# Math in forensics: Sherlock Holmes on steroids

**Probability and statistics**

The following are excerpts taken from “Essential Mathematics and Statistics for Forensic Science” by Craig Adam (Wiley-Blackwell)

Analysis and prediction using probability and statistics are becoming more a part of law enforcement investigation around the world. Fingerprints and DNA are only the tip of the iceberg. A forensic scientist’s work is different from other analytical scientists in that the results of the work is reported to a court of law, with the forensic scientist being the expert witness and subject to cross-examination.

Let’s look at the probabilistic basis for interpreting evidence. The probability of the evidence being at the crime scene (E) given the guilt of the accused may be written as P(E|G). We need to compare the alternative probability of the evidence given that the accused is innocent or not guilty (Ḡ), expressed as P(E|Ḡ). If the ratio of these alternatives is greater than one, then the proposition based on guilt is weighted more than that based on innocence and vice-versa.

Here, the likelihood ratio (LR) is created:

Now we need to consider each of these hypotheses to ensure that they are mutually exclusive and to facilitate the correct evaluation of each probability given specific data. So we now a more general for the LR where the mutually exclusive propositions H_{1} and H_{2} are to be defined:

Here H_{1} is the hypothesis that the prosecution would present and H_{2} for the defense, where H_{1} and H_{2} are mutually exclusive. Chapter 11.2.1 “Estimation and calculation of likelihood ratios” shows some concrete scenarios and examples. Then, the book moves on to Likelihood ratio, Bayes’ Rule and weight of evidence. Population data and interpretive databases in which frequency histograms are constructed from large amounts of data that describe the occurrence of some measurable characteristic within a certain specified population.

I will not go into the details here, but suffice it to say that most basic calculations are straightforward. The problem comes in when trying to understand how these methods should be applied and finding the meaning of the result. The referenced book explores these issues qualitatively using case studies as illustrations.

**Angles and triangles**

Blood spatter can be analyzed by trigonometry. Measuring the angles and distances and then calculating the third point of the triangle----the location of the tracker, is commonly done.

Ballistics trajectories like those from a shotgun can determine distance from the victim to the shooter, the shooter’s approximate height and information about the type of shotgun.

**Blood stain formation**

Math can determine, using fluid dynamics and equations, details about whether the blood is a droplet or a high velocity impact-based spatter. The direction of the impact can also be determined. Two key equations are used for the stain diameter and number of spines respectively:

Variables are:

*D* is the blood droplet diameter

*v* is velocity

Constants are:

ρ is density

η is viscosity

γ is surface tension

Methods for graphical analysis can be verified from these formulae to make calculations and predictions within the range of test data and to some extent outside of it.

A really good reference paper is “Deducing Drop Size and Impact Velocity

from Circular Bloodstains” from the Journal of Forensic Science

Please share your experience and expertise on this subject with our EDN audience

Lagrangian Dynamics and Advanced Differential Equations are not for everyone

LIGO Laboratory mathematics: detecting gravitational waves

Math and logic in the Venus flytrap

The logarithm and the flogarithm and don’t play Powerball

Attacking Ebola mathematically

Active and reactive power flow and flicker on the grid

The Pythagoreans' beautiful fallacy

From Analog Devices we have FMCOMMS Math

Math in forensics: Sherlock Holmes on steroids

Accidental engineering: 10 mistakes turned into innovation

That 60W-equivalent LED: What you don’t know, and what no one will tell you…

6 famous people you may not know are engineers

DC distribution in your house and 42-V cars

10 tips for a successful engineering resume

The 5 greatest engineers of all time

Higgs Pt. 9: What makes King Carl XVI Gustaf think it’s the Higgs Boson?

10 things you may not know about Tesla

Analog Fundamentals: Instrumentation for impedance measurement

Why do users buy test gear? – Part 3: Manufacturing

Changing roles for embedded developers

Are we ready for fusion reactors in space?

ESC Silicon Valley: Be creative by goofing off

7 steps to reusable debounce (and other) code

Connected cars: IoT’s value proposition