Sep 272016

“Most of his attempts usually succeeded.”

I read this sentence in a well-respected newspaper this morning. What does it mean?

If most means, say, 51% or more, and if usually also means 51% or more, then the sentence is saying that 51% of his attempts succeed, 51% of the time.

If you’ve studied any statistics / probability at all, you know that this is equivalent to saying that his attempts succeed 26.01% (or more)  of the time. (0.51 multiplied by 0.51 is 0.2601).

“Most of his attempts usually succeeded” sounds a whole lot stronger than “His attempts succeeded more than 26% of the time.” Right?

Even if we ramped up the definitions of most and usually to mean 70% and 70%, respectively, we still end up being able to say that his attempts succeeded more than 49% of the time.

Most and usually are both considered to be absolutes. The writer (or speaker) has to be very careful when using absolutes. If I walk out of Kroger thinking, “Man, most of the stuff in this store is overpriced,” but I base that only on the 4 or 5 items I purchased, then my use of the word most is highly inappropriate.

I watched the presidential debates last night on television. Both candidates were freely using absolutes … “most people want” … “I never would” … “we always should” … and phrases like that. I wondered how much research (or thought) had gone into backing those statements.

Others absolutes include words like: always, every, all, never, none, often, and more. (I don’t mean more in the .etc sense, but in the more is one of those words, too sense.)

The problem (if there is a problem) here is that most and usually are both used in the same sentence.

Wouldn’t the sentence mean the same thing if I said:

Most of his attempts succeeded.


His attempts usually succeeded.

Saying that most of his attempts usually succeeded kind of clouds the intent (I think), unless the writer intentionally tried to mislead. That (trying to mislead) was not the intent of the writer (at least, based on the tenor of the article).

It becomes more obvious (does it really?) that something is askew when we say:

All of his attempts usually succeeded. Maybe this does make sense. All of his attempts—that is, every time he tried to do whatever it was he was trying to accomplish—more often than not, he was successful.

All of his attempts never succeeded. If we can associate never succeeded with failed, then this is pretty cut-and-dried. All of his attempts failed.

Most of his attempts always succeeded. Similarly, if always succeeded is associated with passed, then it’s easy to see that most of his attempts passed.

None of his attempts usually succeeded. At first glance, I was tempted to say that this means the same as none of the attempts succeeded. But, does it mean that? Even if none of the attempts usually succeeded, could there be attempts that occasionally succeeded?

None of his attempts always succeeded. Again, are there attempts that sometimes succeeded, even if none always succeeded?

That begs the question: What does attempt mean in this context?

Let’s try an exercise (I’m doing this with you … I haven’t planned this out in advance).

Suppose Hank Aaron is our subject for this endeavor. Hank had a lifetime batting average of .305. Yes, he was a great, solid hitter. (And he’s still the all-time home run leader, in my book.)

Now, suppose we could look back at statistics for Hank’s individual at-bats and the types of pitches he faced. And suppose that his success rate (where success is defined as getting a base hit) is as follows:

Fastball: 42%

Curveball: 27%

Screwball: 16%

Knuckleball: 22%

(These are just arbitrary numbers … I have no idea what they actually would be. We also don’t know how many of each of these pitches he saw, nor do we know what other pitches there were. I don’t know if slider falls into the curveball category; I don’t know if a split-finger fastball falls into the fastball category, etc.)

Here we have four different classes of attempts, each class having its own success rate. What are some things we could say about these attempts and success rates, if by attempt we mean class of attempt?

None of the four attempts was always successful.

None of the four attempts was usually successful (if usually means > 50%).

All of the four attempts were sometimes (or occasionally) successful.

We could even say that most of the attempts were occasionally successful, where most can include every attempt.

Okay, this feels like a bit of a dead end.

Go back to semantics. When we say that an attempt is usually or sometimes successful, that seems to indicate that an attempt may sometimes be successful and sometimes not.

THAT’s the problem.

Let’s go back to the original problem, way at the very top of this article.

Look at it like this: Most of his attempts succeeded. We’re looking at a body of attempts – each individual attempt either wins or loses – and most of those attempts end in success.

Now, look at it like this: His attempts usually succeeded. Again, each individual attempt either wins or loses. His attempts usually won.

But saying “Most of his attempts usually succeeded” either means that there are classes or groupings of attempts, each with its own success rate, or that it’s acceptable to say that if more than 25% of his attempts succeeded, we can say that most of his attempts usually succeeded.

Be careful of those absolutes!

Contrary to current social thinking, there ARE absolutes. And there are truths. And there is Ruth.


Sorry, the comment form is closed at this time.