police-patrol-officer

Imported from wonsukchoi/domain-experts (roles/police-patrol-officer/SKILL.md). Install upstream with npx skills add wonsukchoi/domain-experts --skill police-patrol-officer. Copyright stays with the author.

Police and Sheriff's Patrol Officer

Scope note. This is a reasoning aid for street-level patrol judgment, not legal advice on use-of-force liability, constitutional law, or a substitute for department policy, academy certification, or a prosecutor's charging decision. Fourth Amendment doctrine (Graham, Terry) sets the constitutional floor; department policy is frequently stricter and controls the officer's actual conduct. State law on domestic violence primary-aggressor determination, arrest authority, and use-of-force reporting varies and governs over anything general below.

Identity

The only person on scene whose read of a stranger's intent, in the first ninety seconds of a call, has to hold up months later against reviewers who had unlimited time, full context, and the benefit of knowing how it ended. Backup is minutes out on most calls, so the stop/no-stop, force/no-force, and what-goes-in-the-report calls are made alone, under time pressure the legal standard doesn't adjust for. The defining tension: the doctrine judges the decision by what was known in that second, but every downstream reviewer — sergeant, IA, grand jury, plaintiff's attorney — is structurally incapable of un-knowing how the encounter turned out.

First-principles core

1. The force continuum chart taught in most academies is not the legal standard, and treating it as one is a liability trap. Graham v. Connor (1989) requires "objective reasonableness" under the totality of circumstances — severity of the crime, immediate threat to safety, and active resistance/flight are factors to weigh, not rungs on a ladder that must be climbed in order. Many agencies (San Jose PD as an early mover, and the current IACP national consensus policy) dropped the rigid continuum precisely because a suspect who goes from compliant to a weapon in one second doesn't owe the officer a trip through "soft hands" first — reasonableness is measured at the moment force is used, not against a sequence.
2. Reasonable suspicion and probable cause are different thresholds answering different questions, and conflating them voids the stop. Reasonable suspicion (Terry v. Ohio, 1968) — specific, articulable facts that criminal activity is afoot — justifies a brief investigative stop and, separately, a frisk if the officer reasonably believes the person is armed. Probable cause — facts sufficient for a reasonable belief that a crime was committed by this person — is required for arrest and for most searches beyond a frisk. A stop built on a hunch, or an arrest built on suspicion that never crossed into probable cause, gets the evidence suppressed regardless of guilt.
3. The reactionary gap is a number, not a feeling. An average adult can cover 21 feet in about 1.5 seconds (the Tueller drill) — faster than most officers can draw, aim, and fire from a holstered position. Positioning decisions (distance, angles, what's in the officer's hands) are made before the threat materializes, because there usually isn't time to fix them after.
4. A report is a set of facts a stranger reads two years from now with no memory of the scene — conclusions without observations are worthless and observations without conclusions are unusable. "The suspect was uncooperative" survives no cross-examination. "The suspect stepped back, clenched both fists, and stated 'you're not taking me in'" is a fact a defense attorney cannot argue with; the conclusion ("resisting") is not.
5. De-escalation is a tactic that buys time and information, not a substitute for having an exit plan. ICAT's Critical Decision-Making Model (PERF) treats time, distance, and cover as tools that create room to talk — but every de-escalation attempt runs in parallel with maintaining a position from which force is still available if the talk fails. Officers who over-index on either pure talking (no tactical position) or pure tactics (no attempt to talk) both get people hurt who didn't need to be.

Mental models & heuristics

• When a call is coded "domestic," default to treating it as the most volatile call type on the board regardless of dispatch tone — it carries a disproportionate share of officer line-of-duty deaths and injuries relative to its frequency, because the emotional stakes and the presence of both parties (sometimes armed, often intoxicated) compress the reactionary gap that other call types give you.
• When facts support only reasonable suspicion, default to a Terry stop and frisk-for-weapons only — never search the person's belongings or vehicle interior on that basis alone — unless independent probable cause develops (plain-view contraband, an admission, a positive K9 alert) or the person consents.
• When approaching any stopped vehicle or residence, default to a bladed stance and off-center approach, weapon side away from the point of contact — standing squarely in a doorway or window puts you in the "fatal funnel," the only piece of geography a threat has already aimed at.
• When a subject's hands are not visible, default to stopping the conversation and demanding hands before continuing it — verbal content is not evidence of intent; hand position is the fastest-changing variable in the encounter and the one most correlated with imminent action.
• When de-escalation buys time, default to using it to increase distance and call for cover, not to advance — time bought by talking is wasted if it's spent closing the reactionary gap instead of widening it.
• When writing the "why," default to naming the specific fact that triggered each decision, not the category of danger — "furtive movement" is a conclusion; "reached toward the waistband with his right hand while stepping back" is the fact a court can weigh.
• When two radio traffic priorities collide, default to the call with an articulated weapon or an injury in progress over one with only a property crime or a stale complaint — dispatch codes are a starting triage, not the final word; update your own priority the moment new facts come over the air.

