Geographical profiling uses the locations of linked offences and their context to prioritise possible areas connected with an offender’s activity. An anchor point can be a home, workplace or another regularly used place.
A ranked search surface is an investigative aid. Its value depends on linkage, location quality, mobility, the environment and evaluation against suitable alternatives; it does not identify a person or prove an address.
By the end of this topic you will be able to:
- Explain distance decay, anchor points and environmental context.
- Assess linkage, location uncertainty and model assumptions.
- Compare search rankings with suitable baselines.
- Anchor point
- A location regularly used by an offender, not necessarily the home.
- Distance-decay function
- A distance-dependent model relationship whose form and parameters need justification.
- Buffer zone
- A near-site or near-anchor reduction represented by particular models, not a universal known exclusion zone.
- Routine activity theory
- A framework concerning offender, target and guardianship convergence.
- Jeopardy surface
- A ranked spatial output; its scores are not automatically calibrated probabilities.
- Hit-rate
- A search-effort measure whose definition and evaluation boundary must be stated.
Geographical Profiling: Distance-Decay and Routine Activity Theory
Offence locations can reflect opportunities encountered through ordinary movement. Environmental theories provide hypotheses about that relationship rather than a universal rule that all offending occurs near home.
Distance-Decay and Offender Anchor Points
Distance decay describes how occurrence or modelled contribution changes with distance. Its form and parameters depend on the model and data; no universal hundred-metre buffer applies.
Routine Activity Theory
Routine activity theory considers the convergence of a potential offender, a suitable target and inadequate guardianship. Crime-pattern approaches also examine nodes, paths and awareness spaces.
Record roads, barriers, target distribution and travel opportunities rather than assume that straight-line distance explains every event.
Research Traditions and Rossmo’s Contribution
Environmental criminology and investigative psychology contributed to the field. Kim Rossmo’s doctoral work at Simon Fraser University helped develop a systematic geographical-profiling approach.
A single founder story should not erase other research traditions. Distinguish the history of mapping from a specific mathematical model or software implementation.
Rossmo-Style Distance Models and Search Scores
Rossmo-style models combine distance-based contributions from offence sites to rank candidate anchor locations. A buffer component can downweight nearby candidate positions under the model’s assumptions.
The anchor is unknown: the implementation evaluates candidate locations relative to observed sites. It does not begin by drawing an exclusion ring around a known offender’s home.
A numerical score or normalised surface is not automatically a calibrated probability that a person lives in a cell. Explain what the implementation calculates and how its performance was evaluated.
Geographical Profiling: Data, Workflow and Evaluation
Verify why incidents are considered linked, and distinguish encounter, offence and recovery locations. Record geocoding uncertainty and whether coordinates represent an exact site or an administrative approximation.
Define the search area, model version, parameters and relevant constraints. Compare performance with simple baselines rather than assume specialised software is superior.
Evaluate rankings using known-anchor cases where appropriate. Measures such as area searched before reaching a known location depend on how area, ties and the search boundary are defined.
There is no universal five-site threshold or guaranteed 70–90 percent reduction in search effort. More sites do not repair incorrect linkage or biased location data.
Research and Practice across Jurisdictions
In the US, NIJ-sponsored work has examined how to evaluate geographical-profiling software. Publication of an evaluation method does not establish that every product is effective.
In the UK, geographic and behavioural advice must be used within the relevant investigative and legal framework. For India, a nationwide adoption or procurement claim needs a named, current official source.
Rossmo’s Canadian research and policing history are relevant to the field’s development. Australian and other applications likewise need service-specific evidence rather than an assumed universal deployment.
Geographical Profiling: Limits and Alternative Explanations
A profile is not an identification, a search warrant or proof of guilt. Legal authority and case evidence remain separate requirements.
The geometric circle used in some research is different from a convex hull around offence sites. Neither construction guarantees containment of an anchor.
Consider commuting, changing bases, travel networks, victim and target distribution, missing incidents and wrongly linked events. State how these uncertainties affect prioritisation.
Check Your Understanding
Sources
- 1.A Methodology for Evaluating Geographic Profiling Software · Rich and Shively; NIJ-sponsored research
- 2.Kim Rossmo: research and policing background · Simon Fraser University
- 3.Connecting the dots: assessing geographic-profiling software · Derek J. Paulsen, research article