NUMB3RS - Season 2 - "Judgement Call" - Filtering Suspects

To begin our investigation, let's analyze the profile database of six prime suspects. Initially, without any clues or forensic evidence, each suspect is assumed to have an equal chance of committing the crime. This initial baseline probability is called the prior probability.

Suspect Profiles

Suspect	Height	Weight	Hair Color	Facial Hair
Suspect 1	6 ft 3 in	220 lbs	Blonde	Beard
Suspect 2	5 ft 8 in	205 lbs	Black	Beard
Suspect 3	5 ft 4 in	160 lbs	Blonde	Goatee
Suspect 4	6 ft 1 in	190 lbs	Brown	Beard
Suspect 5	6 ft 1 in	215 lbs	Red	Goatee

If we assume only one of these six suspects committed a given crime, the prior probability for any specific suspect is:

$P(S_i) = \frac{1}{6}$

When we receive new, reliable evidence, we can filter our sample space. This changes our calculations to conditional probability: $P(A|B)$, read as 'the probability of event $A$ occurring given that event $B$ is true.'