Monday, July 25, 2016

Novel Techniques to Detect the Multiple-event Murderer: An Exploration of a Researcher’s Field Notes

Expansive fictional depictions of prolific atypical murderers sustain our collective fascination during their near absence from police blotters (DeSpirito, 2016), permeating all aspects of society while appearing in cultural conversations centered on artists (Caldwell, 2016; Ellwood, 2015; Wolcott, 2016), comedians (Everett, 2012; Kimble, 2015) and politicians (Tiffany, 2016). We have become a nation of #Fannibals (Hall, 2013), satiated by the deeds of what are now known as multiple-event murderers (MeMs) (Aamodt & Yaksic, 2015; Yaksic, 2015). Although Warren, Dietz, and Hazelwood (2013) surmise that these atypical murderers do not alter behaviors after learning from public sources, some believe them to become more adept at crime through this exposure. The utilization of multiple murderers as educational devices to entice students reinforces ingrained myths and damages efforts to dismantle them. Why We Love Serial Killers, a recent effort to co-mingle entertainment media with academic insight, resonated as a vapid attempt at garnering fame and profitability (Yaksic, 2014a). This chapter examines deficits exposed during systematic analysis of past work in homicide research (Yaksic, 2015). Detection of the modern day MeM requires forward propulsion stimulated by “anti-disciplinary” thinking and adapting concepts from disparate arenas into the criminological perspective.

Detecting the Devil in the Details

A recent Boston Magazine article titled Profiler 2.0 (Halber, 2016) redefined how criminal justice practitioners are expected to operate in a world saturated with data. We have remained entrenched in arguments about definitions and exclusion criteria in the domain of atypical homicide research (Yaksic, 2015) while efforts to create typologies of different killers are called a popular tactic (Shanafelt & Pino, 2013) and unhelpful in solving crimes (Mark Safarik, personal communication, 20 May 2016). Trepidation in advancing beyond descriptive statistics may be a strategic maneuver, one permitting us to avoid operationalizing data. Studying all criminal offenders with mainstream theory and methods (DeLisi, 2015) is met with apprehension as scrutiny of probabilistic genotyping and predictive algorithms – advancements recently applied to the criminal justice system – (Angwin, Larson, Mattu & Kirchner, 2016; Davey, 2016; Shaer, 2016) continues. Many relationships have been frayed in the struggle to build databases cataloging atypical homicide offenders. Comparably, a Pulitzer Prize winning journalist found himself at odds with colleagues after claiming credit for amassing data on the subject of fatal shootings by police even though similar databases had been built (Swaine, 2016).

Today, much of the criminal justice field has accelerated their effective use of data, evidenced by the National Institute of Justice employment posting for a ‘groundbreaking crime statistics analysis initiative’ (McGough, 2016), the existence of forecasting tools such as PredPol and HunchLab (Chammah & Hansen, 2016) and the Law Enforcement Advancing Data and Science Program (National Institute of Justice, 2016). Still, access to data with appropriate sample size and statistical power is the primary impediment to generating models meant to answer empirical questions about atypical homicide. Data can help determine if commonplace offenders invested in a criminal career behave like serial killers while interspersing murder into a range of offenses. Employing the Radford Serial Killer Database (SKDB) – consisting of 4,274 records across 170 variables (Boyne, 2014) – may aid in resolving queries about the differences between atypical murderers and ordinary criminals experimenting with duality (DeLisi, 2015).

Foraging for the Furtive Few

Some are confronting the false promises of “Big Data” even though exploiting this approach evokes a sense of boundlessness. Yaksic (2015) has proposed creating a dashboard that would apply data mining techniques to the task of identifying discernable patterns in the SKDB data but there may be danger inherent in surrendering ourselves over to a world run by algorithms (Davey, 2016). The aspirations of those interested in weaponizing data to combat the scourge of missing persons[1] (Joshua Zeman, personal communication, 14 May 2016) and unsolved homicides (Michelle Katz, personal communication, 18 May 2016; Sacha Thorpe, personal communication, 19 May 2016) currently outweigh the capabilities of databases because basic details within them, such as the manner in which offenders are apprehended, are often missing. This limitation fosters a reliance on findings from studies that either disregard subgroups within the population of atypical murderers (Kurkjian, 2016) or focus on the broad category of sexual murderers (Beauregard & Martineau, 2016). In response, atypical homicide researchers created an incubator designed to address serial offending (Office of the University Registrar, 2016) and formed a collaborative enterprise with the Homicide Investigation Tracking System (HITS) of the Washington State Attorney General's Office (Tamara Matheny, personal communication, 22 July 2016) and members of the Federal Bureau of Investigation’s (FBI) Behavioral Analysis Unit (BAU) 5 aiming to use research, strategy, and instruction to close the missing data gap (Sarah Craun, personal communication, 2 March 2016).

The culmination of interventions – police investigations, actions taken by the victim, or the involvement of witnesses (White, Lester, Gentile & Rosenbleeth, 2011) – result in the apprehension of MeMs but they sometimes reveal themselves via confessions or miscalculations. Murders can be linked by happenstance as in the “Black Widow” killings where one detective overheard another talking about a similar case (Associated Press, 2008). The transcript of serial murderer James Childers’ statements (United States Department of Justice, 2009) details a decision making process confirming that small choices lead to unanticipated paths (Shanafelt & Pino, 2013). Most serial homicides contain incidents occurring independent of the offender’s ingenuity: escaping capture by way of external situational factors beyond their control – known in common parlance as a ‘lucky break’. These episodes amount to unwarranted political involvement (Yaksic, 2014b), lapses in investigative strategies due to erroneous thinking (Kramer, 2016) or corruption that prevents official inquiry (Schram, Algar, & Tacopino, 2015).

