Today's WOTD was the usage of statistic in various stock phrases along the lines of "I don't want to be a statistic," or "Don't be a statistic." The rationale for hating on this was that, according to the email read on the show, that you cannot avoid being a statistic, that no matter what you do, you are part of one statistical group or another.
This brings in one of the most common fallacies by usage mavens and regular folks everywhere -- trying to apply mathematical logic to language. It's the same logic that is used to argue against "double negatives" (which I prefer to call negative concord or negative agreement, but I won't get into that here) by claiming that "two negatives equal a positive". In this case, the peevologist is applying a strict definition of statistic something along the lines of "a member of a statistical group". I would argue that there are two more useful ways of approaching this problem:
- You could propose that statistic has a secondary, figurative meaning of "someone who, through action or inaction on known risk factors, has put themselves in a negative statistical groups (ex. smokers with lung cancer). This allows us to explain these various phrases all at once, though it does require the qualifier that this usage is fairly restricted.
- Alternately, you could consider the phrase be a statistic is an idiom. In linguistics, an idiom is a phrase that has a meaning that cannot be arrived at by analyzing the components. For example, nothing in the idiom kick the bucket tells us that death is involved, native speakers simply memorize the definition "to die" for the whole phrase.