Number of games played, by itself, is a terrible practice metric for chess. Study time is vital, and the type of game is also extremely important. How the player plays the game is important. Do they just play the same opening every time? If so, they aren't going to learn as much as someone who never plays the same thing twice. Do they review the game? The amount of practice value squeezed out of a single game of chess is vastly different between me and a professional, and in turn vastly different from someone who doesn't care at all about chess and me.
You can play 10000 games of chess and not learn a thing if you aren't setting out to learn.
Exactly. I think Gladwell -- citing that study -- uses the term "deliberate practice," where it's not the same thing over and over, but going beyond what you did the day before or at least doing part of it differently, in order to stretch and acquire more skill.
Gladwell cites a study as only one component of "Outliers." The study itself is the subject of an entire book called "Talent is Overrated." This book should actually be read before Gladwell (which is the order I did). An excerpt from the book is here:
I think his point is more that the common myth of the natural born "genius" is not what society commonly thinks it is.
You can be born mentally genius at something, or innately talented at some physical skill or other, but that the "genius" or "born talented" people we see in society really owe a lot more of their success to practice rather than any innate quality they were born with.
You can play 10000 games of chess and not learn a thing if you aren't setting out to learn.