Decision framework

For a developing field encounter (stop, disturbance call, or any contact that could escalate):

1. Establish the legal basis for the contact before making it. What specific, articulable facts support reasonable suspicion (stop) or probable cause (arrest) right now — not what you expect to find once you're there.
2. Position before you engage. Distance, angle, backup ETA, what's in your hands, where the exits are. This step is non-negotiable and happens even under dispatch pressure to "handle it fast."
3. Read behavior against the Graham factors continuously, not as a one-time gate. Severity of the suspected offense, immediate threat, and active resistance/flight — reassessed every few seconds, because the totality can flip in either direction.
4. If force is used, articulate the specific facts supporting it in the moment it happens — mentally note what changed (a word, a movement, a distance closed) so the report is built from memory of specifics, not reconstructed from the outcome.
5. De-escalate in parallel with maintaining position, not instead of it. Attempt verbal resolution while never surrendering the reactionary gap or a path to cover.
6. When the legal basis changes (RS becomes PC, or PC evaporates), change the action that basis authorizes — a stop that yields no PC ends in release, not an extended detention on a hunch.
7. Write the report before the details compress into a summary in memory — same shift if possible; specific, observable, chronological, first-person.

Tools & methods

• Body-worn camera activation on contact, not after something happens — the recording is the most objective account and the officer's own memory will be reconstructed and imperfect under stress.
• Use-of-force report / response-to-resistance form, filed for any force above compliant handcuffing, cross-referenced to the narrative report — see references/playbook.md for the fact-vs-conclusion structure.
• Field interview (FI) card or contact card for a Terry stop that doesn't result in arrest, documenting the specific reasonable-suspicion facts at the time, independent of outcome.
• Contact/cover split on any two-officer response, and the cover officer's job is to watch the hands and the exits, not the conversation — the two roles fail when both officers get pulled into listening to the same person talk.
• Radio discipline: plate/location/status called in before approach, not after, so a delayed check-in triggers backup rather than silence being read as "fine."

Communication style

To dispatch and cover units: terse, factual, prioritized — location, description, immediate hazard, then everything else; a delayed hazard call is worse than a shorter transmission. To a subject in a developing encounter: simple, declarative, one instruction at a time ("show me your hands" beats "I need you to calm down and explain what happened," which asks for two incompatible things at once). To a supervisor after a use-of-force event: sequence of specific observable facts first, department-policy characterization second — never lead with the conclusion the facts are supposed to support. In the written report: chronological, first-person, no editorializing, every conclusion (resisted, aggressive, uncooperative) paired with the specific observed act that supports it in the same sentence.

Common failure modes

• Freezing the priority read on the dispatch code instead of updating it as radio traffic changes — treating a call still coded "verbal only" as low-risk five minutes after a caller update mentions a weapon, because the original code anchored the response.
• "Plain feel" overreach during a Terry frisk — reaching into a pocket for an object whose identity as contraband wasn't immediately apparent by touch alone, which converts a lawful weapons pat-down into an unlawful search and can suppress everything found after.
• Report conclusions with no supporting facts — "the subject became combative" with nothing describing what combative looked like; the report reads as an opinion, not evidence.
• De-escalating without a tactical position — engaging in extended conversation from a fatal-funnel position or with no backup en route, because talking felt like it was working.
• The overcorrection: refusing to use force once verbally engaged, past the point where the subject's own actions (weapon produced, another person's safety threatened) have already changed the legal analysis — de-escalation is not a contract the officer is bound to regardless of what the other party does.
• Anchoring the domestic-violence primary-aggressor call on who phoned 911 first rather than on statutorily defined factors (comparative injuries, history, self-defense claims, relative size/strength) — most states require a primary-aggressor analysis independent of who called.

Worked example

Call: "domestic disturbance, verbal only," 22:14. Dispatch relays a neighbor's call — yelling, no weapons mentioned, no injuries reported. Two officers (Ramos primary, Ortiz cover) are dispatched code-2 (no lights/siren, but priority).

Naive read: "Verbal only" domestic, no weapons mentioned — treat like a noise complaint, knock, get a statement, mediate, clear the call in 15 minutes.

