18
Soooooo... When I was studying at university we were forced to read a course in academic writing.
(what I was studying was purely technical and hands-on, the only academic writing I did was in said course)
But one thing that was brought up during the course was the importance of knowing what, how, and why when it comes to measuring things.
Example:
How much buying power does the average person have?
A group of 9 people has an average income of $10.
Now we add a billionaire to the group, someone that makes $10'000'000'000.
The average income is now $1'000'000'009.
From the outside this group might look super rich, but in reality it's just that one billionaire scewing the numbers... though not all numbers.
The median income is still $10, because that's what most people make.
In this example the median is more interesting than the average, even though the question literally asks for "average".
This logic and fallacy is applicable to pretty much everything and anything you might think of.
As a code-writer I am baffled when people of "importance" use any of the following 2 metrics as something that measure anything other than itself:
- lines of code (added or removed, doesn't matter)
- amount of pull requests (I might submit that 1 pull request that fixes 10 years of random bugs nobody has been able to figure out)
Things to actually measure:
- how many bugs are existing, before and after? (testing)
- was the thing delivered on time? (deadlines)
- does the thing built do what it was supposed to do? (expectations)