Explain Gain(A=>B) which is P(B|A) - P(B) to a layman

50 Views Asked by At

Let's say that we have a population of colored balls, some of which are defective.

Let $A$ be the event of picking a red ball and $B$ be the event of picking a defective ball.

Based on given data, lets assume that $\Pr(A) = 0.2, \Pr(B) = 0.3$, and $\Pr(B \mid A) = 0.1$

Therefore, $$\operatorname{Gain}(A \implies B) = \Pr(B \mid A) - \Pr(B) = -0.2$$

Conclusion: The probability of observing a defective ball reduces by 0.2 among the red balls, in comparison to the population of all balls.

Question:

1) Is there an easier term than $\operatorname{Gain}$ that explains this to a non-math audience?

2) Is there any known simpler language to express the conclusion than what I have used.

NOTE: While the language in the conclusion seems obvious enough for the example above, my actual use case is a business application where more accessible language might be warranted.