Multiple-event Murderers as Human Beings: Exploiting Resilience and Collapse

Most of the work done in breaking serial killer stereotypes (Kuhn & Coston, 2005; Walters, Drislane, Patrick, & Hickey, 2015) stops short of considering atypical homicide offenders as individuals operating within a set of ‘normal aberrations’, shared cultural and psychological processes or in terms of what motivate our actions as people (Shanafelt & Pino, 2013; Yaksic, 2013). The portrayal of atypical murderers as monsters imbues offenders with fantastical powers, convincing them that they are advanced predators with exceptional abilities and expertise (Wiest, 2016). Adapting the conceptual framework provided by the ‘9 Principles’ (Copeland, 2012) summarizes how atypical murderers remain adaptable in changing terrain. By allowing failure to encourage adjustments in routines, not planning for everything in the beginning, taking risks instead of focusing on safety, fully embracing unexpected changes and events, being imaginative, practicing over theorizing, choosing disobedience over compliance, disallowing authority to limit potential and keeping vision and thinking open and flexible, MeMs can remain resilient and successful. It can be argued that a failure to abide by these principles has contributed to a decline in instances of serial homicide (DeSpirito, 2016; Yaksic, 2015).

Resiliency is ignored during debates about what makes MeMs successful. Because network science can provide a framework for capturing regularities in atypical homicide (Marcos Oliveira, personal communication, 23 June 2016), we can use this concept to conceive the tipping point between offender resiliency and collapse. Modifying the ideas posited by complex network researchers Gao, Barzel, & Barabási (2016) can help us understand how enterprising killers calibrate their actions to maintain base level functionality when environmental changes occur. Thinking about atypical homicide offenders as multi-dimensional systems consisting of a large number of components interacting through a complex network and adjusting to dis­tur­bances in order to remain functional is certainly a paradigm shift. Doing so could allow the prediction of the atypical murderer’s collapse under the weight of errors and internal failures.

The Utility of Criminal Investigative Analysis (CIA) in the 21st Century

What is being done today to uncover undetected, offending MeMs? Discussions about the methods utilized to discover MeMs are often superseded by endeavors to reveal the genetic outlier or negative experience thought to help aid in discovering other such murderers early on in their maturation (Parshley, 2016). The development of processes meant to assist in apprehending these offenders is a popular focal point due to their perceived dangerousness. CIA is often favored as an investigative suite of tools over other conventional techniques – such as linkage analysis programs like Crime Linkage International NetworK (2016) or the use of Bayesian modeling (de Zoete, Sjerps, Lagnado, & Fenton, 2015) – due to the lore surrounding the FBI’s success rate with the program. Shanafelt and Pino (2013) astutely call use of CIA inadvisable while Branson (2013) outright accuses the FBI of creating a matrix devoid of any social or cultural context while they revel in the “magic of the methodology”. Some are inspired by Agent Cooper’s mystical Tibetan Method from the television program Twin Peaks (Blassmann, 1999) and imprudently wield profiling as an armament in a battle of good versus evil (Mains, 2015).

The divide between perceptions and the practical use of CIA was evidenced by Yaksic (The Practicality of CIA in the 21st Century, Survey, 14 March, 2005) in a survey purposed at ascertaining CIA’s further usefulness. While each respondent understood that the perspectives surrounding the definition of CIA vary widely (Scherer & Jarvis, 2014) and knew it to be a useful contrivance, researchers felt that CIA has not adapted to the modern age and is based on an outdated dataset while the inverse was believed by the profilers. These rifts contribute to problems where faulty views and disparaging tones lead to research gaps and limit our scientific understanding of atypical offenders. Northeastern University’s Atypical Homicide Research Group (NUAHRG)[2] was formed to accommodate all viewpoints and reduce the strain of dichotomous relationships among law enforcement professionals and researchers. DeLisi (2015) categorizes these opinions as errors of social construction – a failure to acknowledge that serial murder is an area deserving of study by serious scholars – and stagnancy – a lack of innovation in scholarly thought – and recommends reclaiming this area of inquiry from true crime authors.

The advent of the serial murder entertainment industry has ushered criminal profiling into a sacred space as the cornerstone of forensic science courses. Academic institutions capitalize on university students convinced that becoming a “criminal profiler” is an immediately attainable goal. Although the objective of these programs is to instill a sense of wonderment regardless of how implausible the course, students should be actively discouraged from pursing ‘profiling’ as a career choice due to the abysmal state of grant funding dedicated towards this area of study (Angela Williamson, personal communication, 15 April 2016). The work of Muller (2000) might inspire students to redirect their energy towards other impactful fields of inquiry. Because individuals explain atypical homicide in the terms related to their profession (Shanafelt & Pino, 2013) they can tangentially connect their work to the pursuit of atypical homicide offenders. Environmental scientist Nigel Raine used geographic profiling to look at patterns of bee foraging and perfect the technique (Raine, Rossmo & Le Comber, 2009), electrical engineers Simkin and Roychowdhury (2014) suggest that the likelihood of successive killings is higher soon after a murder than after a long period has passed, physicist Neil Johnson developed a statistical model aimed at identifying behavioral patterns among online supporters of ISIS and used this information to predict the onset of violent events (Johnson, Zheng, Vorobyeva, Gabriel, Qi, Velasquez, Manrique, Johnson, Restrepo, Song, & Wuchty, 2016), Allely (2016) asks if we can predict who will become a mass shooter and the hacker Jester uses IBM's Watson AI to create a monitoring tool, called iAWACS, to analyze a user's "trustworthiness, propensity toward violence [and] openness" (Fung, 2016). The following vignettes comprise efforts by citizen sleuths, journalists, engineers and information technology professionals to create MeM detection tools and use novel approaches to understand how they interact with the world around them.

Utilizing New Tools to Locate MeMs in the Information Age

These ingenious efforts avoid preoccupation with apprehending MeMs so that attention can be redirected to detecting their presence using pattern recognition on characteristics of the victim and offender. The world of data science is fertile ground for folks outside of criminology to make an impact, especially as technical advances outpace practitioner’s understanding. Several projects underway heed Leetaru’s (2016) warning that the big data revolution is being driven by computer science with the majority of sentiment mining tools coming from that area rather than the disciplinary fields whose questions they are attempting to answer. Each platform detailed herein asks ‘where are atypical murderers hiding?’ as their originators were confounded by the findings of Beauregard and Martineau (2016) stating that strategies used by organized sexual murderers may increase the likelihood of detection by police. These architects considered a former FBI BAU 2 Unit Chief’s view that serial murder research has not “identified any radical new ideas over the past few years”[3] to be unsatisfactory and deemed current methodologies to be insufficient for prospectively locating MeMs. Each has committed to gathering data, designing algorithms and combining the efforts of those in the world of data science and engineering with criminal justice practitioners, an approach referred to as translational criminology or research with the potential for real-world implementation (Laub, 2012). Their guiding research question required supporting one of two schools of thought about the prevalence of multiple murderers where either hundreds of serial killers are actively trolling on an unending search for victims or their presence among the ranks of MeMs is steadily declining (Yaksic, 2015). These platforms were designed under the assumption that, aside from those cataloged in the SKDB as captured each year, a great number of active atypical murderers are unaccounted for in the United States.

