Plan Continuation Bias: Why Good Pilots Fly Unstable Approaches

Hey everyone and welcome back to this A320 Knowledge exclusive article.

Here is one of the most uncomfortable statistics in commercial aviation: industry research has repeatedly found that only around 3% of unstable approaches end in a go-around. Read that again. When the aircraft arrives at the gate too fast, too high, or not configured, crews elect to continue roughly 97 times out of 100. And these are not weak pilots, or lazy pilots, or cowboys. They are, overwhelmingly, good pilots: experienced, well-trained professionals who briefed the stabilisation criteria an hour earlier and genuinely believed they would honour them.

So what happens between the briefing and the gate? The answer is a quiet, powerful piece of human psychology called plan continuation bias, and it deserves a place in your mental toolkit right alongside the memory items. Because the data tells us something humbling: the go-around is not primarily a flying problem. It's a deciding problem. In this article we'll unpack why the approach phase is a near-perfect trap for the human brain, the three lies we tell ourselves on the way down, and, most importantly, how to structure your operation so the right decision makes itself.

Let's dive right in.

---

---

Subscribe to the A320 Knowledge YouTube Channel!

---

What Is Plan Continuation Bias?

Plan continuation bias is the deep-seated human tendency to continue with an original course of action even when changing conditions mean it's no longer the best option. Researchers studying flight crew decision-making have described it as an unconscious pull towards the finish line, and they've identified its most dangerous property: it strengthens as you near completion of the plan.

At the top of descent, abandoning the plan costs you almost nothing, and crews divert, hold, or slow down without a second thought. At 500 ft, abandoning the plan means throwing away an entire approach with the runway filling the windscreen. Same decision, wildly different psychological price. The brain, quite literally, does not want to let go of a nearly-finished task.

You already know this bias by its street names: get-there-itis, press-on-itis. And it never travels alone. It brings friends:

• Sunk cost: "We've come all this way, flown the whole arrival, we can't waste it now." The effort already invested feels like a reason to continue. It isn't. It's already spent either way.
• Confirmation bias: once the brain has decided it wants to land, it starts cherry-picking. The speed trend just starting to reduce gets noticed; the altitude still diverging gets explained away.
• Expectation bias: you briefed a landing, you've flown a thousand landings, and so a landing is what your brain expects to happen next. The go-around is, statistically, an exotic event: many pilots fly one for real only once every year or two.

None of this is a character flaw. It's the factory-standard wiring of every human brain, including the sharpest captain you've ever flown with. In fact, experience can deepen the trap: the more approaches you've personally salvaged, the more evidence your brain holds that this one is salvageable too.

---

Why the Approach Is the Perfect Trap

If you were to design a scenario to maximise plan continuation bias, you would design final approach.

The workload curve and the decision curve cross. As you descend, your task load rises: configuration, checklists, ATC, traffic, weather. At exactly the same time, your spare mental capacity for stepping back and evaluating the big picture falls. The moment the aircraft most needs a cold, detached decision is the moment you're least equipped to make one. It's the same cruel arithmetic we saw in the startle effect: the situation demands your best thinking precisely when your brain has the least to give.

The cues arrive incrementally. An unstable approach almost never announces itself with a single dramatic failure. It assembles itself out of small, individually reasonable deviations: five knots fast here, half a dot high there, flap 3 a little late, each one minor, each one explainable, each one nudging the baseline. By the time the picture is undeniably wrong, you've already accepted every piece of it separately. It's the boiling frog, flown at 140 kt.

Landing is the default; going around is a decision. This is the asymmetry at the heart of the 3% statistic. Continuing the approach requires no decision at all: the aircraft is already doing it, and silence continues it. The go-around, by contrast, demands that somebody actively intervene, speak, push the levers forward and dismantle the plan. Under pressure, humans default to the option that requires no action. The system, left alone, lands.

And the gallery is watching. Fuel, slot times, a queue of traffic behind you, passengers with connections, and, if we're honest, your own self-image as someone who doesn't need two attempts. None of these belong in the decision. All of them are in the flight deck anyway.

---

The Three Lies of an Unstable Approach

Listen carefully on the flight deck during a deteriorating approach and you'll hear the bias speak. It uses the same three sentences every time. Learn them, because hearing one out loud, from your colleague or from yourself, is the single most reliable warning sign this trap has sprung:

Lie #1: "It's coming back." The energy is high but the trend is (slightly) improving, and the brain seizes on the derivative rather than the value. Sometimes it genuinely is coming back. The question the lie skips is: will it be back by the gate, or merely heading in the right direction as you cross it?

Lie #2: "We'll be stable by five hundred." This is bargaining, and it's fatal to the entire concept of a gate. The stabilisation height isn't a suggestion of where stability would be nice; it's the pre-agreed point where the decision has already been made for you. The moment you catch yourself renegotiating the gate on the way down, the bias is flying the aircraft.

