Scoble — chubby, bespectacled, and awkward, the unlikeliest of all video personalities — main job for the magazine was to produce a seemingly infinite series of video profiles of startups. Scoble's unwatchable videos mostly consisted of him lapping up tech blather from CEOs of doomed startup ventures about how they would be reinventing some paradigm or another. But as bad as they were, he fell from old-media grace for two main reasons.

First, he picked the wrong backers. The Fast Company Web guy who hired him, Ed Sussman, was loathed by his counterparts at the print magazine and got fired last year. And the videos were sponsored by Seagate, the hard-drive maker. After the company fired CEO Bill Watkins, with whom Scoble had a mutual lovefest, it was only a matter of time before the gravy train ended.

Second, there was Scoble's dangerous overuse of the Web startups he covered. FriendFeed and Twitter provided a steady IV drip of attention, so vital for soothing the damaged ego of a geek who never got over his awkward youth. But Scoble's paid work suffered while he volunteered to provide obsessive entertainment for his fellow Internet addicts.

He and Fast Company are saving face by continuing his column (heavily rewritten or wholly composed, no doubt, by an editor there). And he is, naturally, promising that he's meeting with a lot of companies to plan some exciting new startup. This is what one says in Silicon Valley when one is facing unemployment — the equivalent of talking up one's burgeoning freelance career in New York, or waxing enthusiastic about a script in Hollywood.

What really happened here: Scoble got invited by the pretty girl to the old-media prom. And he just got dumped. How will his ego ever recover?

The theory it's based on is nonsense. It is true that ideas spread virally through the population. But it turns out that there's not a single set of influencers who are reliable Typhoid Marys. Duncan Watts, a former Columbia University researcher who now works for Yahoo, found in a study that the emergence of contagious ideas is random. Repeated experiments found that anyone can start a trend, and it's impossible to predict who those people will be.

Watts's research is not 100 percent conclusive; his models might not translate perfectly to the real world. So let's go there! In April, a study by Canadian research firm Pollara found that word of mouth works — nearly 80 percent said they'd buy products recommended by a friend or family member. But word of mouse? Only 23 percent said they'd buy something touted by a blogger. "This shows that popularity doesn't always equate to credibility," Pollara executive Robert Hutton told MediaPost. "Marketers might have to reconsider who the real influencers are out there."

In backing Scoble, Seagate hoped to buy cheap buzz. It's a convenient fantasy for marketers: Find the one magic guy to woo, then watch him chat up your company to Wall Street traders! Seagate would have been better off sending big hard drives to a dozen bloggers. Or a hundred. Or, for that matter, a random assortment of people, whether or not they have a habit of typing out the contents of their brain every 3 seconds. Anyone — literally — would have been a better choice than Scoble.

When it comes to discussing the sex lives of one's daughters on the record, however, we draw the line. For the crime of oversharing, Watkins has gone from hero to zero — and become our latest Silicon Valley Tool.

Do you have other nominations? Please send them in.