Survivor: Backing the Ravens

Okay - on to this week's slate:

The Patriots at nearly 82 percent to win are this week's biggest favorite, but they're also the most popular team at nearly 32 percent. My feeling is they'll be more heavily used than that for two reasons: (1) People will want to get the win in the bag on Thursday night (and also be able to watch the standalone game); and (2) At this point in the year, the polling data gets skewed by second-chance pools, multiple-strike pools and pools where you can use teams more than once.

For that reason, it's probably better to consider not how owned the Pats are, but how owned they'd have to be before we'd fade them. So let's start by comparing their chances to win vs. teams like the Seahawks, Packers and Ravens. Let's use the Packers who are the highest at 74 percent.

A Patriots win/Packers loss is .82 * .26 = .21. A Packers win/Patriots loss is .74 * .18 = .13. The ratio of .21 to .13 = 1.62. So the payout for taking the Packers would have to be commensurate with that added risk.

If we set aside the polling numbers and just arbitrarily put Green Bay's use at 10 percent in our hypothetical 100-person, $10 buy-in pool, a Packers loss would mean 90 people were left. Let's assume another six people lose on other teams, making it 84 people left (that assumption is close to how many lose when you set aside the main team and another top choice each week). That means you'd divide the $1000 (total money in the pot) by the 84 remaining teams to get your pool equity should the Patriots win and Packers lose. Your $10 becomes $1000/84 = $11.90 in that case.

To find out how many people would have to be on the Patriots for the Packers risk to be worth the reward in this case, we'd look at the ratio: 1.62 and mulitply it by the $11.90 in equity you'd get from playing it safe. That equals $19.28.

So you'd have to have enough people on the Patriots to give you $19.28. In other words, $1000 divided by what = $19.28? That's going to come close to 52. Remember, we posited six of those who lost were from other teams though, so that's 46.

Essentially, for you to take the 10-percent used Packers over the Pats with these odds, you'd need 46 percent of the pool to be on New England.

According to our increasingly flawed polling data, the Packers are even less than 10 percent, and the Pats are only at 32, but if you think nearly half your pool will be on New England - not an unreasonable assumption - you might want to look elsewhere, whether it's the Packers, Seahawks or Ravens.

My picks

1. Baltimore Ravens

The math says the Patriots, but I think the polling data is getting noisier, and nearly 50 percent of many pools will be on New England. The Ravens are stout on the line of scrimmage, and Joe Flacco's able to find his receivers down the field. The Falcons have offensive line injuries, don't travel well and are weak on defense. The biggest worry is Atlanta's passing game - they have one of the league's elite, unguardable receivers and a top-10 quarterback getting him the ball. I give the Ravens a 76 percent chance to win this game.

2. New England Patriots

I'm torn between them and the Ravens in part because I'd love to get the game out of the way on Thursday night. Later in the year as you get down to the end, this might even present a favorable hedging situation on Sunday. In any event, the Pats are the week's biggest favorite, seem to have rediscovered their offense and get the lowly Jets at home. I'll just say there have been so many battles between these two coaches and teams, there's an Army-Navy element where anyone can win regardless of how the two have played all year. Still, I give the Patriots a 78 percent chance to win this game.

3. Seattle Seahawks

They killed me last week, and it was no fluke. Dallas beat them up in their own building, and the score shouldn't have even been that close. But after taking that loss, they should be focused against a weak Rams squad that can't even rush the passer this year. I give the Seahawks a 74 percent chance to win this game.

4. Green Bay Packers

Both teams should be effective moving the ball, and I'd hate to rely on the Green Bay defense for a final stop in a one-score game. I give the Packers a 70 percent chance to win this game.

Notable Omissions

Cleveland Browns - Fading the Jaguars every week has worked like a charm, though last week in Tennessee was nearly a disaster, and now the game is in Jacksonville, and Cleveland is missing its starting center.

Dallas Cowboys - After they destroyed the Seahawks in Seattle, it's hard to see the Giants team that failed to show up in Philly as a serious threat, especially without Victor Cruz and Dominique Rodgers-Cromartie. But these teams know each other well, the Giants had been playing well in the weeks prior and the Cowboys defense might be playing over its head.

Washington Redskins - The Titans are bad, to be sure, but Kirk Cousins has been a generous source of interceptions, and the Washington secondary is prone to giving up big plays.