tae808 wrote:
Hi Mike,
This question came to my mind after working on one of
Magoosh problem.
In the "dead-ball" era of 1900-1919, Major League Baseball hitters in both leagues hit an average total of 370 home runs each season, more than 60% percent less than those in the 1920s.The explanation states: Here, though, we are talking about comparing the "average total" --- the average of several seasons is not guaranteed to be a whole number: it could be a decimal. Therefore, we aren't really counting whole things anymore, so we don't used "fewer"
But you said if I I can count what's being percentized, then I need to use "fewer" which is contrary to the explanation. (I can count how many home runs they hit...)
Please let me know. Thanks
Dear tae808,
I'm happy to respond.

We are starting to get into territory that's at the outer edges of what the GMAT might expect you to know. Admittedly, sometimes I write a question that pushes issues in this region, as I did in that baseball question.
Here's the thing. What does the word "
countable" really mean? Let's think carefully about this. The colloquial definition is that "we can count it," but think about that. If we are guaranteed that we can count something, then we are guaranteed that it will come in positive integer quantities. If I buy eggs, the number of eggs will always be a positive integer; if the carton contains part of an egg, that would be disgusting, and I certainly wouldn't count that as part of my purchase. If I work with a group of people, the number of my fellow employees is always a positive integer. This is the deep idea. Having 2/7 of a person or of an egg simply wouldn't make sense.
This is yet another reason that so many units are not countable. If I take a long walk, I could by chance walk exact four miles, but most real world distance fall between the integers. Most real times are not precisely a whole number of hours, etc. It makes perfect sense to talk about a distance of half a mile, or times of a fraction of a second, etc. Thus, even though we are talking about units that, ostensibly, we could count, we are not guaranteed that they are positive integers, and so "
countable" doesn't apply to them.
Similarly with averages. Let's consider a school example. Each class has a certain number of students, and students are countable: we are guaranteed beyond a shadow of a doubt that the classrooms have no decimal parts of students actively participating in the classroom. Thus it is perfectly correct to say:
This class has fewer student than that class.
Now, suppose we start taking averages. Suppose there are several four-grade classes and several fifth grade classes. If we start talking about the average fourth-grade class or the average fifth-grade class, we are no longer guaranteed these numbers are integers. Those two averages could be, say, 32.7 and 29.7, so it would be correct to say:
The average number of students in the fifth-grade class is less than that average in the fourth-grade classes. As with physical units, it may work out by chance that an average comes up neatly to an integer, but we are not guaranteed at the outset that the result will be a positive integer, and this means that it doesn't qualify as countable.
Similarly, if we are talking about percents of averages --- well, if averages are not countable, then percents of averages are not countable, even though percents of individual students would be countable.
To qualify as countable, something must come with the binding guarantee that it will always only come in positive integer quantities, that if it comes in any decimal or fraction part it will cease to be what it is (as a fraction of an egg no longer counts as an egg!)
Does all this make sense?
Mike

Thank you so much. I think I understand 95%... But one thing...
What you wrote: "Fewer than 20% of the people are still in the room." is wrong...?
Since there can be 21.5% less people in the room...? Or is it correct since 20% is a whole number...?