Survivor: Surviving Week 2

Three facts need to be acknowledged before we proceed to this week's slate: (1) The main source of carnage in Week 1 was the Steelers; and (2) They were my pick (The Bucs knocked a few others out, but roughly half as many people were on them); and (3) The math in Week 1 wasn't right. I used Steelers chances to win vs. the Patriots chances to win, when I should have compared Steelers win/Pats loss for Pats win/Steelers loss. It would not have changed the pick once I put the Steelers at 80 percent, but I realized writing up Week 2 that I did it wrong. I'm pretty sure I've made this mistake before (in prior years), and it discounts the chance-to-win variable too much. I'm sorry about that. It's likely it skewed a couple of my prior picks.

I also wrote (scroll down) about this at some length in East Coast Offense, so I won't re-hash the entire thing here, but the gist was I'm sorry for the people who took my advice and lost.

That said, I'm not that sorry for them in the same sense I wouldn't be that sorry for someone who died walking a tight rope between two skycrapers or free-climbing up a treacherous mountain in bad weather. The moment you send in your entry fee, you're almost guaranteeing yourself a premature demise. Consider if you're in a 100-person pool every year, you'll be lucky to win in your lifetime. Twice would be a huge success.

It sucks if you lose in Week 1 though because you don't even get the entertainment bang for your buck, but you can play either to maximize your chance of winning the whole thing or survive for a long time while much of your pool does too.

From some of the emails I've gotten since the Steelers lost it seems many people prefer the latter and dislike my "pot odds" methodology, preferring simply to survive. I disagree, but if you want to play it that way, you absolutely can, and this article actually picks teams on that basis, too. Last week it would have been the Pats (82%) followed by the Colts (81%). I had the Steelers at 80, but picked them because fewer people were on them. Or you can use Vegas' market-based numbers in the grid.

I'm still going to make my picks based on two variables - likelihood of winning and also the payout in pot equity should you win, but you can focus only on the first factor if you like.

Finally, if you're not inclined to read the strategy article, and/or you don't really understand all the math here, I'll illustrate with a quick example why surviving at all costs can't be the best strategy: Imagine if last week in your 100-person pool, you knew 99 had the Colts. And let's further imagine you had only two choices, the Raiders or the Colts. Which would you choose? If you take the Colts, you share the same fate as everyone else - a 1 in 100 chance to win the pool no matter what happens in this game. You'd have a 100-percent chance of making it to Week 2, and a 1 in 100 chance of winning. If you took the Raiders, you'd have an 81-percent chance to be out Week 2, and a 19 percent chance of winning your entire pool then and there! So which would you prefer: (1) Stay alive for another week and have a one-percent chance to win your pool; or (2) Most likely be out by Week 2, but have a 19-percent chance to win your pool?

That's an extreme example - you could never know that 99 people were on one team, but you can see that it's not just the odds of winning, but also the pick distribution that matters.

And sadly, the Vegas odds pick was Denver (which nearly won people some pools when Indy and New England were on the ropes late). The math only favored Pittsburgh when I substituted my own odds for the market's.

Okay - on to Week 2.

This is a slate with a lot of possibilities. Let's start with the Pats who are 26 percent used, but are at 85 percent to win. The Texans are 17 percent used, but at 80 percent. When you run the numbers, a Texans win/Pats loss in a $10, 100-person pool leaves 74 left and an equity stake of $13.51. A Pats win/Texans loss leaves you with 83 poeple and $12.05 in equity. The ratio of 13.51/12.05 = 1.12. But a Pats win/Texans loss is (85%*20%) = 17% and a Texans win/Pats loss is (80%*15%) = 12. 17/12 = 1.42. Clearly the Pats are the better play according to Vegas.

What about the Packers, who are 77 percent but only six percent used. If New England won and Green Bay lost, there would be 94 people left, giving you $10.64 in equity. 13.51/10.64 = 1.28. But that payout would not justify the increased risk as a Pats win/Packers loss is 19.55 percent likely while the reverse is 11.66 percent. 19.55/11.66 = 1.68.

So the Pats are Vegas' pick. But - and I do this at my own peril - I see them in the 75-80 percent range. So given their heavy use, I'm looking elsewhere.

My picks

1. Houston Texans

I'm not in love with this team, but the Texans should be better in nearly every phase of the game. Yes, the Titans' offensive line seemed improved, but Jake Locker didn't, and he'll likely be asked to do a lot more this week. I give Houston a 79 percent chance to win this game.

2. Baltimore Ravens

I thought they looked good against Denver until the third quarter, and that coincided with the defense getting stuck on the field after John Harbaugh failed to challenge Wes Welker's clear drop. The Ravens were winning at that point, but after that third-down conversion their defense seemed to run out of gas in the high altitude. In any event, I still think this is a playoff team, and I'd expect it to bounce back in Week 2. I give Baltimore a 77 percent chance to win this game.

3. New England Patriots

The Pats couldn't stretch the field last week, and it's only going to get worse without Shane Vereen and likely Danny Amendola against a stingy Jets defense. The good thing is the Jets offense isn't likely to do all that much, and Tom Brady should find ways to move the ball. I give the Patriots an 77 percent chance to win this game.

4. Green Bay Packers

The Redskins were out of sync last week, and while Robert Griffin played better in the second half, it's unclear whether he's really the same player who was so dynamic on the ground last year. The Packers are always going to be elite on offense with Aaron Rodgers, but the secondary is still a problem. I give the Packers a 74 percent chance to win this game.

5. Philadelphia Eagles

The Chargers seem to be the same team as last year's, which means they can lose even when they're playing well. In Philadelphia against an exhaustingly fast offense, they probably won't. I give the Eagles a 74 percent chance to win this game.

6. Chicago Bears

They should be able to handle Minnesota, but Jay Cutler is mistake prone, and the Bears didn't look that sharp last week. I give Chicago a 70 percent chance to win this game.

Notable omissions:

Cincinnati Bengals - The Steelers have owned the Bengals in recent years, and I don't want to bet against Mike Tomlin after a loss.

Atlanta Falcons - I see the Falcons and Rams as rough equals, and while Atlanta has home-field advantage, there's not enough separation here.

Oakland Raiders - The Raiders looked okay in Week 1, but I'm not going to stake my life on them even against Jacksonville which should get a big lift with Chad Henne taking over for Blaine Gabbert.