How to Make Better Decisions Under Pressure

On January 15, 2009, at 3:27 PM, US Airways Flight 1549 struck a flock of Canada geese at 2,800 feet over the Bronx. Both engines lost thrust simultaneously. Captain Chesley "Sully" Sullenberger had 208 seconds.

Air traffic control offered him two runways — Teterboro in New Jersey and LaGuardia behind him. Either option seemed reasonable. Either option would have killed everyone on board.

The National Transportation Safety Board later ran simulations. When pilots in the simulator were told immediately about the engine failure and allowed to turn back to LaGuardia with zero delay, they landed safely. But when the simulation added a thirty-five-second delay — a realistic estimate for the time a human brain needs to assess the situation, process the options, and commit to a course of action — the planes crashed short of the runway or into buildings. Every time.

Sully didn't return to LaGuardia. In those 208 seconds, he vetoed the option that looked correct, identified the Hudson River as the only viable landing site, and put the aircraft on the water. All 155 people survived.

What happened inside Sully's head during those seconds is the subject of this post — not because you'll ever land a plane on a river, but because the neural mechanisms that allowed him to override a bad option under extreme time pressure are the same ones available to you in every high-stakes decision you face. And unlike landing a plane, they can be practiced.

The Veto Window

Sully's brain generated the same initial impulse as every other pilot in the simulation: go back to LaGuardia. That impulse was automatic, pattern-matched from thousands of hours of training. The difference was in what happened next. Sully's prefrontal cortex — the part of the brain responsible for executive override — vetoed the impulse before it became action. Not because he had better instincts. Because he ran the computation fast enough to see that the obvious answer was wrong, and he had the neurological capacity to override it in the window between impulse and action.

This window exists in every decision. The email you almost sent. The price you almost quoted. The hire you almost approved. The impulse arrives — the brain has already generated a response based on pattern matching. The question is whether your prefrontal cortex has the resources and the practice to evaluate that response before it becomes irreversible.

Two things close the veto window: time pressure and hardware degradation. When you're sleep-deprived, the prefrontal cortex loses connectivity with the emotional systems it's supposed to regulate — the same hardware degradation that makes self-regulation collapse under stress. When you're stressed, cortisol narrows attentional focus, making the first available option feel more compelling than alternatives — the same escalation trap that makes sunk cost decisions so reliably wrong under pressure. When you're in a meeting where everyone is waiting for your answer, the social pressure compresses the window further. The impulse is the same. The ability to veto it shrinks.

The Emotional Downshift

In 2007, Matthew Lieberman and his colleagues at UCLA published the fMRI study that made this mechanism visible. Participants viewed images of faces expressing strong emotions. When they simply looked at the faces, the amygdala — the brain's threat-detection center — activated as expected. When they were asked to label the emotion — to say "angry" or "afraid" — amygdala activation diminished. The right ventrolateral prefrontal cortex activated simultaneously, and its activation was inversely correlated with the amygdala. The more the labeling circuit engaged, the more the threat circuit quieted.

This isn't a metaphor for "talking about your feelings." It's a measurable neural mechanism. The act of labeling recruits the prefrontal cortex into a process that was, a moment ago, running entirely on automatic threat detection. The label doesn't eliminate the emotion. It changes the brain's relationship to it, creating a gap between the signal and the response.

Katharina Kircanski and colleagues extended this to real behavior in 2012. Participants with spider phobia were divided into groups. One group was asked to label their emotions while approaching a live tarantula — "I feel anxious, my heart is racing, I'm afraid the spider might move." One week later, the labeling group showed lower skin conductance responses and approached the spider more closely than groups that used distraction, reappraisal, or no strategy at all.

Chris Voss, the FBI's former lead international kidnapping negotiator, built his entire negotiation methodology around this mechanism. During the 1993 Chase Manhattan Bank robbery, Voss used mirroring — repeating the last few words of the hostage-taker's statements back to him — combined with what he would later call "tactical empathy." The technique didn't involve agreeing, sympathizing, or problem-solving. It involved reflecting what the other person was experiencing. All hostages were released. Voss later wrote that labeling — naming the other person's emotion explicitly — is the single most effective tool in high-stakes negotiation, because it activates exactly the circuit Lieberman's study identified: it recruits the prefrontal cortex into a threat response that would otherwise escalate unchecked.

The Reappraisal Switch

In 2014, Alison Wood Brooks at Harvard ran three experiments that tested the simplest possible version of this intervention. Before a stressful performance — karaoke singing, public speaking, and a math test — participants were randomly assigned to say either "I am anxious," "I am excited," or "I am calm."

The "excited" group outperformed both other groups on every measure. They sang better (rated by the karaoke software's pitch-tracking algorithm). Their speeches were rated as more persuasive and more competent by independent judges. Their math scores were higher. Saying "I am calm" — the advice most people give themselves before a stressful event — performed worst.

The mechanism is arousal reappraisal. Anxiety and excitement share the same physiological signature — elevated heart rate, heightened alertness, increased cortisol. Telling yourself to calm down requires the brain to suppress an arousal state it's already committed to, which is expensive and usually fails. Reframing the arousal as excitement keeps the physiological activation but changes the cognitive interpretation from "threat" to "opportunity." The body stays ready. The brain stops fighting itself.

Before your next investor pitch, board meeting, or high-stakes negotiation: say "I'm excited" out loud. Not as affirmation. As reappraisal. You're not lying to yourself. You're relabeling a physiological state that's already present, and the relabeling changes how the prefrontal cortex processes what comes next.

The Pre-Mortem

Gary Klein formalized the pre-mortem technique in 2007, drawing on research by Mitchell, Russo, and Pennington showing that "prospective hindsight" — imagining an event has already occurred — increases the ability to generate explanations for it by thirty percent compared to imagining it might occur. The difference is grammatical but neurologically significant: "the product launch failed" activates different retrieval pathways than "the product launch might fail." The past tense makes the scenario feel real, which makes the brain search harder for explanations.

The standard version of risk assessment asks "What could go wrong?" The pre-mortem asks "What did go wrong?" and works backward. The shift from hypothetical to retrospective is the same shift that makes implementation intentions more effective than goals — it converts an abstract possibility into a specific scenario the brain treats as concrete. Pair the pre-mortem with a checklist for the decision process itself and you've built a systematic guard against both cognitive bias and procedural drift.

The pre-mortem is especially valuable for decisions where the team has already reached consensus, because it gives people permission to voice concerns they've been suppressing. "What could go wrong?" invites optimism bias. "The launch failed — why?" invites honesty. It's a structural workaround for the Dissent Tax that silences teams: instead of asking people to disagree with the plan, you're asking them to explain an outcome. The social cost is lower, and the information yield is higher.