Catch Me If You Can: Traversing the Criminal Mind

Posted: 5 Feb 2017

See all articles by Francis Kuriakose

Francis Kuriakose

Erasmus University Rotterdam

Deepa Iyer

University of Cambridge, Students

Date Written: February 4, 2017


The mind of a criminal works in very specific ways. A crime is a manifestation of the working of a criminal mind. Crime is therefore, a ‘trace’ of a criminal mind. During the planning of a criminal act, the mind of a criminal offender is subject to several types of logical fallacies like over generalization, tunnel vision and dichotomy of perspectives and these fallacies reflect in the crime scene. A criminal offender leaves two kinds of traces in a crime scene - (i) expressive traits that reveal dimensions of his personality and (ii) instrumental traits that bring out his motive for committing the crime. Both these traces are left while the criminal interacts with the victim and the environment of crime. This paper argues that by examining these traces back to the criminal, crimes can be prevented. Criminal linkage is a domain of crime investigation that works with behavioural dimensions of the criminal with laws of probability to link previously random criminal events. Crime linkage works on two theoretical assumptions related to behaviour dimensions - consistency and inter-individual variation. Assumption of consistency states that dimensions of behaviour of an offender across crime scenes is consistent, i.e., in a given circumstance offender behaviour is predictable. Assumption of inter-individual variation suggests that behaviour of one offender is different from another. The advantage of crime linkage is that information from various crime files can be pooled together for fresh analysis and interpretation. A number of variables like signature traces of the offender, modus operandi and temporal distance between two crimes can be used as indications of crime linkages. In recent times, Bayesian reasoning from statistics has been used to aid criminal linkage with productive results. Bayesian method of probability testing differs from other traditional approaches called frequentist methods. In traditional frequentist method, future probability of an event is determined using its ratio of repeatability. However, there are many instances where an event is not repeatable, for instance, a terrorist attack. According to Bayes’ rule, future probability of an event is a degree of belief dependent on prior information and conditional evidence. Bayesian reasoning is a process by which observations (new conditional evidence) are used to update the probability that a hypothesis is true given ‘a priori’ characterization of its plausibility (prior information). There are several strengths to this method - it allows for honest representation of uncertainty, initiating investigation from expert knowledge (solved crime data base), linking non-obvious patterns and handling missing data. This paper examines the application of Bayesian reasoning to criminal linkage for specific types of crimes. The major conclusion of the paper is that crime investigation has become more effective with interdisciplinary applications of behavioural science and statistics. This is because of the advantage of pooling previously unrelated information and building systematic representation of the offender and working out probability. ‘Probability and proof’ are intimately interlinked. Bayesian reasoning in crime linkage goes beyond attribution, correlation and prediction to bring out biographical, motivational and psychological profile of the offender. It is a systematic approach and a mathematical representation of behaviour model of a criminal offender.

Keywords: Crime linkage, Bayesian network, Marginal probability, Suspect prioritisation

JEL Classification: C11; D03

Suggested Citation

Kuriakose, Francis and Iyer, Deepa, Catch Me If You Can: Traversing the Criminal Mind (February 4, 2017). Available at SSRN:

Francis Kuriakose (Contact Author)

Erasmus University Rotterdam ( email )

Burgemeester Oudlaan 50
Rotterdam, 3062

Deepa Iyer

University of Cambridge, Students ( email )

United Kingdom


Here is the Coronavirus
related research on SSRN

Paper statistics

Abstract Views
PlumX Metrics