The Murder Accountability Project (MAP) is an outgrowth of a 2010 national reporting project conducted by Thomas Hargrove at the E.W. Scripps Co. that aimed to investigate the declining homicide clearance rates in the United States (Crawley, 2015). MAP seeks to improve public awareness of the problem of unsolved homicides and identified bodies. Accurately accounting for unsolved homicides and making FBI murder data more widely and easily available to police, journalists and the general public is of importance as more than 216,000 homicides have not been solved since 1980. MAP seeks to improve the quality of reporting by police by using Freedom of Information laws to obtain information from state and local governments about unsolved homicides that many major police departments decline to supply to the FBI’s voluntary Uniform Crime Report (UCR) and Supplementary Homicide Report (SHR).

During the national reporting project, Hargrove created the Serial Killer Detector, a computer algorithm that flags potential serial killings within the FBI’s SHR (Editor & Publisher, 2010). This tool is now accessible on MAP’s interactive website (Hargrove, 2015) where the user can consider if a serial killer may be active within a specific community. The Scripps study identified an alarming number of unsolved killings of women in 161 clusters nationwide involving 1,247 deaths of women of similar age who were killed by similar means. Serial murderers are known to operate in a tight geographic area, selecting, killing and disposing of their victims within this zone. The SHR data available on the "Search Cases" tab of the MAP website can be helpful to homicide investigators testing their theories about homicide suspects who may have killed across multiple jurisdictions or within the same jurisdiction over time.

Hargrove used multivariate analysis on data reported to the FBI to calculate that at least one multiple-victim killer had operated in Indiana (Hargrove, 2014). Hargrove provided the names of fifteen potential serial homicide victims to the chief of police in a letter four years before serial killer Darren Vann was arrested. So far, one killing on the list is being investigated as part of Vann’s series. Next, a trend was spotted on the MAP website among the FBI’s case-level homicide data by selecting the “Search Cases” tab, “Atlanta” under metro areas, “Female” under victim’s sex and “strangulation” under weapon –  indicating that strangulation deaths of African American women occurred in Atlanta until the pattern abated in 2007. Hargrove turned to the NUAHRG to introduce the existence of this series to a member of the BAU. An analyst at the FBI’s Violent Criminal Apprehension Program (ViCAP) is reviewing the findings[4].

Rebecca Kramer and other citizen sleuths suspect that it may be advantageous to manually collect homicide event data. They understand that reliance on government information obtained from the noncompulsory FBI UCR/SHR may have limited the scope of MAP’s objectives. Kramer set out to investigate the ostensible murder of scores of college aged men in or near bodies of water in the hopes of identifying a pattern in the data. Kramer has persisted in collecting information on 400 instances of this phenomenon despite general opinion that these stories cannot be substantiated. The working theory, that men are being drugged at bars or lured to bodies of water whereby they are drowned and left at the homicide scene by at least one male or female serial murderer, can be investigated using advanced mapping technologies and sophisticated analytical approaches in Risk Terrain Modeling (RTM). Because the risk of crime is both place-based and situational – since victims often encounter offenders because of the activities they pursue and the environments they occupy – it is important to take note of the clustering of environmental risk factors and their spatial influences (Caplan & Kennedy, 2016).

RTM is a diagnostic technique that accounts for the different factors contributing to the spatial and temporal dynamics of illegal behavior, identifies the risks inherent in features of a landscape and models how they co-locate to create unique behavior settings for crime. Here, to assess the possibility that these drowning deaths are homicides, the co-location of bars and water was explored as contributing to conditions suitable for the victimization of unwitting men. The interaction effect can be used to compute a probability of criminal behavior occurring in the future. Because the data points within these programs are human beings moving in structured environments, it is critical to consider spatial factors of the topography as influential and enabling to criminal behavior. Swarm Intelligence – the idea that humans perform actions based on cues taken from each other (Forman, 2016) – should be surveyed in conjunction with RTM. Swarms are thought to occur as one event but they can be a series of people visiting a location over time, each making decisions based on physical cues left by those who have been there before. Marrying RTM and Swarm Intelligence makes logical sense as criminals operating independently can target the same places, evident in South Los Angeles where the bodies of 55 victims of five serial killers were located between 1984 and 2007 (Los Angeles Times, 2010).

Citizen sleuths like Kramer can be a productive member of the investigative team (Halber, 2013; Peters, 2013) but are often overlooked as a resource. Resultantly, Kramer turned to the NUAHRG to conduct outreach to secure cooperation from law enforcement and refine the project’s inclusion criteria. Modifications to the data collection procedures would require re-reviewing thousands of online records but, as a benefit to Kramer’s ongoing efforts, data scientist Peter Brendt is creating a method to collect and collate records from millions of news websites simultaneously. Brendt (2016) writes that previous efforts to connect human investigators with computers meant to aid in solving crimes have failed due to the idiosyncrasies inherent within both systems. The FBI’s ViCAP, the most widely available resource for those investigating potential linkages across crimes, is an application predicated on the idea that entry of all homicides or sexual assaults is not necessary for the program to be viable (United States Department of Justice, 2016). FBI analysts request that “cases of a serial nature” are prioritized for entry into ViCAP, providing autonomy to local law enforcement officers to select cases for upload to the platform. The onus to determine, prima facie, the likelihood that a case appears to be a part of series is a major limiting factor as it falls to those with little or no experience working a series of murders. ViCAP represents a true intersection of Brendt’s (2016) point that humans are hampered by how they react to a lack of resources, political affiliations, jurisdictional conflicts and distrust of others while computers are limited by data models not designed for complex comparisons, lack of granularity for attribute comparisons and variable weights of attributes of crimes that change by area, cultural predominance and victimology.

