He shared a story of how he basically did a small con to build credibility.
He surveyed 10 of his friends, 2 were using Linux.
A year later 4 were using Linux.
So...
Now, this is a non-representative + insufficient sample size of “4 friends” up from 2 friends...
But the sentence runs in the article...
“According to Linux Home Journal, Linux adoption is growing 100% + doubling every year”.
BusinessWeek uses that growth rate and quotes it.
Then?...
“According to BusinessWeek, Linux use is rapidly doubling every year!”
So from 2 friends to 4 to a random article to BusinessWeek he CONJURED...CREDIBILITY.
We believe stuff other people believe that other people have LED them to believe...
Sometimes it’s innocent mistakes (like Popeye + Spinach TL;DR it was decimal error of iron content)
...but sometimes it’s just willfully misleading.