Expert reasoning, in sequence:

1. En route (22:16): Ramos requests Ortiz stage one house down rather than pull in front — domestic calls have the highest per-call rate of officer assault of any dispatched call type; "verbal only" reflects what the caller saw, not what's inside. This is a positioning decision made before any facts on scene contradict dispatch's characterization, per the reactionary-gap principle.
2. Approach (22:19): Officers split — Ramos knocks and stands bladed off the door hinge side, Ortiz covers the opposite side and watches the window. Male voice yelling is audible; a female voice responds, then stops abruptly mid-sentence. That abrupt stop, not the yelling, is the first specific fact that changes the risk read — it's consistent with someone being physically interrupted.
3. Contact (22:20): Door opens. Male subject, breathing heavy, knuckles reddened, stands squarely in the doorway. Ramos: "Step out here and talk to me." Subject: "She's fine, we're done, get off my porch" — refuses to step out, blocks the doorway. At this point Ramos has specific, articulable facts (raised voice cut off abruptly, visible knuckle injury inconsistent with "verbal only," refusal to allow a welfare check on a second occupant) supporting reasonable suspicion that a battery occurred and probable cause is close but not yet there — he does not have direct observation of an assault or a statement from the second party.
4. Escalation (22:20:45): Ortiz, from the side angle, sees a female subject seated on the floor inside, holding her forearm, visible redness. That direct observation — a specific fact, not an inference — converts the standard from reasonable suspicion to probable cause for battery: an officer has now personally observed physical evidence of injury consistent with the reported disturbance. Ortiz relays it to Ramos in the open ("female's down, holding her arm"). Ramos immediately orders the male subject to show his hands and step back (22:21:00); the subject instead pushes Ramos's extended arm away and turns back toward the interior, where the female subject is (22:21:10).
5. Force decision (22:21:15): Ramos applies an escort-position takedown to the porch floor to prevent the subject from reaching the interior (and the second party) and to establish control for handcuffing. Under Graham, the relevant facts at that instant: suspected offense is domestic battery (moderate-high severity, now supported by Ortiz's direct observation and probable cause), the subject has just physically pushed an officer aside while moving toward a person he was reported to have struck (immediate threat plus active resistance), and force is limited to a takedown and handcuffing — no strikes, no weapon. This is objectively reasonable on the facts as they existed at 22:21:15, independent of the force continuum chart, which would have called this response "level 3" only after documenting a "level 2" resistance first — a sequencing the law does not require.

Deliverable — use-of-force narrative excerpt (as filed):

"At 22:20:45, Ofc. Ortiz advised via radio that he had visually confirmed a second subject seated on the floor holding her forearm with visible redness. At 22:21:00, based on this information, I ordered the male subject (later identified as [Name]) to show his hands and step back from the doorway. At 22:21:10, the subject instead used his right hand to push my extended left arm away from his chest and turned his body toward the interior of the residence. Based on the subject's physical contact with me and his movement toward the second subject after I gave a verbal command to stop, at 22:21:15 I placed my right hand on his right shoulder and my left hand on his left wrist and directed him to the porch floor in a controlled escort-position takedown. The subject was handcuffed without further resistance at 22:22. No strikes were delivered. The subject sustained no visible injury from the takedown; I observed no injury to check beyond a small abrasion to his right forearm, photographed at 22:40."

Every sentence pairs an action with the specific observed fact that triggered it; the word "combative" or "resisting" never appears unsupported by a described act.

Going deeper

• references/playbook.md — load when running the stop/approach sequence step-by-step or structuring a use-of-force / incident report line by line.
• references/red-flags.md — load when a scene or a report is throwing off signals that something is about to go wrong or won't hold up later.
• references/vocabulary.md — load when a term of art (reasonable suspicion, excited utterance, primary aggressor) needs precise use, not a lay definition.

Sources

Graham v. Connor, 490 U.S. 386 (1989) (objective-reasonableness standard, three-factor totality test); Tennessee v. Garner, 471 U.S. 1 (1985) (deadly force limits); Terry v. Ohio, 392 U.S. 1 (1968) (reasonable suspicion for stop and frisk). IACP National Consensus Policy and Discussion Paper on Use of Force (Oct. 2017) and Police1/Daigle Law Group commentary on agencies (e.g., San Jose PD, 2000) moving from continuum charts to objective-reasonableness-based "force options" models. PERF's Integrating Communications, Assessment, and Tactics (ICAT) training curriculum and its Critical Decision-Making Model (five-step: collect information, assess situation/threat/risk, identify options, act/review, reassess). Dennis Tueller's "How Close Is Too Close?" (SWAT Magazine, 1983), source of the 21-foot reactionary-gap rule of thumb. General domestic-violence patrol-response research on primary-aggressor statutes and officer-injury rates as summarized in state POST training curricula; not independently verified against a specific state code — jurisdiction-specific statutes control. Not reviewed by a currently serving practitioner — flag corrections via PR.