Brendt is developing a “knowledge-based” system with the LAMP software bundle of the Linux operating system, Apache HTTP Server, MySQL relational database management system (RDBMS), and PHP programming language designed to search for intersections in unidentified bodies, unsolved homicides and missing persons datasets. This ‘Weather Map’ may identify linkages using website screen scraping, data corroboration, deep learning and automatization. Brendt is motivated by the belief that these data constitute “hallmarks of a serial killer” and that their victims lay somewhere within it. The ambitious scope of this homicide offender forecasting tool would require working with incomplete and overlapping data of differing granularity, thousands of man-hours, top shelf hardware and algorithms “on the edge of technology”.

Brendt stresses that for this system to work, it would need to be as adaptable as atypical murderers have been and change along with their evolution. Current convention surrounding definitions must be ignored because, as Brendt states, “a serial killer can become a spree killer during his endgame”. The amount of data needed for the project would increase perpetually along with the demand for its quality because datasets become more limited, volatile and uncontrollable as they swell in size. The software system is being built modularly due to the necessary combination of techniques with parts dedicated to data entry, analysis and automatic data reading. While a full productive system is slated for future release, twenty potential serial homicide series have been detected by the ‘Weather Map’ across the United States to date.

Detecting MeMs using the ‘Weather Map’ will locate only those offenders that are well into their careers and kill frequently. Highly mobile serial murderers (HMSM) and burgeoning offenders are beyond the bounds of search parameters as they spread their victims outside of the time and geographical horizons. Further development of the recognition patterns must be completed to make the rulesets more logical in order to surpass the spatial and chronological restrictions. Because Donoho (2015) posits that coping with large scale cluster computing is hampering our ability to make appropriate judgements and holds us back from data analysis strategies that we would otherwise eagerly pursue, Joseph Johaneman, a student dual majoring in data analytics and applied mathematics, plans to build on Brendt’s work by constructing a small computing cluster from single-board computers (SBC) with the aim of locating homicide victims of HMSM (i.e. long haul truckers). Patterns in the SKDB will be used to train the cluster and design a mechanism to track HMSMs by creating a probabilistic function on the central hub of their movement based on multiple drop sites, hotzones and victim locations. The Manhattan Distance formula and buffer zones used in Kim Rossmo’s geographic profiling tool (Rossmo, 1995) are replaced by Mahalanobis Distance, the space between two probability distribution sets. Anchor points are used and based in the space and time of a trucker’s routines along established routes. Incorporating a temporal element, the constant range of the semi-trailer truck’s speed and assigning weights to favored body dump locations to ascertain their selection probability will enhance current strategies. SBCs allow for good performance per watt, the ability to conduct deep learning activities essential to the project and for police departments and newspapers to build their own systems due to their low cost (Baun, 2016). Johaneman will address whether the isolation and anonymity of driving invites the violence-prone (Strand, 2016) and provide what he learns to the FBI’s ongoing Highway Serial Killings Initiative (HSKI), a program implemented to attribute more than 500 murder victims from along highways to some 200 potential suspects.

Oklahoma State Bureau of Investigation Intelligence Analyst Terri Turner, the inspiration for HSKI, arduously identified a series of seven homicides in late 2003 which victimized truck stop prostitutes and encompassed multiple states and jurisdictions. She coordinated case information between all the agencies involved and monitored law enforcement and open source information for new incidents. Turner partnered with ViCAP and the BAU to facilitate the exchange of case information and provide investigators with “best practices” in working these cases. Johaneman hopes to use his work to identify future killer truckers like Bruce Mendenhall or Adam Leroy Lane before they are able to wantonly victimize others.

The Importance of Debunking Ingrained Myths

The search for atypical murderers among the incarcerated may seem counterintuitive but many unidentified victims result from an offender’s self-imposed silence. Angela Williamson plans to obtain the DNA of HMSMs and enter those genetic details into the Combined DNA Index System (CODIS) (Angela Williamson, personal communication, 15 April 2016) to retrospectively match offenders to victim’s remains (White et al., 2011), a method used to no avail during the East Area Rapist-Original Nightstalker (EAR-ONS) investigation. Many of the legacy axioms included in the EAR-ONS criminal profile (D’Ambrosia, 1998), such as those surrounding race and gender, must be controverted to approach this project without cognitive bias. The stereotype of the Caucasian male serial killer allowed groups of offenders to continue victimizing others as African American serial murderers have historically been discounted by criminologists who thought them incapable of enduring through a career of serial murder. Yaksic (2006) normalized the notion that African American MeMs are as abundant as their Caucasian counterparts while Farrell, Keppel, and Titterington (2011) demonstrated the differences in characteristics between male offenders and their underestimated female equivalents.

White et al. (2011) and Hickey (2014) call for the study of offenders that kill less than three victims since they are excluded from databases. Offenders that spread their victims outside of the capabilities of the ‘Weather Map’s’ time horizon could theoretically continue to evade detection (Brendt, 2016). SKDB data indicate that the majority of two-victim offenders kill for financial gain while those killing three or more victims normally do so for enjoyment. The severity of the offender's crimes concomitantly increases as their victim count rises. The odds of two-victim offenders committing a third murder escalates if they are white, use multiple methods, kill for enjoyment and commit the first two homicides close together in time.

