Puzzles, Mysteries, and Lies

I've been reading Malcolm Gladwell's opinions of the Enron collapse on the New Yorker's site. Mr. Gladwell all but acquits Enron senior management saying that, of course, they obfuscated the truth through questionable accounting standards because that's what they needed to do to preserve the value of the company. He sites Gregory Treverton's distinction between puzzles and mysteries to make his point.

Mr. Treverton has defined puzzles as being problems with incomplete data sets and mysteries as complete but confusing data sets. In other words: to solve a puzzle, you must infer a missing but crucial piece of data; to solve a mystery, you must simply connect the dots. Mr. Gladwell asserts that, while the boys from Houston were convicted for creating a puzzle, they were actually guilty of creating a mystery.

The part that sticks in my craw (and yes, despite overuse, I still have a craw), is that Gladwell makes the distinction between puzzle and mystery as if all the facts of the case could be tracked in a simple checklist - that if you had a clipboard with all the facts necessary to solve the problem listed, you could go right down the list and place a tic mark next to them all. The point he seems to miss is that context is, in and of itself, a data point. Enron management not only failed to provide context for some data points, but out-and-out lied about the context in some cases.

As his kicker, Mr. Gladwell sites the case of Cornell University students reviewing Enron's financials and posting a recommendation of "sell" on the University's website as proof that the dots could be connected. The truth is, just because someone was able to solve a puzzle doesn't make it a mystery.