Survivor: Backing the Texans

Surviving Week 1

What an interesting year 2010 was. I lost in Week 2 with Dallas, even though I thought Green Bay was far more likely to win that week. Why did I do it? You can check out my reasoning in detail here (both in the article and the comments). The shorter version is there are but two variables to consider if you want to *win* Survivor: (1) How likely a team is to win that week; and (2) What payout in terms of expected pool equity does a win with that team provide.

When we start the year, our pool equity (assuming the commissioner doesn't take a vig for himself) is equal to our entry fee. We pay our $100 in a 100-person pool, and we own a 1/100th share of $10,000. As the season goes on, and people die off, our equity grows. When there are 50 people left, for example, our share is 1/50th of $10,000 which is $200.

But people do not die off in our pools linearly or uniformly. Some weeks hardly anyone perishes, and others take down substantial portions of the pool. As such our equity can go from $100 to $120 dollars from Week 1 to Week 2, but it could go from $200 to $1000 dollars from Week 6 to Week 7, if 40 of the 50 remaining survivors lose.

Unfortunately, great rewards also come with risk. To get that five-fold jump in equity, you had to avoid the team that took down 80 percent of your pool. And typically, a team like that is going to be the biggest favorite remaining on the board. Hence the only way to surf that wave, so to speak, is to take a team that's a lesser favorite, and therefore more likely to lose that week. Otherwise, if you only take the team most likely to win (and other people are doing the same thing), then you will have the greatest chance of surviving a long time, but your equity will grow at a glacial pace, and you will have to win more weeks and go further to win the entire pool.

In years where pools go 17 weeks, typically the entrant who takes the biggest favorite every week will win it (or split it with several other people). But in years where the pool ends in Week 9 (and this happens often), the person who went against the grain will be left standing when massive upsets knock out the majority. At the season's onset, we really have no idea what kind of year it'll be - a favorite-heavy one like 2009, or a short-season where carnage is everywhere. All we can do is recognize that uncertainty and give ourselves the best chance to win overall. The way to do that is to consider likelihood of winning each week, and amount of expected equity growth should we advance to the next round.

As such, I am going to rank the teams EXACTLY as I did last season - in terms of win likelihood, multiplied by payout. That it did not fare well last year in no way shakes my faith in its efficacy as the best method, assuming *winning* is your ultimate goal. If your goal is to get entertainment from hanging around longer, regardless of how much your pool equity grows each week, then my method is NOT the best. It is only the best at winning the entire pool. My method presumes there is no difference between losing in Week 2 or Week 10, so long as you get $0 for both outcomes.

For those of you who simply want to last as long as possible regardless of the expected equity, you can still benefit from this article as every week, I list the team most likely to win, irrespective of the payout. Just pick that team and ignore the rest. But just know that I will not be doing that.

Okay, that out of the way - on to this week's picks:

My Picks

1. Houston Texans

Maybe Kerry Collins is competent enough to keep up with the Texans, especially if they're without Arian Foster, but it's tough to see the 38-year old journeyman coming out of retirement a few weeks ago and leading this team to a win on the road. Moreover, Jim Caldwell strikes me as a below average coach who benefited from Tom Moore's system and Peyton Manning's all-time skills. Without that in place, you have a team with below average talent going on the road to face a rival whom they've tormented for years and who wants nothing more than some long overdue payback. Vegas has this as an 78 percent game, but I've got it at closer to 85. As the pick distribution is so evenly spread out, the "pot odds" variable isn't significant enough to sway me to another option.

2. New England Patriots

I hate to take a road team, especially in a division rivalry where the opponent (the Dolphins) is not a total doormat. That said, the Pats are one of the most projectable elite teams every year given their quarterback and coach, and after a seven-month layoff, that projectability has great value. I give the Patriots a 76 percent chance to win this game.

3. San Diego Chargers

They're the biggest Vegas favorite on the board, and they're at home, but I have a bad vibe about this game for a few reasons: (1) The Chargers are one of those "sabermetric darling" teams where all their key indicators say they're Super Bowl contenders, but they always seem to underperform; (2) the Chargers under Norv Turner tend to start out slowly for whatever reason; and (3) the Vikings are a little dangerous with a veteran quarterback who's played well in spurts over the last two years, a dominating running back, and one of the best pass rushers in the league. Brad Childress' departure can only help, too. I give the Chargers a 75 percent chance to win this game.

Notable Omissions

Arizona Cardinals - I'd rather not gamble on Kevin Kolb's first game with his new team, incidentally one that was pretty terrible last year.

Cleveland Browns - I need to see the likely improvement before I'm sold on them as my survivor choice.

Kansas City Chiefs - The Bills aren't good, but they're scrappy and can get the ball down the field. If the Chiefs don't put them away early, it wouldn't surprise me to see Ryan Fitzpatrick mount a comeback, especially if Todd Haley tries to use Thomas Jones to kill the clock.

Philadelphia Eagles - Super Bowls aren't won on paper, and the Rams are at home and should be tougher on both sides of the ball this year.