Limitations of Programs on the Edge of Tomorrow

Each of these initiatives is informed by basic rulesets established from cataloged knowledge about how MeMs select and kill their victims. The assumption that atypical murderers abide by patterns, namely those where victims are clustered within a geographic location, is at the foundation of each program. Consideration should be given to findings that indicate differences between the planning behaviors of American and South African serial murderers (Sorochinski, Salfati, & Labuschagne, 2015) and that intervals between successive killings may be predictable (Lange, 1999). Relying solely on the presence of tightly grouped victim caches would make the task of MAP unsound but the algorithm also accounts for offenders using the same kill method over their series, a fairly consistent measure (Hargrove, Witzig, Icove, Harry, Arntfield, Yaksic, Lang, & Wolf, 2016). Kramer may be overestimating the prevalence of the serial murderer in modern society but some researchers subscribe to the belief that virtually anyone can emerge as a perpetrator of atrocities given the "right" combination of factors (Shanafelt & Pino, 2015). This logic dictates that large swaths of recovered bodies could be indicative of an unsolved serial homicide series. Yaksic (2015) argues that the phenomenon is declining, partially due to the attention that theories about MeMs receive before more logical possibilities are considered. Feedback from the NUAHRG highlights our preoccupation “with serial killers, which in reality are very rare”[5]. Because extraneous factors may prevent HMSMs from frequenting the same locations, Johaneman may encounter issues identifying the truck stops that drivers prefer. Researchers may be ‘censoring’ the data on offenders that claim two victims by failing to consider their burgeoning status and the further victims their discovery precluded them from amassing, disallowing their chance to evolve into a “traditional” serial murderer. Williamson may be unable to compel offenders to supply their DNA profiles. Brendt’s ‘Weather Map’ platform incorporates the ambiguous concept of escalation over a killer’s career and stipulates that offenders do not revert to previous routines in killing method. Educating the ‘Weather Map’ to detect patterns based in this logic will needlessly restrict the program by not accounting for the variability inherent in the decision making process of human beings. Premature capture always interrupts a MeM’s plans and negates any opportunity to learn if escalation is a well-founded concept. The ‘Weather Map’ program would overlook offenders altering their techniques dependent on their victim type.

The Blood Red Flag of Domestic Violence

The subset of multiple-method murderers contains a growing population of offenders that first practice violence on their spouse, brutalizers that choke their wives during domestic disputes before targeting members of the public. Rehearsals for these campaigns include instances of intimate partner violence and culminate in cruelty towards others (Shifman & Tillet, 2015). The precursory link between private brutality and public savagery is a crucial part of the MeM detection labyrinth, an unacknowledged corollary due to investment in thinking of these offenders as loners without families. Returning home to continue their vicious attacks is unsurprising because the phenomenon of atypical murder has been succinctly rooted in male dominance for millennia. Serial rapists, devoid of enjoyment, frequently eliminate the only witness to their crime as part of a scheme to avoid detection. Non-sexual serial killers have gravitated to melding their features with characteristics of spree offenders since aspects of the modern age have made it difficult to behave like their past counterparts (Yaksic, 2015). Deceit can be readily identified in the ‘Age of Authenticity’ and is infrequently tolerated which forces transparency upon atypical murderers, jeopardizing their ability to exercise the dual facets of their personality that contribute to success. MeMs adapted to a more open world by becoming proficient at sharing their contempt with others through outward displays of anger and revenge.

Broadening the scope of offenders comprising the MeM definition to include these unsparing men requires giving motive additional consideration. Anger is not considered a key component of the serial sexual homicide offender's motivation, according to Myers, Husted, Safarik, and O'Toole (2006), but adding ever increasing spree murderers (Kirby, Graham, Green, 2014) to the cohort of atypical murderers, calls for reexamining 'toxic masculinity' (Hamblin, 2016) in this context. The recent actions of Eulalio Tordil, (Alexander & Bui, 2016) Mainak Sarkar (Hamilton, Watanabe, & Winton, 2016), Cedric Ford (Sanchez, Berlinger, & Flores, 2016) and Edward Acquisto (Mettler, 2016) encapsulate the social issues and sentiments ongoing in American culture: male entitlement, hypermasculinity, stringent gender roles, unabashed sexual expectations, objectification, and unchecked rage – all coupled with flagrant use of firearms. These offenders reinforce antiquated viewpoints on manliness, an abstraction requiring a mixture of anger (Hayes, 2016), brute force and a domineering attitude towards others. Society at large has taken notice of intimate partner violence of late because of the tendency for offenders to pose a threat to others after harming or killing their significant other.

What we knew about atypical homicide in the ‘traditional’ sense has given way to hard-won realizations over the past decade: more impatient and incompetent offenders are acting out, often erratically, while leaving footprints on social media or electronic devices (Felton, 2016; Meyjes, 2016). Criminologists must resolve their misgivings on these changes because algorithms learn by consuming information upon which the system builds a model of the world. If a system is trained on faulty data, it will have a harder time generating the appropriate answer (Crawford, 2016). These learning systems must be aware of circumstances resulting in outcomes such as offenders killing freely due to apathetic enablers (Crimesider Staff, 2016) and law enforcement missteps (Holman, 2016), premature captures due to unforeseen handicaps (Alexander & Bui, 2016) or an inability to fulfill plans because potential victims alter their routine (Hamilton et al., 2016) so that accurate information is fed into the platform.

“One-off” intimate partner and familial murderers sometimes exhibit what investigators call hallmarks of serial murder activity (Truesdell, 2016) as in the killings perpetrated by Gregory Scott Hale (Radar Staff, 2014), John Robert Charlton (Carter, 2016), Blake Leibel (Hamilton, 2016), Hasib Bin Golamrabbi (Mejia & Rocha, 2016) and the killer of Karen Perez (Harris, 2016) which involved cannibalism, dismemberment, exsanguination, scrawlings, strangulation during rape or outward statements of idolization of other serial killers. These cases could prove problematic for the ‘Weather Map’ as they may be scraped and cataloged along with other serial killers as the program will assume these victims are part of a series of murders. The ‘Weather Map’ may also capture offenders like Vernon Primus due to his link as a person-of-interest in open homicide cases raised speculation of his ‘serial killer’ status (Murphy, 2016).

Welcoming Contributions from the Uninitiated

As the nation's crime monitoring system struggles to move into the 21st century (Rosenfeld, 2016) there are still many areas in which contributions can be made by those adjacent to the realm of criminology. Boyd (2016) stresses shunning the conventions of an academic infrastructure which compels the pursuit of acceptable lines of inquiry to obtain kudos, emboldening researchers to be aware of the context they finds themselves and to pay attention to the incentives and pressures that influence their decisions while developing impactful questions that matter regardless of funding battles or the fads and obsessions of their disciplines. Leetaru (2016) calls for understanding the nuances of data but warns that answering a question of interest may take more than plugging a dataset into an algorithm. Because deep-learning software recognizes patterns in digital representations of data (Hof, 2013) it can be used to overcome noted drawbacks and unite those with institutional knowledge about MeMs with those that understand how they actually behave. This reconciliation would place researchers on the forefront of an indispensable defense science for decades to come.

