia PM Najib Razak said that the course flown was a deliberate action.
  • Somewhere among this sequence of turns, ZAS put the plane on autopilot again, and at 02:22 had reversed the previous power-off sequences to begin the three-minute power restoration sequence to the SDU.
  • Back in the cockpit, ZAS resumed control and headed towards Penang, which initially perplexed analysts. However, ex-pilot Simon Hardy used to fly transcontinentally across Australia. Captain Hardy looked at MH370’s circuitous route map, and realized that this was exactly the method he used to fly past Ayer’s Rock17 to give the best view. From the cockpit, ZAS could have looked down upon the lights of Georgetown, the biggest city. ZAS originally was from Penang and this would have been his farewell fly-by.
  • Maybe ZAS decided deep in the dark of night to put the plane on autopilot and end his own life (e.g. by depressurizing the plane once again). But if the 777 had crashed uncontrolled into the sea along the southern arc, then there would have been a large debris field. Certainly one piece of flotsam or wreckage would have been found by searchers after a one-year vigil. So this is less likely than the alternative. More likely, ZAS would have set the plane on autopilot, and then when it ran out of fuel, he managed a controlled glide and changed course sharply to further thwart detection. The 777 in theory could have then ditched at least 118 miles (185km) away from the BTO-predicted path. Despite what were probably rough conditions, ZAS could have landed the plane well-enough that it broke into only a few pieces, minimizing wreckage and quickly sinking before the emergency transmitters had a chance to activate. Thus ZAS’s last acts would have comprised the finishing touches on a monstrous crime.
  • The 25 March 2015 suicide by 28-year old co-pilot Andreas Lubitz which killed 144 people on Germanwings 9525 from Barcelona to Dusseldorf brought pilot suicide more sharply into focus. Farther back in time was the EgyptAir flight 990 incident in 2002 when the co-pilot killed all 217 passengers after the airline punished him for sexual misconduct. Pilots killing themselves and everyone else on board is obviously a topic airlines and pilots are reluctant to discuss either off or on the record.


ZAS’s state of mind is irrelevant: what happened was premeditated, not impulsive.

Even if ZAS possessed almost superhuman organizational skills and adhered to a precise timetable of obfuscation, murder and deceit, alone he couldn’t have done everything ascribed to him. And if he was part of a team, then who was on it? Killing oneself and 238 others on the spur of the moment requires a deeply depressed mental state. Being part of a suicide team speaks to a different mindset. One thing is certain, however. Given the complexity of what transpired that night there’s no way that whatever happened took place on the spur of the moment. It was planned well in advance. So let’s now look for clues as to what ZAS’s may have been feeling, and whether his possible states of mind support a premediated plan to kill all aboard.

  • The New Zealand Herald quoted an anonymous friend of ZAS in a 26 March 2014 article entitled “Pilot in Wrong State of Mind to Fly” – Friend18. The man reported that “Captain Zaharie was ‘terribly upset’ when his wife told him she was leaving . . . .” In addition, “He had been facing serious family problems . . . and relationship problems with another woman he was seeing.” Joking aside, problems with (ex-) wives and (current/ex-) girlfriends have driven many a man to contemplate flamboyant self-destruction as a way out. But it’s still quite a leap to go from taking one’s own life to plotting to murder 200-300 other people at the same time.
  • Recall, too, from the 31 May blog The Facts of the Matter that ZAS was unique among all MH370 crew and passengers in having a completely free personal calendar after 8 March. So if ZAS’s guilty, he’d been planning the exact date for some time.
  • In June 2014 the London Independent reported unnamed investigators saying that they had recovered from ZAS’s home flight simulator deleted “evidence of routes programmed to take a plane far out into the Indian Ocean and practicing landing using a short runway on an island.”19 Standing in complete opposition was the FBI’s comment that they had “found nothing suspicious whatsoever”20 when they examined the recovered flight recorder data. I find it curious that nothing was published as a follow-up to reconcile these contrasting points of view. A pilot friend assures me that many pilots build elaborate at home simulators because they love to fly, and having gone to the trouble and expense then from time-to-time use their simulators to attempt reckless maneuvers such as landing on too-short runways, highways or flying under bridges. That ZAS may have erased such flights of fancy from his system–even ones showing landings on deserted Indian Ocean islands–shouldn’t be taken as evidence of any criminal intent.
  • ZAS was a distant cousin of former Malaysia deputy Prime Minister Anwar Ibrahim and a political supporter. Friday 7 March produced a trial verdict that would end Anwar’s political aspirations (yet again) and send him back to jail for more years as his sodomy acquittal was overturned on appeal by the prosecution. ZAS had appeared in a recent photo wearing a “Democracy is Dead” T-shirt. The press made a big deal of this happening the afternoon before the fateful flight. Some articles purport that ZAS was in court to hear the verdict in person.
  • In August 2014 a New Zealand airline executive Ewan Wilson and a journalist Geoff Taylor teamed up to write ‘Goodnight Malaysian 370’: The Truth Behind The Loss of Flight 370 where the principal thesis is pilot suicide by ZAS. The authors couldn’t find a solid motive for ZAS’s suicide, either. ( The book received good reviews on Amazon.


