Area of Opportunity

Area of opportunity is a statistical concept that is not normally relevant to fingerprinting because inquiries normally start from a crime. When there is no evidence that an act of wrongdoing has occurred when an identified person denies depositing a fingerprint a trier of fact could easily come to the wrong conclusion by not understanding the concept. For example it would be very easy for someone to think:

(a) it is not unlikely for a police officer to enter a crime scene without permission and lie about it, and

(b) misidentifications are very rare.

(a) and (b) could have different areas of opportunity in the thinker’s mind. Thought (a) might be embracing a large area – “it is not unlikely that a police officer might enter a crime scene - somewhere”. Thought (b) might be only be embracing the immediate situation in front of them – “the chances of a misidentification involving the people in front of me are small”.

The correct way to think about this situation is:

(a) it is not unlikely for a police officer somewhere in the world to enter a crime scene without permission and lie about it and

(b) it is not unlikely for a misidentification to occur somewhere in the world

so when a denied identification is encountered, it is the balance of these two similar probabilities that the officer is lying. Another correct line of thinking would be:

(a) without the fingerprint evidence there is no reason to believe that an unauthorised police officer entered this crime scene, so that is very unlikely to have happened here, and

(b) the chances of a misidentification in any one case are small so that is very unlikely to have happened here

so again it is the balance of two similar probabilities.

A statistician would say the prior probability of a police officer entering the crime scene and lying about it when there is no evidence to support the proposition is very low. This is transformed into a medium posterior probability after a fingerprint ID which has a very low probability of error.

It would be very wrong to think:

(a) even though there is no physical evidence that a police officer entered the crime scene we have a fingerprint ID so that makes it not unlikely. And compare that with

(b) misidentification is very unlikely

(a) on its own is OK but not if it is being compared with (b), that would be applying the fingerprint evidence twice. “Not unlikely” is not enough on its own to accuse someone of lying.

Nobody needs to think about areas of opportunity when the case starts from a physical act of wrongdoing because everything relates to the local case.

(a) we know the criminal was in the crime scene so there is a fair chance that he or she deposited a fingerprint and

(b) the chances of a misidentification in any one case are small

That is not a balance of similar probabilities.

In statistical speak, a medium prior probability that the criminal left a fingerprint while in the crime scene is transformed into a very high posterior probability after a fingerprint ID which has a very low probability of error.

Back to previous document