About

P. Jeffrey Brantingham is a professor at UCLA in the Anthropology department. His research focuses on the study of human behavior in complex environments. He employs mathematical and computational models to understand the mechanisms that generate behavioral patterns. Additionally, he utilizes a range of computational methods to learn features of behavioral systems from real-world data. His current research examines crime patterns in space and time, as well as the interaction between online and offline behavioral systems. Brantingham is also the co-editor-in-chief of the Journal of Quantitative Criminology.

Research topics

• Political Science
• Computer Science
• Computer Security
• Law
• Psychology
• Medicine
• Criminology
• Economics
• History
• Linguistics
• Econometrics
• Medical emergency
• Social psychology

Selected publications

• Target Choices of Inner-City Illegal Taggers Demonstrate Consistency and Specificity

  CrimRxiv · 2025-10-22

  preprintOpen access

  Offender populations are diverse, with individual offenders preferring specific target types. While extensive research has focused on high-impact crimes, the consistency and specificity of lower-impact offenses like tagging remain underexplored. Consistency refers to how stable an offender’s preferred target type is, and specificity addresses whether different offenders prefer different target types. We examined these patterns among taggers, a type of illegal graffiti writers, in Ghent, Belgium, using graffiti removal data. Our dataset comprised 1,651 non-gang related tags by 248 taggers who have been observed at least twice in an inner-city area. We used the Hunter-Gaston Diversity Index (HGDI) for target preference consistency and the Weighted Nestedness Metric based on Overlap and Decreasing Fill (WNODF) for specificity. Observed values were compared with Monte Carlo simulations of random target choices. On average, each tagger produced 6.657 tags across 2.427 target categories. Most tags are clustered in a few types, with 32 taggers responsible for 50% of all tags. The observed HGDI (0.587) was significantly lower than the mean simulated (0.698), indicating consistency in target preferences. The observed WNODF (25.537) was also significantly lower than the mean simulated (31.608), suggesting no collective taste but rather specificity. Our results reveal a moderate level of consistency and specificity in taggers’ target preferences within an inner-city setting.

  PublisherOA PDFDOI

• Modelling the spillover from online engagement to offline protest: stochastic dynamics and mean-field approximations on networks

  ArXiv.org · 2025-07-17

  preprintOpen access

  Social media is transforming various aspects of offline life, from everyday decisions such as dining choices to the progression of conflicts. In this study, we propose a coupled modelling framework with an online social network layer to analyse how engagement on a specific topic spills over into offline protest activities. We develop a stochastic model and derive several mean-field models of varying complexity. These models allow us to estimate the reproductive number and anticipate when surges in activity are likely to occur. A key factor is the transmission rate between the online and offline domains; for offline outbursts to emerge, this rate must fall within a critical range, neither too low nor too high. Additionally, using synthetic networks, we examine how network structure influences the accuracy of these approximations. Our findings indicate that low-density networks need more complex approximations, whereas simpler models can effectively represent higher-density networks. When tested on two real-world networks, however, increased complexity did not enhance accuracy.

  PublisherOA PDFDOI

• Calling the police as an interdependent security game

  Journal of Mathematical Sociology · 2025-02-06

  article1st authorCorresponding

  Calling to report crime represents public cooperation with the police. When rational individuals are predicted to report (and when not) is still poorly understood. We study an interdependent security game under threat of a costly event that can only occur once or is perceived as so costly that the threat of the event occurring more than once is (in foresight) perceived as no more costly than the event occurring only once. Our analysis suggests how the interactions among the benefits, costs and neighborhood effects of police response might affect reporting. When there is spatial contagion of crime, rational individuals may choose to report when more of their neighbors report. When there is spatial contagion of deterrence, the relationship is reversed.

