Base Actions on Current and Expected Fire Behavior
mental-model established
Source: Fire Safety
Categories: decision-makingrisk-management
Transfers
Standard Fire Order #3 — “Base all actions on current and expected fire behavior” — is one of the Ten Standard Fire Orders that govern US wildland firefighting operations. It encodes a deceptively simple discipline: do not act on what the fire was doing when you last checked, and do not act on what you wish it were doing. Act on what it IS doing right now and what it WILL do based on identifiable factors.
Key structural parallels:
-
Current vs. expected as distinct cognitive tasks — the order separates two questions that untrained decision-makers habitually merge. “Current behavior” requires direct observation: where is the fire, how fast is it moving, what is it consuming? “Expected behavior” requires prediction: given wind forecasts, terrain ahead, fuel type, and time of day, what will the fire do in the next hours? A fire that is currently creeping slowly through damp ground-level fuel may be 30 minutes away from reaching a dry slope where it will crown and run. A decision based only on current behavior deploys resources for a creeping fire; a decision based on expected behavior prepares for a running crown fire. In incident response, the parallel is distinguishing between current system symptoms and projected cascade effects.
-
Prediction must be grounded and named — firefighters do not predict fire behavior through intuition alone. They use specific inputs: the Haines Index (atmospheric stability), fuel moisture readings, wind speed and direction, slope percentage, time of day (diurnal wind patterns). The order requires that the prediction be based on identifiable, communicable factors. This makes the expectation auditable: after an incident, reviewers can ask “what factors did you base your expected behavior on?” and evaluate whether the prediction was reasonable. In software engineering, this maps to requiring that capacity planning decisions cite specific metrics (request rate trends, resource utilization curves) rather than vague appeals to “growth.”
-
Action must track changing reality — the word “base” is critical. Actions are not decided once and then executed; they are continuously re-based on updated observation and prediction. If the wind shifts, the expected behavior changes, and actions must change with it. This creates a discipline of continuous recalibration that resists the sunk-cost trap: resources committed to a plan based on yesterday’s fire behavior must be redeployed when today’s behavior diverges. In project management, this maps to re-evaluating a roadmap when market conditions change rather than completing a plan because it was already approved.
-
The order is a decision filter, not a decision rule — it tells you what to base your decision on, not what decision to make. Two commanders observing the same fire behavior may make different tactical choices depending on available resources, crew fatigue, and values at risk. The order constrains the input, not the output. This transfers to any domain where decision frameworks need to standardize the evidence basis without prescribing the conclusion.
Limits
-
Prediction tractability does not transfer — fire behavior is governed by physics. Given fuel type, moisture content, wind speed, slope angle, and atmospheric stability, an experienced firefighter can predict fire movement with useful accuracy over a span of hours. This tractability is a feature of the physical domain, not of the mental model. In organizational contexts, “expected behavior” of a market, a competitor, or a complex software system is not derivable from a small set of observable inputs. Importing the fire order’s confidence in prediction leads to overconfident forecasting in domains where the equivalent physics do not exist.
-
Two inputs are not enough — the fire order works because in wildland fire, the decision-relevant factors genuinely reduce to fire behavior. Other factors (crew morale, resource availability, political pressure) are handled by other orders and protocols. In most organizational decisions, current state and predicted trajectory are necessary but radically insufficient inputs. Strategy must also weigh resource constraints, stakeholder alignment, regulatory environment, and opportunity costs that have no analog in fire behavior analysis.
-
Temporal horizon mismatch — fire behavior prediction is useful over hours to a day. The order’s rhythm of “observe, predict, act, re-observe” maps onto fast tactical decisions. Many organizational decisions require prediction over months or years, where the compounding of uncertainty makes “expected behavior” more aspiration than forecast. Applying the fire order’s confident predictive stance to long-horizon decisions imports a false precision.
-
The order assumes observable current state — firefighters can see the fire. They have direct sensory access to current behavior. In many organizational contexts, “current state” is itself uncertain: the team does not know how many users are affected, how much revenue is at risk, or what the competitor is actually doing. The order’s clean separation of current and expected behavior assumes that current behavior is known, which is often the harder problem in non-fire domains.
Expressions
- “What is the fire doing? What will it do?” — the canonical two-question formulation, used in fire briefings
- “Base your actions on current and expected behavior” — the full order, invoked in training and after-action reviews
- “The fire doesn’t care about your plan” — the folk corollary, emphasizing that plans must yield to observed reality
- “Are we planning for the fire we have or the fire we’ll have in an hour?” — the predictive challenge, applied in incident management
- “What’s the expected behavior of this system under load?” — software engineering transfer of the two-question discipline
Origin Story
The Ten Standard Fire Orders were established in 1957 by a task force convened after the Inaja fire in California, which killed 11 firefighters. The task force, led by the US Forest Service, studied military general orders as a structural model and drafted ten orders intended to prevent the specific failure modes identified in fatality investigations. Order #3 — “Base all actions on current and expected fire behavior” — addressed the recurring finding that crews had acted on outdated or wishful assessments of fire conditions.
The orders were refined after subsequent tragedies, including the 1994 South Canyon fire (14 fatalities) and the 2013 Yarnell Hill fire (19 fatalities). In both cases, investigations found that crews had failed to update their assessment of fire behavior as conditions changed — acting on the fire they had briefed, not the fire they faced. The order’s emphasis on “current AND expected” — not one or the other — reflects this hard-won lesson about the gap between briefed conditions and actual conditions.
References
- National Wildfire Coordinating Group, Incident Response Pocket Guide (PMS 461), current edition — contains the Ten Standard Fire Orders
- “Report of Task Force on Study of Fatal and Near-Fatal Fires,” USDA Forest Service, 1957 — origin of the Ten Standard Fire Orders
- Maclean, N. Young Men and Fire (1992) — Mann Gulch disaster, foundational text on fire decision-making
- Marsh, R. “Yarnell Hill Fire Serious Accident Investigation Report,” Arizona State Forestry Division, 2013 — case study of failure to re-assess expected fire behavior
Related Entries
Structural Neighbors
Entries from different domains that share structural shape. Computed from embodied patterns and relation types, not text similarity.
- Choice Point (navigation/mental-model)
- Ticket Rail (food-and-cooking/metaphor)
- Observe and Interact (/mental-model)
- Attachment Styles (folk-taxonomy/mental-model)
- Internal Working Model (manufacturing/metaphor)
- Excalibur (mythology/metaphor)
- Jevons Paradox (economics/mental-model)
- Inspect and Correct (food-and-cooking/mental-model)
Structural Tags
Patterns: pathflowmatching
Relations: causeenableselect
Structure: cycle Level: specific
Contributors: agent:metaphorex-miner