A research note on the methodological and theoretical considerations for assessing crime forecasting accuracy with the predictive accuracy index

•The Predictive Accuracy Index is a common metric utilized to indicate how accurate crime forecasting techniques are when predicting future crime.•The PAI traditionally relies on area in the denominator, overlooking more appropriate units.•We argue that length or street segments should be considered in the denominator of the PAI calculation to better represent practical applications of crime forecasting.

[1]  Robin Bryant,et al.  Evaluation of the MPS Predictive Policing Trial (redacted) , 2015 .

[2]  Lisa Tompson,et al.  (Looking) Back to the Future: Using Space—Time Patterns to Better Predict the Location of Street Crime , 2010 .

[3]  Jeremy Carter,et al.  A Penalized Likelihood Method for Balancing Accuracy and Fairness in Predictive Policing , 2018, 2018 IEEE International Conference on Systems, Man, and Cybernetics (SMC).

[4]  David Weisburd,et al.  How Much Time Should the Police Spend at Crime Hot Spots? Answers from a Police Agency Directed Randomized Field Trial in Sacramento, California , 2014 .

[5]  Ned Levine,et al.  The “Hottest” Part of a Hotspot: Comments on “The Utility of Hotspot Mapping for Predicting Spatial Patterns of Crime” , 2008 .

[6]  D. Weisburd THE LAW OF CRIME CONCENTRATION AND THE CRIMINOLOGY OF PLACE , 2015 .

[7]  Michael Leitner,et al.  Forecasting of Crime Events Applying Risk Terrain Modeling , 2015 .

[8]  Grant Drawve,et al.  Predictability of gun crimes: a comparison of hot spot and risk terrain modelling techniques , 2016 .

[9]  Grant Drawve,et al.  Enhancing Data-Driven Law Enforcement Efforts: Exploring how Risk Terrain Modeling and Conjunctive Analysis Fit in a Crime and Traffic Safety Framework , 2018, American Journal of Criminal Justice.

[10]  Grant Drawve,et al.  A Metric Comparison of Predictive Hot Spot Techniques and RTM , 2016 .

[11]  Xiaoyun Wu,et al.  Outcome Reporting Bias in Randomized Experiments on Substance Use Disorders , 2018, Journal of Quantitative Criminology.

[12]  Leif D. Nelson,et al.  False-Positive Psychology , 2011, Psychological science.

[13]  Charlotte Gill,et al.  Testing the “Law of Crime Concentration at Place” in a Suburban Setting: Implications for Research and Practice , 2017 .

[14]  Alese Wooditch,et al.  The point break effect: an examination of surf, crime, and transitory opportunities , 2015 .

[15]  Tao Cheng,et al.  Novel evaluation metrics for sparse spatio-temporal point process hotspot predictions - a crime case study , 2016, Int. J. Geogr. Inf. Sci..

[16]  Michael D. Porter,et al.  Rotational grid, PAI‐maximizing crime forecasts , 2018, Stat. Anal. Data Min..

[17]  Alese Wooditch,et al.  Using Space–Time Analysis to Evaluate Criminal Justice Programs: An Application to Stop-Question-Frisk Practices , 2016 .

[18]  Haojie Zhu,et al.  A Spatio-Temporal Kernel Density Estimation Framework for Predictive Crime Hotspot Mapping and Evaluation , 2018, Applied Geography.

[19]  Paul A. Zandbergen,et al.  Reference data and geocoding quality: Examining completeness and positional accuracy of street geocoded crime incidents , 2013 .

[20]  Mamoru Amemiya,et al.  Applying Crime Prediction Techniques to Japan: A Comparison Between Risk Terrain Modeling and Other Methods , 2018 .

[21]  Joel M. Hunt Do crime hot spots move? Exploring the effects of the modifiable areal unit problem and modifiable temporal unit problem on crime hot spot stability , 2016 .

[22]  Alese Wooditch,et al.  Do Stop, Question, and Frisk Practices Deter Crime? , 2016 .

[23]  Wilpen Gorr,et al.  Chronic and Temporary Crime Hot Spots , 2017 .

[24]  Jeremy D. Barnum,et al.  Using Vulnerability and Exposure to Improve Robbery Prediction and Target Area Selection , 2019, Applied Spatial Analysis and Policy.

[25]  Liping Fu,et al.  Identification of crash hotspots using kernel density estimation and kriging methods: a comparison , 2015 .

[26]  Paul A. Zandbergen,et al.  Kernel density estimation and hotspot mapping: examining the influence of interpolation method, grid cell size, and bandwidth on crime forecasting , 2014 .

[27]  Shane D. Johnson,et al.  Predictive Crime Mapping: Arbitrary Grids or Street Networks? , 2016, Journal of Quantitative Criminology.