United 737 Runway Excursion Exposes Pilot Protocol Gaps

A United Boeing 737 MAX slid off a wet Houston runway after the captain skipped safety briefings and lowered autobrakes. What this reveals about airline safety culture.

A captain who preferred a gentle stop over following the book. A first officer who saw a wet runway while his colleague saw a dry one. An air traffic controller pressing for speed on a surface already flagged for reduced braking. United Flight 2477's runway excursion at Houston's George Bush Intercontinental Airport on March 8, 2024, did not involve a mechanical failure or a freak weather event. It involved a chain of human choices, each individually defensible, collectively disastrous. The left main landing gear collapsed and separated. The aircraft slid into the grass. All 166 souls on board walked away, but the NTSB findings released in late 2025 paint a picture that should concern every frequent flier: the safety net held, but several layers of it had already been cut away before the wheels even touched the ground.

The Anatomy of a Skipped Checklist

United's Standard Operating Procedures require pilots to conduct a threat forward briefing during the top of descent phase. This briefing exists for a specific reason: it forces the crew to verbalize known hazards before they become time-critical problems. Captain Alireza Johartchi did not perform one. He also did not print the landing performance data, another procedural step designed to give crews hard numbers on stopping distances for the conditions they are actually facing.

These are not obscure regulations buried in supplementary manuals. They are core operational procedures that every line pilot at a major carrier is expected to execute on every flight. Skipping them does not mean the pilot is incompetent. It means the pilot has developed a personal risk calculus that diverges from the airline's published risk calculus. That divergence is where accidents live.

The Houston ATIS was broadcasting runway condition codes of 3/3/3, indicating a surface described as "slippery when wet" with noticeably reduced braking deceleration. First Officer Michael Dickson recalled the runway appearing wet. Captain Johartchi recalled it as dry. Two pilots in the same cockpit, looking at the same runway, arriving at opposite conclusions about surface conditions. Without a threat briefing to force that discrepancy into the open, each pilot operated on his own mental model. The captain, as pilot flying, was the one whose model determined the autobrake setting.

He set it to 2, then lowered it to 1. In post-incident interviews, Johartchi told investigators he preferred a lower braking setting, favoring a gradual deceleration using available runway length over a more aggressive stop. This is a preference, not a procedure. United's wet runway operations call for specific deceleration techniques precisely because pilot preferences are unreliable when conditions are ambiguous.

The ATC Pressure Problem Nobody Wants to Solve

Houston Tower had traffic stacking up. United 2477 was following an Embraer ERJ-145, with another Boeing 737 on approach behind it. The controller cleared 2477 to roll to the end of Runway 27, then added four words that changed the landing dynamics: "Keep your speed up."

Controllers issue speed instructions constantly. They are a fundamental tool for maintaining separation and throughput at busy hub airports. But those instructions create an implicit pressure environment that interacts with pilot decision-making in ways the system does not adequately account for. A pilot who has already decided the runway is dry and has already lowered the autobrake setting now receives external validation for carrying extra speed. Each link in the chain reinforces the previous one.

This is not a Houston problem. It is a systemic tension baked into the architecture of hub operations at every major airport in the National Airspace System. Airlines need high throughput to make hub economics work. Controllers need to sequence arrivals tightly to deliver that throughput. Pilots absorb the resulting pressure, sometimes consciously, sometimes not. The FAA has studied this dynamic for decades without resolving it, because resolving it would mean accepting lower landing rates at the busiest airports in the country.

IATA data shows that runway excursions accounted for 21% of all aviation accidents over the 2015 to 2024 decade, with 104 recorded events and landing operations responsible for 80% of them. The rate climbed in 2024 to its highest point since 2020. Lateral excursions, the category that describes what happened in Houston, represent 52% of all runway excursions. These are not exotic failure modes. They are the single most common category of aviation accident, and the contributing factors repeat with depressing regularity: unstabilized approaches, excessive touchdown speed, failure to go around, and inadequate use of available braking systems.

United's Safety Culture Under the Microscope

The cockpit voice recorder captured the immediate aftermath. "Oh my god, I can't believe that happened. God darn it. I am so sorry," Johartchi said. Then a more telling exchange: a union representative contacted during the post-incident communications reportedly advised the captain to stop talking. "Don't say another word." That is standard legal counsel, not evidence of a cover-up. But the speed with which the institutional machinery shifted from safety investigation to liability management illustrates a tension that exists at every major carrier.

