Using a Simple Math Tool to Judge Shoe Prints in Court

In England and Wales, people who look at shoe prints left at crime scenes often face a tricky job. They have to decide how likely it is that a particular pair of shoes made the mark, and then explain their thoughts to judges and juries. The way they do this can make the evidence look strong or weak, which matters a lot in court. Right now, most experts use a step‑by‑step method that can be confusing. They compare the print to many known patterns and then give a rough estimate of how good their match is. This can lead to mistakes or disagreements, especially when the evidence is not clear. A new idea uses something called a Bayesian Network. Think of it like a decision tree that shows how different pieces of information connect. For example, one branch might ask whether the shoe size is common in the area; another branch could check if the tread pattern matches a known brand. By feeding real data into this tree, the expert can get a numerical score called a likelihood ratio. This number tells how much more likely it is that the suspect's shoes made the print versus any other shoe.The paper shows several sample trees built with data from real crime labs and opinions of a current UK shoe‑print expert. These examples illustrate common problems, like missing size information or unclear tread wear, and show how the tree can still give useful guidance. When the researchers tested their model with a real court case, they found that it made the expert’s reasoning easier to follow and less open to misinterpretation. The authors also ran a sensitivity check. They changed the input numbers from different sources—such as police reports, shoe‑stores, and other labs—to see how the likelihood ratio would shift. The results gave a range of possible values, so lawyers could present upper and lower bounds instead of one single figure. This adds honesty to the evidence. In short, using a Bayesian Network could help forensic shoe‑print experts explain their findings more clearly and make the legal process fairer. It also shows a way other forensic fields could adopt similar tools to improve transparency.