The variable demands of academic life have transformed higher learning into an unattractive pursuit for those looking for a career suffuse with field work. Defending new science to peers can create an atmosphere of animosity and study findings may take years to be adopted. The stress experienced by young academics was noted as one of seven challenges facing the sciences (Belluz, Plumer, & Resnick, 2016). The restrictive path of scholarship can be overcome when the study of atypical homicide is approached as a neophyte or outsider. Boyd (2016) characterizes academic disciplines as myopic and judgmental of those operating outside the acceptable boundaries and encourages those from diverse experiential, cultural and political backgrounds to contribute. This is an area ripe for disruption as important research findings can be made readily, such as the discovery that most partners in serial killer teams fit into one of three roles that aid the Principal Offender: the Enlisted Accomplice, the Witting Facilitator and the Idle Witness (Yaksic & Hargrove, 2016). Hester Brink (personal communication, 5 July 2016) is collecting the input of outsiders for use in the investigation of cold cases while other research avenues focus on the interplay between a genetic predisposition to violence and environmental factors (Parshley, 2016) and the roles depression (Levenson, Ransom & Crimaldi, 2016; United States Department of Justice, 2009), emotional regulation (McGinty, Kennedy-Hendricks, Choksy, & Barry, 2016) and mental illness (Eyster, 2012; Gelinas & Hadjistavropoulos, 2016) play in the lives of atypical murderers. Pioneering work is underway to deduce the likelihood that the decrease in serial homicide is correlated with an observed increase in spree and mass homicides. The projects outlined in this chapter have built on the work of Cai (2015) and demonstrate the need for engagement in practices that will enable the field to operationalize data for the social good.


This chapter could not have been completed without the contributions of the following individuals: Mike Aamodt, Dalal Alrajeh, Peter Brendt, Joel Caplan, Thomas Hargrove, Joseph Johaneman, Rebecca Kramer, Ravi Shroff and Angela Williamson. This chapter is dedicated to the memory of Michelle McNamara.


Aamodt, M., & Yaksic, E. (2015). Serial murder: Separating fact from fiction. Webinar hosted by the Justice Clearinghouse. Retrieved from

Alexander, K. L., & Bui, L. (2016). Suspect lost eyeglasses and was unable to flee after shooting, prosecutors say. The Washington Post. Retrieved from

Allely, C. (2016). Can we predict who will become mass shooters? The Conversation UK. Retrieved from

Angwin, J., Larson, J., Mattu, S., & Kirchner, L. (2016) Machine Bias. ProPublica. Retrieved from

Associated Press. (2008). "Black Widows" Guilty In Homeless Murders. Retrieved from

Baun, C. (2016). Mobile clusters of single board computers: an option for providing resources to student projects and researchers. Springerplus. 5:360.

Belluz, J., Plumer, B., & Resnick, B. (2016). The 7 biggest problems facing science, according to 270 scientists. Vox. Retrieved from

Blassmann, A. (1999). The Detective in 'Twin Peaks'. Thesis for Freiburg University. Retrieved from

boyd, d. (2016). We Are to Blame for the State of Social Science Research. Items. Social Science Research Council. Retrieved from

Boyne, E. S. (2014). Serial homicide collaborative brings research data together. Criminal Justice Update (CJ Update), An Online Newsletter for Criminal Justice Educators. From Routledge and Anderson. 43(1), 2. Retrieved from

Branson, A. (2013). African American serial killers: Over-represented yet under acknowledged. The Howard Journal of Criminal Justice. 52(1):118.

Brendt, P. (2016). The Serial Homicide Offender ‘Weather Map’. [White Paper].

Cai, S. (2015). Mining Modus-operandi Patterns of Swedish Serial Burglaries. [White Paper]. Retrieved from

Caldwell, S. (2016). We Know You Gotta Satisfy Your Hamilton Fix, So Here Are Two Great Videos. Vulture. Retrieved from

Caplan, J. M., & Kennedy, L. W. (2016). Risk Terrain Modeling: Crime Prediction and Risk Reduction. CA: Univ. of California Press.

Carter, M. (2016). Man charged with slaying of Ingrid Lyne tells police: I’m ‘not a normal person’. The Seattle Times. Retrieved from

Chammah, M., & Hansen, M. (2016). Policing the Future. The Marshall Project. Retrieved from

Copeland, MV. (2012). Resiliency, Risk, and a Good Compass: Tools for the Coming Chaos. Wired Magazine. Retrieved from

Crimesider Staff. (2016). Cops: Killing spree suspect inspired by movie "The Purge". Retrieved from

Crawford, K. (2016). Artificial Intelligence’s White Guy Problem. The New York Times. Retrieved from

Crawley, M. (2015). Thomas Hargrove’s award-winning methods. Scripps News Online. Retrieved from

Crime Linkage International NetworK (C-LINK). (2016). Retrieved from

D’Ambrosia, L. (1998). Read A&E Cold Case File FDLE Criminal Profile on EAR-ONS. Florida Department of Law Enforcement. Retrieved from

Davey, M. (2016). Chicago Police Try to Predict Who May Shoot or Be Shot. The New York Times. Retrieved from

DeLisi, M. (2015). Mayhem by Occupation, in Sex Offenders: A Criminal Career Approach (eds. Blokland, A., & Lussier, P.), John Wiley & Sons, Ltd, Chichester, UK.

DeSpirito, L. (2016). Is There a Decline in Serial Homicide? Unpublished manuscript. Northeastern University, Boston, MA.

de Zoete, J., Sjerps, M., Lagnado, D., & Fenton, N. (2015). Modelling crime linkage with Bayesian networks. Science and Justice. 55(3):209-17.