Is there an explanation that permits ZAS to have acted alone?

No smoking gun exists, but from the above certainly ZAS emerges as a person who may have been planning this act and have been in a poor state of mind. But there simply cannot be only one perpetrator. The power off/ power reset actions alone require a second pair of hands. And even it if didn’t, there’s no logical reason for a suicidal individual (i.e., ZAS) bent on making the plane disappear to restore power to the SDU. It’s neither necessary nor helpful to his aims. And there’s also the inconvenient part about having to be in two places at once, flying the plane one moment and then in the next burrowing below deck in the avionics bay, or stepping over thirty rows of suffocated people to fiddle directly with the SDU back in mid-plane.

But surely with more imagination we can make this all fit? ZAS is barricaded in the cockpit and his co-pilot and crew know that something is very wrong. They start pulling circuit breakers in an attempt to get the armored cockpit door open. (I’m not aware of such an override mechanism, and, if one exists, then it won’t be discussed publicly.) The pilot sees from the changes to his indicators in the cockpit what they’re trying to do, and he depressurizes the plane (as he puts on his own separate oxygen supply) to stop them dead.

If the co-pilot is really smart, he gets into the avionics bay and disconnects the cockpit emergency oxygen supply. He then informs that pilot that there’s no emergency oxygen supply to the cockpit. ZAS doesn’t care: he programs the plane’s autopilot one last time to head to Antarctica and takes MH370 up to 43,000’ as he depressurizes the cabin. The frantic crew pulls switches and circuit breakers until they die. Then at 02:22 someone is still alive and re-boots the SDU, bringing the SATCOM back to life . . . ouch . . . need another theory.

New idea: the same crew member(s) who disabled the power re-enable it when the cabin is depressurized, before they lose consciousness. Maybe this fits within the time parameters available, but then the plane crashes uncontrolled in the Southern Indian Ocean and life jackets, seat cushions and bits of plastic are strewn across the surface where satellites and/or search planes spot them.

*  *  *  *  *

 We can keep going and going here. At the margin, there could be a workable explanation that fits through the Conjecture Filter. But doesn’t it read like the minutes of the Geocentric Astronomy Club of Rome annual general meeting of 1544, the year after Copernicus published On the Revolutions of the Celestial Spheres (proving that the sun didn’t revolve around the earth)?



Recall that we’re examining conspiracies in the 14 June blog entry, One Conspiracy to Rule Them All. The following conjectures have the advantage of not requiring cover-ups or state-sponsorship to work. Unfortunately, they are even worse than pilot suicide at meeting the conditions of the Conjecture Filter.


Mechanical failure/electrical fire left a ghost flight: If there was a fire in the avionics bay, then the outbreak would be some time around 01:21 when the transponder powered down. (The ACARS went off-line a few minutes on either side of 01:21, but it’s impossible to say exactly when.) The pilots didn’t make any distress calls because they were busy following disaster protocol: aviate (fly) first, navigate second and communicate third. The pilot turned the plane around and pointed it west towards the 13,000’ runway on the resort island of Langkawi. (Langkawi is closer than KLIA and less crowded, too). Then smoke overcame the pilots and the plane flew into the SIO and was never seen again. Former pilot and UK-based lawyer (and pilot) James Healy-Pratt described what might have happened:

…it could take 90 seconds for the pilots to lose consciousness, but the plane could continue to fly on using ‘fly-by-wire’, a system that replaces the conventional manual flight controls of an aircraft with an electronic interface.”21

Furthermore, one prolific MH370 commentator maintains that the Rockwell Collins CMU-990 unit controls the transponder, ACARS and VHR voice radio on a Boeing 777. If it fails, allegedly all three systems go out of commission22. In turn, that would make an airborne fire a more likely cause of a crash.

However, I couldn’t find any support on-line from the JIT, Boeing, the IG or other authorities. When I wrote to Jeffrey Wise about this being possible, he replied, “Without knowing the details, sounds ridiculous. The 777 is super-robust with multiple redundancies of every component.”23 The June 2014 ATSB report further refutes the commentator’s assertion: “In the case of MH370, there were multiple redundant communications systems fitted to the aircraft (3 x VHF radios, 2 x HF radios, SATACOM system, 2 x ATC transponders).24 There’s no way Boeing is going to include all those redundant systems and yet have them all rely on the performance of a single component (i.e., the Rockwell Collins CMU-990).

