The core model is the Recognition-Primed Decision (RPD) model, developed by Gary Klein around 1985: experts match a situation to patterns from experience, recognize it as typical, retrieve a course of action that worked before, and mentally simulate it before acting. Klein, Calderwood and Clinton-Cirocco's fireground study interviewed 26 experienced commanders about 156 decision points and found that in fewer than 12 percent of cases did they compare two or more options. In a famous 2009 adversarial collaboration, Daniel Kahneman and Gary Klein concluded that expert intuition can be trusted only in high-validity environments with stable, learnable cues and adequate feedback — conditions met by firefighting, chess, and anesthesiology, but not by long-term political or stock-market forecasting. Naturalistic decision making suits fast, recognizable, single-expert situations; deliberate, high-stakes, novel, multi-stakeholder decisions instead need structure. Argumentree supports those deliberate decisions with structured pro/con argument trees, multi-dimensional rating that aggregates into consensus scores, AI extraction of arguments, and a full audit trail that makes expert reasoning inspectable rather than locked in one person's head.

How do experts decide in a burning building, with seconds to spare? Not by weighing options — by recognizing a pattern and acting. That's naturalistic decision making.
Naturalistic decision making (NDM) studies how experts decide in the messy real world — under time pressure and uncertainty — and finds they recognize rather than compare. Gary Klein's Recognition-Primed Decision model is the centerpiece. The catch: this expert intuition is only reliable in predictable domains with good feedback. Outside those, you need deliberate structure — which is where tools like Argumentree come in.
In one of Gary Klein's most-told cases (Sources of Power, 1998), a lieutenant led his crew into what looked like a routine kitchen fire. But the water had no effect, the room was far hotter than a kitchen fire should be — and strangely quiet. Uneasy and unable to say why, he ordered everyone out. Moments later the living-room floor collapsed: the real fire was raging in a basement nobody knew was there.
The lieutenant first credited a "sixth sense." Klein's interview revealed the truth — the fire's behavior violated his mental model, and his expert pattern-recognition flagged the mismatch before he could consciously explain it. That is recognition-primed decision making in action.
Developed by Gary Klein around 1985, RPD describes the expert's fast path:
Match the situation to patterns from years of experience and see it as typical.
Recall a course of action that has worked in similar situations before.
Mentally play it out — 'would this work here?' — before committing.
If the simulation looks good, act. If not, tweak it or try the next option — one at a time, not in parallel.
Klein's team studied 26 fireground commanders (~23 years' experience) across 156 real decision points. In fewer than 12% did anyone compare two or more options. Experts recognize; they don't deliberate. It's the empirical rebuttal to the classical rational model.
Skeptic Daniel Kahneman and intuition champion Gary Klein spent years in an "adversarial collaboration" and published a joint answer in 2009. Their verdict: intuition is trustworthy only when both conditions hold.
Stable, learnable cues and regularities — not chaos. Firefighting, chess, anesthesiology qualify.
Enough repetition with rapid, accurate feedback to actually learn the cues.
Where those fail — long-term forecasting, stock picking, novel one-off strategy — confident intuition is the illusion of validity, and you need deliberate, structured reasoning instead.
NDM is for the lone expert acting fast in a domain they know cold. Most organizational decisions are the opposite: novel, high-stakes, irreversible, and shared across many people — exactly the conditions where intuition is least reliable and structure matters most — the realm of deliberate decision making and collaborative decision making. Argumentree is built for those, on argument mapping:
Turn an expert's gut call into explicit pro/con arguments others can examine — so "I just know" becomes reasoning the group can test.
When the environment is too unpredictable for reliable intuition, rate and weigh arguments to ground the decision in evidence, not confidence.
AI extraction pulls experts' reasoning out of meetings and documents, so hard-won pattern knowledge is recorded, not lost when they leave the room.
Every decision keeps a record of the reasoning behind it — reviewable later against how it actually turned out.
Confident intuitions formed in unpredictable domains feel just as real but aren't reliable.
Recognition fails when a situation only superficially resembles a familiar one.
Sticking with an initial read despite cues that contradict it — the opposite of the lieutenant who broke fixation.
Naturalistic decision making (NDM) studies how experienced people actually make decisions in real-world conditions — under time pressure, uncertainty, high stakes, and incomplete information — rather than in artificial lab choices. Its signature finding is that experts rarely compare options; they recognize a situation as familiar and act on the first workable course of action.
RPD, developed by research psychologist Gary Klein around 1985, describes how experts decide fast: they match the current situation to patterns from experience, recognize it as typical, retrieve a course of action that has worked before, and mentally simulate it. If the simulation looks good, they act; if not, they adjust or try the next option. It is serial — one option at a time — not a parallel comparison.
Klein, Calderwood and Clinton-Cirocco interviewed 26 experienced fireground commanders (about 23 years of experience on average) about 156 real, non-routine decision points. In fewer than 12% of cases was there any evidence of comparing two or more options. Commanders overwhelmingly recognized the situation as typical and went straight to a suitable action — directly contradicting the classical 'weigh every alternative' model.
In a famous 'adversarial collaboration,' skeptic Daniel Kahneman and intuition champion Gary Klein agreed (2009) that expert intuition is trustworthy only under two conditions: (1) a high-validity environment with stable, learnable cues, and (2) enough practice with rapid, accurate feedback to learn those cues. Firefighting, chess, and anesthesiology qualify; long-term political and stock-market forecasting do not — which is why those 'expert' intuitions so often fail.
No. Expert intuition (NDM) is fast but built on years of pattern recognition with feedback — it looks instant but is deeply informed. Impulsive decision making is fast without that foundation: acting on emotion or the first urge with no recognition of a learned pattern. The difference is whether the speed rests on genuine expertise in a predictable domain or on none at all.
When a call is novel, high-stakes, and shared, structure beats gut. Make your team's reasoning explicit and decide with confidence on Argumentree.
Start Free