Risk a Lot to Save a Lot

Transfers

Alan Brunacini, longtime Phoenix fire chief, codified the three-tier risk management principle that became a cornerstone of NFPA 1500 (Standard on Fire Department Occupational Safety, Health, and Wellness Program). The principle is deceptively simple: risk a lot to save saveable lives, risk a little to save saveable property, risk nothing to save what is already lost.

Key structural parallels:

• Risk calibrated to present salvage value — the principle forces the decision-maker to assess what can actually be saved right now, not what was at stake when the incident began. A building that was worth saving twenty minutes ago may now be fully involved, with no one inside. The acceptable risk drops to zero regardless of the building’s value, the insurance, or the political pressure. This transfers to any domain where sunk costs tempt continued investment: a failing product launch, a losing military position, a deteriorating patient. The model insists on reassessing salvage value continuously, not once at the start.
• Explicit three-tier classification — Brunacini did not say “balance risk and reward.” He created three discrete categories with different rules. This is the model’s analytical power: it forces the decision-maker to classify the situation before acting, rather than making a continuous risk-reward judgment under stress. Is there a saveable life? Then aggressive action is warranted. Is there only property? Then defensive action. Is there nothing left to save? Then withdraw. The discrete categories prevent the continuous sliding that stress and institutional momentum produce.
• Anti-sunk-cost mechanism — the most dangerous sentence in firefighting is “we’ve already committed resources.” Brunacini’s model explicitly breaks the sunk-cost linkage: past investment does not justify future risk. A crew that has been fighting a fire for two hours has no more reason to continue than a crew that just arrived, if conditions have deteriorated past the point of salvage. This transfers to venture capital (additional funding rounds for a dying startup), military operations (reinforcing a losing position because of casualties already sustained), and healthcare (aggressive treatment for a patient who has passed the point of benefit).
• “Saveable” as the key qualifier — the word “saveable” does enormous work. It is not “risk a lot to save a lot” — it is “risk a lot to save saveable lives.” The qualifier demands honest assessment: can this person actually be rescued, or are we risking live firefighters to recover a body? The transfer to other domains is in the distinction between optimism and assessment. A project may be “important” without being “saveable.” A patient may be “young” without being “treatable.” The model insists on the adjective that matters.

Limits

• Assessing “saveable” is the hard part — the model tells you what to do once you know whether lives are saveable, property is saveable, or nothing is saveable. It does not tell you how to make that assessment. In a burning building with poor visibility, determining whether occupants are still alive is the most difficult judgment a fireground commander makes. The model provides a decision framework but assumes the hardest input is available.
• Institutional momentum overrides the model — fire departments have strong cultural pressure toward aggressive action. “Risk nothing to save nothing” requires a commander to order withdrawal while cameras are rolling, politicians are watching, and subordinates want to fight. The model is analytically sound but socially costly to apply. The same limit applies in business: telling the board “this product cannot be saved” requires the same kind of moral courage that ordering a withdrawal from a fire does.
• The three tiers are not always distinguishable — in practice, the line between “saveable lives” and “nothing saveable” is not crisp. Is there a 10% chance someone is still alive inside? 5%? 1%? The model’s three discrete categories assume a clarity that ambiguous situations do not provide. The transfer to medicine is direct: “saveable” is a probability, not a binary, and the model provides no guidance on the threshold.
• It assumes unified command — Brunacini’s model works because the incident commander has authority to order withdrawal. In organizations without clear authority structures — committees, partnerships, consensus-driven cultures — no one can declare “nothing is saveable, risk nothing.” The model requires someone with the authority to call the loss, and many organizations deliberately avoid creating that authority.

Expressions

• “Risk a lot to save a lot, risk a little to save a little, risk nothing to save nothing” — the full three-tier formulation
• “Is there a saveable life?” — the first question in the decision tree, determining whether aggressive action is warranted
• “We’re in a defensive posture” — fire service language for the second tier: protecting exposures and saving property, not making interior attack
• “Nothing showing” — the assessment that transitions from offensive to defensive or to withdrawal
• “Everybody goes home” — the companion maxim that sets the constraint: the overriding goal is zero firefighter casualties, and Brunacini’s risk model is the means
• “Don’t trade lives for property” — the second-tier rule stated negatively: no firefighter should die for a building
• “Sunk cost” — the economic concept that Brunacini’s model operationalizes in the field: past investment does not justify future risk

Origin Story

Alan V. Brunacini served as fire chief of the Phoenix Fire Department from 1978 to 2006 and authored Fire Command (1985, revised 2002), one of the most influential texts in American fire service leadership. His risk management principle was incorporated into NFPA 1500, which standardized fire department occupational safety. The three-tier model was a response to a persistent pattern of firefighter fatalities in structures that were already lost — firefighters dying to save buildings that were going to collapse regardless.

The model drew on Brunacini’s observation that firefighters’ willingness to take risks was not calibrated to outcomes. Institutional culture, adrenaline, and the desire to fight rewarded aggressive action regardless of whether anything could be saved. The three-tier model was designed to insert a deliberate assessment step between the arrival at the scene and the commitment of personnel.

References

• Brunacini, A.V. Fire Command (2nd ed., 2002) — the full risk management framework
• NFPA 1500: Standard on Fire Department Occupational Safety, Health, and Wellness Program (2021 edition)
• Gasaway, R.B. Size-Up, Assessment and Decision Making in the Hazard Zone (2019) — cognitive aspects of fireground decision-making
• International Association of Fire Chiefs. “Rules of Engagement for Structural Firefighting” (2009)