United has faced heightened scrutiny since a series of operational incidents across 2023 and 2024 drew FAA attention. The airline operates the largest 737 MAX fleet in the world, with over 350 aircraft in service or on order. Scale of that magnitude means statistical exposure to events like Houston is higher simply by volume. But it also means that procedural compliance gaps, if they exist systemically rather than in isolated cases, propagate across thousands of daily flights.

The question the NTSB investigation ultimately needs to answer is whether Johartchi's decision to skip the threat briefing and override the autobrake setting reflects an individual deviation or a cultural norm. Airlines conduct Line Operations Safety Audits, known as LOSA studies, specifically to measure the gap between published procedures and actual line operations. The results are almost always sobering. Pilots routinely deviate from SOPs in minor ways that individually carry negligible risk but collectively erode the safety margins the procedures were designed to protect.

United's competitors face identical challenges. Delta, American, and Southwest have all experienced runway excursions with similar causal chains. The IATA Operational Safety Audit program data shows that IOSA-registered carriers have an all-accident rate of 0.98 per million flights versus 2.55 for non-IOSA carriers, demonstrating that structured safety programs work in aggregate. But aggregate statistics offer no comfort when your specific flight is the one sliding off the pavement.

The Contrarian Take: The System Worked

Here is the uncomfortable truth that gets lost in the coverage. Every person on United 2477 walked off that aircraft. The landing gear collapsed as designed, absorbing energy and preventing a far worse outcome. The runway safety area beyond the pavement existed precisely for this scenario. Emergency response was immediate. The evacuation was orderly.

Aviation safety is not built on the assumption that humans will always follow procedures perfectly. It is built on the assumption that they will not, and that the system needs enough redundancy to catch failures when they cascade. Houston was a case study in Swiss cheese alignment: the threat briefing hole, the wet runway perception hole, the reduced autobrake hole, and the ATC speed pressure hole all lined up simultaneously. But the physical infrastructure and aircraft design held.

That is not an argument for complacency. It is an argument for understanding where the real risk reduction comes from. Procedure compliance campaigns have diminishing returns because human attention is finite and cockpit workload is already high. The more productive investment is in systems that compensate for predictable human lapses: enhanced ground proximity warnings tuned to runway remaining distance, automatic braking systems that reference actual surface conditions rather than pilot-selected settings, and runway status lights that give crews real-time friction data independent of ATIS reports that may be stale by the time the aircraft touches down.

Boeing has been developing enhanced autobrake logic for the 737 MAX family that incorporates real-time deceleration monitoring, but fleet-wide retrofit timelines remain unclear. Honeywell's SmartRunway system, which provides automated callouts for excessive speed and insufficient runway remaining, is available but not mandated. The technology to prevent Houston-type events exists. The regulatory and economic will to mandate it fleet-wide does not, at least not yet.

What This Means for Travelers

Passengers have no control over whether their captain conducts a threat briefing or what autobrake setting gets selected. That is by design. Commercial aviation's safety record, with a fatal accident rate below one per five million flights for IATA carriers, is built on layers of institutional oversight that individual passengers cannot and should not need to manage.

But the Houston excursion offers a useful lens for understanding how to think about airline safety as a consumer. The relevant question is not whether an airline has ever had an incident. Every major carrier has. The relevant question is what the airline does afterward. Does it mandate additional simulator training focused on the failure mode? Does it update its SOPs? Does it participate in voluntary safety reporting programs that allow line pilots to flag procedural pressure without fear of discipline?

United has publicly committed to cooperating fully with the NTSB investigation and reviewing its wet runway procedures. Whether that review produces meaningful changes or becomes another compliance exercise will only become visible over time, through LOSA data and recurrent training syllabi that passengers never see.

For travelers flying through Houston or any high-traffic hub during wet weather, the practical takeaway is this: the system is remarkably safe, but it is safe because of redundancy, not perfection. Every layer matters. When a crew skips one, the remaining layers have to hold. In Houston, they did. The next time the cheese slices align, they might not. That is not a reason to avoid flying. It is a reason to insist that airlines, regulators, and manufacturers keep investing in the layers that catch us when the human ones fail.