Behavioral Signatures, Modus Operandi Evolution, and Geographic Profiling Across the 20 Most Prolific Serial Killers: A Mixed-Methods Comparative Study

Serial homicide research has produced a wealth of individual case studies and small-sample typological analyses, yet systematic, internationally comparative, multi-method investigations of the most prolific offenders remain strikingly scarce. The organized–disorganized typology and Canter’s circle hypothesis stand as the dominant analytical frameworks, but researchers have rarely tested them together on an internationally diverse sample of extreme-prolificacy offenders. This study examined behavioral signatures, modus operandi evolution, and geographic profiling patterns across the 20 most prolific serial killers in the historical record, a sample spanning 12 countries across six continents. A convergent parallel mixed-methods design integrated qualitative behavioral content analysis with quantitative geographic and statistical analysis. Data drawn from court records, official inquiries, academic case literature, and investigative journalism were coded across four behavioral domains: modus operandi elements, signature behaviors, escalation indicators, and geographic variables. Statistical analyses included Mann–Whitney U tests, Kruskal–Wallis H tests, and Spearman rank correlations, and behavioral coding reliability was assessed using Cohen’s kappa and intraclass correlation coefficients. The prediction that organized offenders would display greater behavioral consistency than disorganized offenders (H1) was not supported, U = 16.5, p = .874, r = .056; both groups yielded equivalent mean Signature Richness Scores, M = 5.00. By contrast, the prediction that hedonistic-sadistic offenders would show greater career escalation than power/control offenders (H2) was supported with a large effect, U = 61.0, p = .014, r = .552. The predicted marauder majority (H3a) did not emerge; commuters (45%) slightly outnumbered marauders (40%), and all three disorganized offenders were commuters. The expectation of greater geographic range among organized offenders (H3b) was likewise not supported, U = 15.0, p = .694, r = .112. Integrated behavioral-geographic profiling (H4) was partially supported: 45% of cases showed non-redundant investigative value alongside 80% concordance. A near-significant positive correlation between Signature Richness and the Escalation Index (ρ = .422, p = .064) pointed to a theoretically meaningful trend. Taken together, these findings challenge the categorical utility of the organized–disorganized typology for extreme-prolificacy populations while offering robust cross-cultural support for fantasy-driven escalation in sadistic serial offenders. The commuter majority and the disorganized-commuter paradox unsettle circle theory assumptions and indicate that geographic profiling models require revision for this population. The study discusses concrete implications for investigative practice, database development, profiling training, and future research.