The author, Ray Gustini, went 5-11, a paltry 31.25% success rate, with his picks last week, illustrating the point I made in another thread, which is that most pundits make terrible picks. By contrast, my overly simplistic algorithm (pick team with better record; in case of a tie, pick home team) went a much more respectable 13-3 (81.25%). I would be willing to bet $1000 right here and now that my algorithm will kick his scattershot method by at least 10% over the course of the season, and probably more closer to 30%.
The problem with sportswriters' picks is that most writers are incorrigible bandwagon jumpers (they can't help themselves), meaning they're short-sighted and get distracted by short-term variations and anomalies (i.e., they believe in "hot" and "cold" streaks); and even worse, they want to make a splash by being the one to have prognosticated that improbable pick that stuns everyone. As successful -- that is to say, experienced -- bettors and investors know, however, "The Trend is Your Friend." They don't become overly excited by short-term aberrations because they know that in the end, everything regresses toward the mean. My algorithm, while certainly not sophisticated and unable to take into consideration factors unique to specific games or predict the unpredictable, at least strips all emotion and sensationalism from the picking process . . . and is therefore, considering how simplistic it is, remarkably successful.