There have been much criticism of how accurately evidence can be presented in the field of fire arms and tool mark evidences. One of the main concerns within this field is that the community has not yet succeeded in supplying the “known or potential error rate. From a statistical point of view, the nature of tool mark surfaces makes probability densities nearly impossible to estimate in an accurate manner. Random match probabilities cannot be estimated easily because of the spectrum of variability of data. Continual effort has been made in order to estimate RMPs and identification error rates. Quantifiable measures of uncertainty of forensic analyses have been recommended by the community, but the relevance of these mathematical functions is another matter. ”. Critics argue that this field does not have a strong statistical basis of tool mark identification. This argument is not a statement of fact and is not even a well-grounded argument due to lack of logic. One point that shoots down this argument is the fact that it does not address whether or not firearm and tool mark identification actually works. A second point to discount this argument is that this claim has not been proven to be a statement of fact. A third point is that this criticism assumes that examiners will testify to an absolute certainty rather than a practical certainty. Statistical models have in fact been implemented to firearms and tool mark identification. However, each model has its own limitations. The relevancy of these models to this field have not yet been determined. This is also true in various other fields of forensic science such as fingerprints, trace evidences, shoe prints or tire tracks. Forensic science aside, other important aspects lack robust models such as the appointment of judges, the prediction of weather and the future of criminal behavior of a convict.
The evolution of firearm and tool mark identification has been in the works since as early as 1932. Various studies have been made to create and, at the same time, test the statistical probabilities of finding firearm or tool mark striation matches by random chance. Separate studies were developed to account for lack of empirical based studies within this field. The progression of technology has led to examiners being able to scan and analyze firearm and toolmark RMPs in three dimensional nature. The potential creation of automated computer programs is the result of this technological advancement. These programs can scan for RMPs and allow the decreasing human error.
A study conducted by Weller “used confocal microscopy to analyze breech face markings on 90 cartridge cases test fired from ten consecutively manufacture pistol slides.” This research pointed out a clear separation of the matching and nonmatching test fires within the experiment. A slight overlap was also indicated from the results. The calculated RMP that resulted from this study turned out to be 1 in 3.00 *10^(-5). Gambino and Petraco used statistical pattern recognition methods that were both similar in nature to both striation marks and primer shear marks on 9mm cartridge cases. In both studies done, the rates that were calculated low appeared to be at a very low numerical value (less than .02 in both cases).
Current State of the Use of RMPs in Firearm and Toolmark Identification
In this section of the article, RMPs are discussed in how they are usually “error rates” and how these rates are difficult to establish through the process of firearm and tool mark identification. It is also discussed how it is currently unknown whether it is feasible to apply RMP to the firearms and tool mark identification. Tool markings play a key role in the reasoning for why RMPs are not 100 % reliable. The level of difficulty to develop a universal mathematical model that can accurately predict the random tool marks left by some tools is also discussed.
Absolute Versus Practical Identification and Subjectivity
The final section informs us on the relevancy of certain evidence. For example, the writers talk about the concept of uniqueness and the allowed use of “absolute doubt” but the forbidden use of “reasonable certainty.” The authors of these studies continue on to say that subjective aspects in firearms and tool mark identification are not viewed positively in examination of these two. It is argued that subjective viewpoints are only to be used when examiners determine similarities are sufficient enough for identification. The authors finish their argument stating that subjectivity has a use in all actions and opinions of both the system of justice and in our common livelihood.