And the fire/ghost flight hypothesis also fails the Conjecture Filter in two vital areas: active pilot control at least as late as 02:25 (and maybe 02:40, or all the way to the end at 08:19), plus the curious power-off/power-on sequence. At the common sense level, no one has explained how a fire or explosion serious enough to disable communications and knock out the crew would also somehow allow the plane to keep flying another six or seven hours. We need another theory.


Punctured fuselage and slow asphyxiation: After 8 March, a Stanford University student quickly published what at the time seemed a sensible explanation for MH370’s demise. He started from a 2013 FAA Airworthiness Directive for the 777 which noted widespread “cracking in the fuselage skin underneath the SATCOM antenna adapter”.25 If this had happened to MH370, then a skin rupture would have explained the failure of the SATCOM system and a slow decompression which left the passengers unconscious (most were probably already asleep anyway), and the crew disoriented and unable to deploy their oxygen masks. The autopilot would have sent the plane north and east where it crashed in the South China Sea.

Depressurization early in the flight was a nice hypothesis, but was swiftly superseded by the Inmarsat ping data, radar tracks describing the flight course reversal, and the power-off/power-on sequence. In any event, the Boeing 777 (registration 9M-MRO) that made the fateful flight didn’t have the design flaw identified by the FAA directive.


A lightning strike shorted out all comms and plane either vaporized or crashed: Weather was clear for hundreds of miles in all directions. The 777 (and all other modern aircraft) is well-shielded from a lightning strike. But assume that a massive ‘bolt out of the blue’ (positive lightning strike) shorted the internal cables and knocked out the ACARS and transponder simultaneously. Jeff Wise’s website from 7 September 2014 has two IG members exchanging views on the lightning hypothesis26:


“Alex Siew”: “A lightning strike which managed to penetrate into the aircraft would have fried all the cables/wires and caused a total electrical failure, resulting, among other things, a broken data path between the AIMS at the front and the SDUs at the back.”

“Gysbreght” replied, noting that the weather report at midnight was benign and wrote, “If a lighning [sic] strike managed to penetrate into the aircraft and ‘fried all the cables/wires’, caused a total electrical failure but the SDU somehow managed to continue functioning on its internal battery, why did it cease functioning sometime between [01:07:48] and [02:03:41], and become alive again with a log-on request at [02:25:27]?”


Alex Siew continued to argue his case for a positive (“bolt out of the blue”) lightning strike in this same discussion thread, with other participants beseeching him to cease his line of enquiry. Why? Because a lightning strike powerful enough to short out all communications wouldn’t have left the plane flying another eight hours. And if the plane was felled by a lightning strike, then all the radar, BTO and BFO data were fabricated, too. And why falsify the data to cover up an accident? So we also rule out a lightning strike, at least from the plausible conjectures list.


None of these conjectures stand up to the “beyond reasonable doubt” standard

Resorting to a legal analogy, I’d say that the balance of probability is that the Inmarsat data are legit, and reinforce what the US and Australia already knew from undisclosed military-sourced data: that MH370 crashed into the Southern Indian Ocean around 08:19 Saturday 8 May. But ‘balance of probability’ is good enough for a conviction only in civil cases, and even here it would take a persuasive litigator to convince jurors to ignore the contrary evidence.

But with 239 souls in limbo, certainly the level of proof must be set a notch higher at ‘beyond all reasonable doubt’ as the case in US criminal law. I don’t see any jury—much less one comprised of informed MH370 followers—finding that the Inmarsat ping-inspired calculations meet this higher standard. If MH370 is really lying at the bottom of the ocean, almost US$100 million has been spent searching for fifteen months without so much as a seat cushion found. And until that happens, a jury in a capital trial would remain unable to reach a verdict. (That is, unless Australia, the UK or US shares more definitive information in respect of radar tracks, acoustical evidence and satellite imagery.)

*  *  *  *  *

On the preceding basis, I vote to acquit Zaharie Ahmad Shah of the charge of mass murder. But if ZAS didn’t do it, who or what did? That’s the subject of the 14 June blog entry, One Conspiracy to Rule Them All.





3 MH370 – Definition of Underwater Search Areas [Australian Transport Safety Bureau, 2014 June (revised 2014 August)], pg. 35


5MH370 – Definition of Underwater Search Areas [Australian Transport Safety Bureau, 2014 June (revised 2014 August)], pg. 28

6MH370 – Definition of Underwater Search Areas [Australian Transport Safety Bureau, 2014 June (revised 2014 August)], pg. 57

7What Happened to Malaysia Airlines Flight 370?”, April 2015.




11What Happened to Malaysia Airlines Flight 370?”, April 2015





16Source: Boeing executive in conversation with the author (March 2014).

17, 16 April 2015





22  see Comments section, Sy Gunson contributions

23Jeff Wise (email to Bradley West, 28 May 2015)

24MH370 – Definition of Underwater Search Areas [Australian Transport Safety Bureau, 2014 June (revised 2014 August)], pg. 34

25 quoting FAA Directive of 26 Sep 2013 at