  PublisherDOI

• An Ecological Model of Place-based Deterrence

  Journal of Quantitative Criminology · 2025-10-29

  articleOpen access1st authorCorresponding

  Abstract Purpose The central premise behind place-based policing is that an intervention narrowly targeted to a location is able to suppress crime for some period of time. The crime-free survival time in a place ends with prolonged exposure to police action, known as initial deterrence decay, or after police have left, known as residual deterrence decay. The purpose of the present work is to understand the origin and character of deterrence decay at an aggregate spatial scale. Methods Deviating from previous efforts that explain deterrence decay based on the psychology of offender decision-making, the present work borrows ideas from theoretical ecology to model place-based deterrence as a form of competition between police and offenders over space. Deterrence decay emerges as a byproduct of this competition. Results When measured on an aggregate spatial scale, the model suggests that the waiting time to the emergence of crime and disorder from the onset of place-based policing actions should be gamma-like in distribution. The waiting time from the end of a place-based police action should be exponentially distributed. Conclusion If the model is a reasonable approximation of reality, then attempts to schedule place-based maintenance visits to counteract deterrence decay may be of limited value.

  PublisherOA PDFDOI

• Target Choices of Inner-City Illegal Taggers Demonstrate Consistency and Specificity

  Deviant Behavior · 2025-07-31

  articleOpen access

  Offender populations are diverse, with individual offenders preferring specific target types. While extensive research has focused on high-impact crimes, the consistency and specificity of lower-impact offenses like tagging remain underexplored. Consistency refers to how stable an offender’s preferred target type is, and specificity addresses whether different offenders prefer different target types. We examined these patterns among taggers, a type of illegal graffiti writers, in Ghent, Belgium, using graffiti removal data. Our dataset comprised 1,651 non-gang related tags by 248 taggers who have been observed at least twice in an inner-city area. We used the Hunter-Gaston Diversity Index (HGDI) for target preference consistency and the Weighted Nestedness Metric based on Overlap and Decreasing Fill (WNODF) for specificity. Observed values were compared with Monte Carlo simulations of random target choices. On average, each tagger produced 6.657 tags across 2.427 target categories. Most tags are clustered in a few types, with 32 taggers responsible for 50% of all tags. The observed HGDI (0.587) was significantly lower than the mean simulated (0.698), indicating consistency in target preferences. The observed WNODF (25.537) was also significantly lower than the mean simulated (31.608), suggesting no collective taste but rather specificity. Our results reveal a moderate level of consistency and specificity in taggers’ target preferences within an inner-city setting.

  PublisherOA PDFDOI

• <scp>PlayBest:</scp>Professional Basketball Player Behavior Synthesis via Planning with Diffusion

  2024-10-20 · 6 citations

  article

  Dynamically planning in complex systems has been explored to improve decision-making in various domains. Professional basketball serves as a compelling example of a dynamic spatio-temporal game, encompassing context-dependent decision-making. However, processing the diverse on-court signals and navigating the vast space of potential actions and outcomes make it difficult for existing approaches to swiftly identify optimal strategies in response to evolving circumstances. In this study, we formulate the sequential decision-making process as a conditional trajectory generation process. Based on the formulation, we introduce PlayBest (PLAYer BEhavior SynThesis), a method to improve player decision-making. We extend the diffusion probabilistic model to learn challenging environmental dynamics from historical National Basketball Association (NBA) player motion tracking data. To incorporate data-driven strategies, an auxiliary value function is trained with corresponding rewards. To accomplish reward-guided trajectory generation, we condition the diffusion model on the value function via classifier-guided sampling. We validate the effectiveness of PlayBest through simulation studies, contrasting the generated trajectories with those employed by professional basketball teams. Our results reveal that the model excels at generating reasonable basketball trajectories that produce efficient plays. Moreover, the synthesized play strategies exhibit an alignment with professional tactics, highlighting the model's capacity to capture the intricate dynamics of basketball games.

  PublisherDOI

• Gang Ecological Diversity in the Hollenbeck Area of Los Angeles, 1978–2012

  Oxford University Press eBooks · 2024-01-23

  book-chapter1st authorCorresponding

  Abstract Criminal street gangs are paradoxically both ephemeral and durable social forms. Unique gangs might emerge to only disappear a short while later, while others may persist for decades. These dynamics raise questions about the ecological diversity of gangs over time and space. This chapter examines gang diversity in an area of Los Angeles over a 35-year period. It looks to annual counts of the number of uniquely named gang cliques identified in gang-related homicides. The chapter finds that gang diversity fluctuates over time from a minimum of 8 to a maximum of 34 unique cliques. Homicide is concentrated among a small number of unique cliques; one uniquely named clique is tied to 86 homicides over this period. Most cliques are around for a very short period of time—the median lifespan of a clique is seven years, and 41% of all cliques appear for just one year. While the homicides per clique is relatively stable over time, it is clear that more cliques on the landscape is associated with more homicides per clique; there were 63% more homicides with 34 unique cliques on the ground compared to when there were just 4. Overall, the chapter estimates that this area of Los Angeles could support as many as 40 unique cliques at any one time. It considers several explanations for the observed patterns and discuss how gang diversity offers a unique window into the challenges communities with chronic gang problems face.

