Quantitative Identification Criteria for Secondary Toolmarks on Bullets Shot from AK-74

Abstract

The paper compares the effectiveness of identifying low-count land marking patterns using the method of consecutively matching striations (CMS method) and the probabilistic method. The study was conducted on 48 pairs of matching toolmarks and 66 pairs of non-matching toolmarks. The criterion for assigning a positive match that was determined using the CMS method yielded one false positive in the analysis of known non-matches. Approximately 50% of matching pairs of toolmarks identified by the expert were omitted in the analysis. The application of the probabilistic method showed that out of the same 16 pairs of matches identified by the expert, no pair will be missed if the probability value of 5 х 10-3 is taken as the identification criterion. No false positive statement will be made either. Thus it is shown that the probabilistic method of toolmark identification is more effective. It was demonstrated to be practically impossible to justify elimination when comparing low-count toolmark patterns.

Keywords

• land markings
• consecutively matching striations
• poisson distribution
• firearms identification
• forensic ballistics
• matching toolmarks