Arkansas a good bet against Alabama 

September, 22, 2010

I've been doing these systems of the week since the start of college football season and, for the most part, they've been pretty successful. Last week, Sal from nailed the Arizona-Iowa game. He was also all over the Stanford-UCLA game the week before.

But these systems do raise people's ire. I got plenty of e-mails last week, from readers and handicappers, telling me they thought Sal was way off and needed to have his head, as well as his system, checked for banking on the Cats. They just weren't buying. Part of the reason is that the stat-based systems a lot of wiseguys use reach so far back, to the early 1980s, readers think half of the positive situations produced by the data are irrelevant.

So when I called Big Al McMordie about this week's System of the Week I asked him -- given that the numbers he's basing his decision on stretch back 30 years -- how can he be so sure they will apply to today? "I've been doing this a long time," he told me. "Patterns develop. If I see a situation that has favorable results close to 100 times, even if that means it's just three or so games a year, I can feel confident the pattern will hold."

That's how we got to the two systems he offered up for this weekend.