Lie #3: "I've landed worse than this." Probably true. Also completely irrelevant. Every runway excursion in history was preceded by a career of approaches that worked out. Your past saves are not evidence of safety; they're evidence of exposure.

🚨 Real-World Example: "We were high and fast from an ATC shortcut, and I remember saying all three, almost word for word. 'It's coming back... we'll have it by five hundred... it's fine, I've had worse into here.' We touched down long on a wet runway and stopped with less margin than I ever want to see again. In the debrief my FO admitted he'd wanted to call it at a thousand feet but I sounded so confident. That sentence has stayed with me longer than the landing has."

---

Making the Go-Around the Default

You cannot brief away a cognitive bias, and you cannot out-discipline your own wiring in the moment. What you can do is restructure the decision so the bias never gets a vote. Three moves accomplish this:

1. Treat the gate as a decision already made. The entire value of stabilised approach criteria, typically 1,000 ft, with some operators allowing 500 ft in visual conditions, is that they convert a judgement call into a tripwire. Not stable at the gate means go around: not "assess", not "discuss", not "nearly". The decision was made in a calm room, by rested people, with the full weight of industry accident data behind it. Your only job at the gate is not to overrule them with 30 seconds of adrenaline and a windscreen full of runway.

2. Pre-commit, out loud, in the briefing. As we covered in Effective Leading From the Flight Deck, hard limits belong in the briefing: "If we're not configured by 5 miles, say something." Go one step further and pre-commit the outcome, not just the callout: "If either of us calls unstable at the gate, we go around. No discussion." Psychologists call this an implementation intention, and it works because the decision is made before the bias switches on. On the day, nobody decides anything. They just execute what was already agreed.

3. Reframe what a go-around is. A go-around is not a failed approach. It is the system working exactly as designed: the safety net catching what it was built to catch. Airbus's own guidance urges crews to be go-around minded on every approach, and the healthiest framing I know is this: every approach is a go-around, with an option to land if the criteria are met. Fly it that way and the go-around stops being an exceptional, face-losing event and becomes the pre-loaded default, with the landing as the earned outcome.

---

The Pilot Monitoring: The Circuit Breaker

The 3% statistic hides an important detail: in a large share of continued unstable approaches, one pilot knew. The information was on the flight deck; it just never became a decision. That is a monitoring and assertiveness failure, not an awareness failure, and it's why the PM role is the circuit breaker in this entire system.

✔ Call the parameter, not an opinion. "Speed plus fifteen" is a fact that demands a response. "Bit fast, isn't it?" is an invitation to bargain. Standard deviation callouts exist precisely because they're bias-proof: there's no polite way to argue with a number.

✔ Use the agreed escalation. Deviation call, corrective call, then the decision call: "unstable, go around." If the pre-commitment was briefed properly, that final call isn't a challenge to the PF's authority; it's the execution of a joint decision made an hour earlier.

✔ Flatten the gradient in advance. The junior FO who stayed silent didn't lack courage on the day; they lacked permission granted before the day. If you're the captain, the sentence "if you call it, we fly it" costs you five seconds in the briefing and hands your colleague the circuit breaker with your fingerprints already on it.

💬 First Officer's Reflection: "The best captain I ever flew with briefed the same line every single sector: 'Any go-around call from you is my go-around.' Months later I finally had to use it, ten knots fast and half a dot high at the gate. TOGA, no discussion, not a flicker of annoyance. In the debrief he thanked me. I'd have followed that man anywhere, and I've briefed his line on every sector since I got my own command."

---

After the Approach: Close the Loop

One last defence, and it's the one that protects your next approach. If you continued when you shouldn't have, and every honest pilot eventually does, the worst possible response is to let it quietly become normal. Unstable-but-it-worked-out is how deviance gets normalised, one uneventful landing at a time.

So close the loop. Debrief it factually, the way we covered in Mastering Communication and Conflict: what did we see, when did we see it, what stopped us acting? File the report your operator's just culture exists for; every unstable approach report is a data point that makes the case for the next crew's go-around. And privately, notice which of the three lies you told. That's not self-flagellation. That's calibration, and it's what separates experience from mere exposure.

---

Key Takeaways

• Plan continuation bias strengthens as the plan nears completion, which makes final approach the most psychologically loaded decision point in the entire flight, and continuing the path of least resistance.
• Learn the three lies: "it's coming back", "we'll be stable by five hundred", and "I've landed worse". Hearing any of them, from your colleague or yourself, is the trap announcing itself.
• Move the decision out of the moment: treat the gate as a decision already made, pre-commit the go-around in the briefing, and fly every approach as a go-around with an option to land.
• Empower the circuit breaker: factual deviation callouts, a briefed escalation, and a no-fault response to the go-around call turn the PM into the strongest defence the system has.

---

Subscribe to the A320 Knowledge YouTube Channel!

---

If you found this exclusive article helpful, consider sharing it with others or leaving a comment below with your thoughts.

I'll see you in the next one.

A320 Knowledge