Donoho, D. (2015). 50 years of Data Science. Princeton NJ, Tukey Centennial Workshop. Retrieved from

Editor & Publisher. (2010). Hunting Serial Killers: a Scripps Howard News Service Investigation. Retrieved from

Ellwood, M. (2015). 'American Horror Story: Hotel' Recap: Calling all serial killers. Condé Nast. Retrieved from

Everett, C. (2012). ‘Wanderlust’ star Jennifer Aniston: My boyfriend Justin Theroux looks ‘like a serial killer’. New York Daily News. Retrieved from

Eyster, D. (2012). District Attorney Determines Fatal Officers-Involved Shooting Of Serial Killer Outside of Fort Bragg Was Legally Justified. Office of the District Attorney County of Mendocino. Retrieved from

Farrell, A. L., Keppel, R. D., & Titterington, V. B. (2011). Lethal Ladies Revisiting What We Know About Female Serial Murderers. Homicide Studies 15(3):228-252. 

Felton, R. (2016). Kalamazoo shooter saw 'devil' on Uber app and blames visions for killing spree. The Guardian. Retrieved from

Forman, R. (2016). Crime Prevention by the Numbers. Rutgers School of Criminal Justice Newsletter. Retrieved from

Fung, B. (2016). How artificial intelligence could help warn us of another Dallas. The Washington Post. Retrieved from

Gao, J., Barzel, B., & Barabási, A.L. (2016). Universal resilience patterns in complex networks. Nature. 530, 307–312.

Gelinas, B. L., & Hadjistavropoulos, H. (2016). "Am I Becoming a Serial Killer?" A Case Study of Cognitive Behavioral Therapy for Mental Illness Anxiety. Behav Cogn Psychother. 44(3):374-9.

Halber, D. (2013). Web sleuths help solve cold cases. The Boston Globe. Retrieved from

Halber, D. (2016). Profiler 2.0. Boston Magazine. Retrieved from

Hall, E. (2013). Meet the “Hannibal” Fannibals, TV’s Newest and Most Intense Fandom. Buzzfeed. Retrieved from

Hamblin, J. (2016). Toxic Masculinity and Murder. The Atlantic. Retrieved from

Hamilton, M. (2016). Her blood was 'drained' from her: Canadian heir charged with torture killing of girlfriend in WeHo. Los Angeles Times. Retrieved from

Hamilton, M., Watanabe, T., & Winton, R. (2016). For UCLA shooter Mainak Sarkar, sudden rage after years of intense academic studies. Los Angeles Times. Retrieved from

Hargrove, T. (2014). Indiana police were contacted about a potential serial killer back in 2010. Scripps News. Retrieved from

Hargrove, T. (2015). Murder Accountability Project. Retrieved from

Hargrove, T., Witzig, E., Icove, D. J., Harry, B. E., Arntfield, M. A., Yaksic, E., Lang, H., & Wolf, I. (2016). Accounting for Murder: A New Tool for Homicide Investigators. Federal Bureau of Investigation Law Enforcement Bulletin. Forthcoming.

Harris, C. (2016). 'I Don't Want to Die': Boyfriend Allegedly Filmed Rape and Murder of 15-Year-Old Girl Found Under Sink. People Magazine. Retrieved from

Hof, R. D. (2013). Deep Learning: With massive amounts of computational power, machines can now recognize objects and translate speech in real time. Artificial intelligence is finally getting smart. MIT Technology Review. Retrieved from

Holman, R. (2016). Cedric Ford case shows need for law enforcement to communicate. The Wichita Eagle. Retrieved from

Johnson, N. F., Zheng, M., Vorobyeva, Y., Gabriel, A., Qi, H., Velasquez, N., Manrique, P., Johnson, D., Restrepo, E., Song, C., & Wuchty, S. (2016). New online ecology of adversarial aggregates: ISIS and beyond. Science. 352(6292):1459-1463.

Kimble, L. (2015). Steve Carell: 40-Year-Old Virgin nearly shut down because he looked like serial killer. People. Retrieved from

Kirby, S., Graham, J., & Green, M. (2014). The Cumbria spree killing – how mobility affects the policing of critical incidents. International Journal of Emergency Services. 3(1):34-48.

Kramer, J. (2016). Dr. Henry Lee Profiles Today’s Serial Killers. CT News Junkie. Retrieved from

Kuhn, J., & Coston, C. (2005). The Myth That Serial Murderers are Disproportionately White Males, in Demystifying Crime and Criminal Justice (eds. Bohm, R. M., & Walker, J. T.). Oxford University Press, New York, NY. Retrieved from

Kurkjian, A. (2016). How Serial Sexual Murderers Are Apprehended (Unpublished thesis). John Jay College of Criminal Justice, New York, NY.

Lange, R. (1999). A Cusp Catastrophe Approach to the Prediction of Temporal Patterns in the Kill Dates of Individual Serial Murderers. Nonlinear Dynamics, Psychology, and Life Sciences. 3(2):143-159.

Laub, J. H. (2012). Translational Criminology. Translational Criminology. Retrieved from

Leetaru, K. (2016). Are We Mining Data Instead Of Answering Questions? Forbes. Retrieved from

Levenson, M., Ransom, J., & Crimaldi, L. (2016). Taunton attack victims are remembered. The Boston Globe. Retrieved from

Los Angeles Times. (2010). Map: Serial killers in South L.A. Retrieved from

Mains, K. L. (2015). Ken’s Corner. Retrieved from

McGough, M. (2016). Job Opportunity! NIJ seeks a social science researcher with experience partnering with law enforcement to lead a groundbreaking new initiative. Office of the Director National Institute of Justice. Retrieved from

Mejia, B., & Rocha, V. (2016). Brothers arrested after parents' bodies found next to note: 'Sorry, my first kill was clumsy'. Los Angeles Times. Retrieved from

Mettler, K. (2016). The crime spree began in a cemetery. An hour later, two men in their 80s were dead. The Washington Post. Retrieved from

Meyjes, T. (2016). Serial killer-obsessed schoolboy jailed for 27 years after murdering two strangers. Metro. Retrieved from

Murphy, M. (2016). NYPD investigates whether Caribbean murder suspect is serial killer. PIX11 Investigates. Retrieved from

Muller, D. A. (2000). Criminal Profiling Real Science or Just Wishful Thinking? Homicide Studies, 4(3), 234-264.

Myers, W. C., Husted, D. S., Safarik, M. E., & O'Toole, M. E. (2006). The motivation behind serial sexual homicide: is it sex, power, and control, or anger? J Forensic Sci. 51(4):900-7.