  PublisherDOI

• A macroarchaeological view of mobility

  Journal of Archaeological Science Reports · 2024-11-29

  article1st authorCorresponding
  PublisherDOI

• Optimal policy for control of epidemics with constrained time intervals and region-based interactions

  Networks and Heterogeneous Media · 2024-01-01

  articleOpen access

  &lt;p&gt;We introduce a policy model coupled with the susceptible–infected- recovered (SIR) epidemic model to study interactions between policy-making and the dynamics of epidemics. We considered both single-region policies as well as game-theoretic models involving interactions among several regions and hierarchical interactions among policy-makers modeled as multi-layer games. We assumed that the policy functions are piece-wise constant with a minimum time interval for each policy stage, considering that policies cannot change frequently in time or be easily followed. The optimal policy was obtained by minimizing a cost function that consists of an implementation cost, an impact cost, and, in the case of multi-layer games, a non-compliance cost. We show, in a case study of COVID-19 in France, that when the cost function is reduced to the impact cost and parameterized as the final epidemic size, the solution approximates that of the optimal control in Bliman et al, (2021) for a sufficiently small minimum policy time interval. For a larger time interval, however, the optimal policy is a step down function, quite different from the step up structure typically deployed during the COVID-19 pandemic. In addition, we present a counterfactual study of how the pandemic would have evolved if herd immunity was reached during the second wave in the county of Los Angeles, California. Finally, we study a case of three interacting counties with and without a governing state.&lt;/p&gt;

  PublisherDOI

• Optimal policy for control of epidemics with constrained time intervals and region-based interactions

  arXiv (Cornell University) · 2024-08-04

  preprintOpen access

  We introduce a policy model coupled with the susceptible-infected-recovered (SIR) epidemic model to study interactions between policy-making and the dynamics of epidemics. We consider both single-region policies, as well as game-theoretic models involving interactions among several regions, and hierarchical interactions among policy-makers modeled as multi-layer games. We assume that the policy functions are piece-wise constant with a minimum time interval for each policy stage, considering policies cannot change frequently in time or they cannot be easily followed. The optimal policy is obtained by minimizing a cost function which consists of an implementation cost, an impact cost, and, in the case of multi-layer games, a non-compliance cost. We show in a case study of COVID-19 in France that when the cost function is reduced to the impact cost and is parameterized as the final epidemic size, the solution approximates that of the optimal control in Bliman et al, J. Optim. Theory Appl., 189, 2021, for sufficiently small minimum policy time interval. For a larger time interval however the optimal policy is a step down function, quite different from the step up structure typically deployed during the COVID-19 pandemic. In addition, we present a counterfactual study of how the pandemic would have evolved if herd immunity was reached during the second wave in the county of Los Angeles, California. Lastly, we study a case of three interacting counties with and without a governing state.

  PublisherOA PDFDOI

Recent grants

• AWARE U.S.-China Cooperative Research on Pleistocene Hunter-Gatherers of the Tibetan Plateau

  NSF · $137k · 2002–2006

• Early Human Occupation of the Tibetan Plateau and the Development of High Altitude Adaptations

  NSF · $190k · 2009–2012

• DHB: Mathematical and Simulation Modeling of Crime Hot Spots

  NSF · $750k · 2006–2009

Frequent coauthors

• Andrea L. Bertozzi

  40 shared

• George Tita

  35 shared

• Patricia L. Brantingham

  Simon Fraser University

  33 shared

• George Mohler

  Boston College

  28 shared

• Martin B. Short

  27 shared

• A. Krivoshapkin

  Institute of Archaeology and Ethnography

  18 shared

• Li Jinzeng

  Santa Fe Institute

  16 shared

• Ya. Tserendagva

  16 shared

Education

• B.A.

  UCLA Anthropology