National Institute of Justice. (2016). The Law Enforcement Advancing Data and Science Program. Office of Justice Programs. Retrieved from

Office of the University Registrar. (2016). Course Descriptions. Northeastern University. Retrieved from

Parshley, L. (2016). Can Your Genes Make You Kill? Science's search for the roots of violence. (2016). Popular Science. Retrieved from

Peters, J. (2013). Can a Group of Amateur Internet Detectives Catch the Golden State Killer? Slate. Retrieved from

Radar Staff. (2014). Satanist Cannibal Killer Gregory Hale Emulated The Night Stalker – ‘I Am Beyond Good And Evil’. Radar Online. Retrieved from

Raine, N. E., Rossmo, D. K., & Le Comber, S. C. (2009). Geographic profiling applied to testing models of bumble-bee foraging. Journal of the Royal Society Interface. 6:307-319.

Rosenfeld, R. (2016). Documenting and Explaining the 2015 Homicide Rise: Research Directions. National Institute of Justice. Office of Justice Programs. Retrieved from

Rossmo, D. K. (1995). Place, space, and police investigations: Hunting serial violent criminals, in Crime and place: Crime prevention studies (eds. Eck, J. E., & Weisburd, D. A.). 4:217-235. Monsey, NY: Criminal Justice Press. Retrieved from

Sanchez, R., Berlinger, J., & Flores, R. (2016). Who was Kansas shooter Cedric Ford? CNN. Retrieved from

Scherer, J. A., & Jarvis, J. P. (2014). Criminal Investigative Analysis: Practitioner Perspectives (Part One  of Four). FBI Law Enforcement Bulletin. Retrieved from

Schram, J., Algar, S., & Tacopino, J. (2015). Busted ex-police chief blocked FBI probe of Gilgo Beach murders. New York Post. Retrieved from

Shaer, M. (2016). The False Promise of DNA Testing. The Atlantic. Retrieved from

Shanafelt, R., & Pino, N. W. (2013). Evil and the common life: Towards a wider perspective on serial killing and atrocities. New Directions in Crime and Deviancy, 252-273.

Shanafelt, R., & Pino, N. W. (2015). Rethinking Serial Murder, Spree Killing, and Atrocities: Beyond the Usual Distinctions. New York: Routledge.

Shifman, P., & Tillet, S. (2015). Op-Ed. To Stop Violence, Start at Home. The New York Times. Retrieved from

Simkin, M. V., & Roychowdhury, V. P. (2014). Stochastic modeling of a serial killer. J Theor Biol. 355:111-6.

Sorochinski, M., Salfati, G., & Labuschagne, G. N. (2015). Classification of Planning and Violent Behaviours in Serial Homicide: A Cross-National Comparison Between South Africa and the US. Journal of Investigative Psychology and Offender Profiling. 12(1):69–82.

Strand, G. (2016). Killers on the Road: Driving's Link to Violence. Opinion. National Geographic. Retrieved from

Swaine, J. (2016). No one except @thecounted, @fatalencounters and @killedbypolice 8:22 AM - 20 Apr 2016 Tweet. Retrieved from

Tiffany, K. (2016). Who called Ted Cruz the Zodiac Killer, why, and is he? The Verge. Retrieved from

Truesdell, J. (2016). Murder of Seattle Mom Ingrid Lyne Leads Investigators to Ask: Could Suspect Be a Serial Killer? People Magazine. Retrieved from

United States Department of Justice. Federal Bureau of Investigation. (2009). Transcription of James E. Childers Interview. CASE: 308J-PG-78010. The Inter-Mountain. Retrieved from

United States Department of Justice. Federal Bureau of Investigation. (2016). Violent Criminal Apprehension Program. Part 1: Sharing Information to Stop Serial Offenders. Retrieved from

Wolcott, J. (2016). Is American Psycho Too Bloody for Broadway? Vanity Fair. Retrieved from

Walters, B. K., Drislane, L. E., Patrick, C. J., & Hickey, E. (2015). Serial murder: Facts and misconceptions. Science and the Courts, 1(5), 32-41.

Warren, J. I., Dietz, P. E., & Hazelwood, R. R. (2013). The collectors: Serial sexual offenders who preserve evidence of their crimes, Aggression and Violent Behavior. 18, 666-672. 

Wiest, J. B. (2016). Casting cultural monsters: Representations of serial killers in US and UK news media. Howard Journal of Communications, 27.

White, J. H., Lester, D., Gentile, M., & Rosenbleeth, J. (2011). The utilization of forensic science and criminal profiling for capturing serial killers. Forensic Science International. 209:160–165.

Yaksic, E. (2006). Can a demographic make you psychopathic? Poster presented at Northeastern University Research Expo. Retrieved from

Yaksic, E. (2013). Breaking the stereotype of a serial killer. Op-Ed. Clarion Ledger. Retrieved from

Yaksic, E. (2014a). A review of Why We Love Serial Killers: The Curious Appeal of the World’s Most Savage Murderers. Journal of Forensic Research and Criminal Studies. 1:1-5.

Yaksic, E. (2014b). A guest post: The Felix Vail case. Retrieved from

Yaksic, E. (2015) Addressing the challenges and limitations of utilizing data to study serial homicide. Crime Psychology Review. 1(1):108-134.

Yaksic, E., & Hargrove, T. (2016). Killers in pairs and applications of gathered data. Presentation at the Confronting Homicide in a Changing World Homicide Training Seminar, Green Bay, Wisconsin.

[1] bootscallahan (22 July 2016). [TUTORIAL] Welcome to the missing & unidentified subreddit! Message posted to reddit. Available at
[2] Fox and Yaksic (2015). Northeastern Universitys Atypical Homicide Research Group. Available at:
[3] Hilts, Mark. Unit Chief of the FBIs Behavioral Analysis Unit 2. Interview conducted by Thomas Hargrove. Scripps Howard News Service. 22 April 2010.
[4] Hargrove, T. (15 April 2016). A prolific serial killer in Atlanta? Message posted to Northeastern University’s Atypical Homicide Research Group electronic mailing list, archived at Homicide@Listserv.NEU.EDU
[5] Harry, B. (31 May 2016). Drowning young men. Message posted to Northeastern University’s Atypical Homicide Research Group electronic mailing list, archived at Homicide@Listserv.NEU.EDU