Week 1 NFL Survivor Strategy: Two Good Options

September 7, 2011 - by David Hess

Welcome to the first installment of our weekly series of posts exploring data-driven strategies for getting an edge in NFL Survivor contests. This post includes analysis for NFL Week 1, based on NFL predictions from our algorithmic Team Rankings models and other relevant data.

This analysis is also meant to complement our Survivor Predictor tool, which provides a grid view of win odds predictions by week, from now until the end of the season. This presentation of prediction data will help you quickly grasp the current week’s best Survivor pick options, while also analyzing future schedule considerations when you are torn between two or three picks.

Each week, we’re going to present a table that lays out the most important info that we use to make our choices. This table contains data on three main factors, for each of the week’s favorites:

– Likelihood of winning. We obviously want to survive, which entails choose a winning team.
– Popularity with the public. We want to avoid choosing the consensus favorites, if prudent. The only way to win is for your opponents to lose, while you survive. That means we’re typically going to be rooting for the top favorites each week to lose, and take out as many of our opponents as possible.
– Future value. Because you can only pick a team once, if all else is equal, we’re going to want to save teams that have lots of easy games in the future, and choose teams where it may not get any better than the current week.

OK, now that you know the plan, let’s get to the table and analysis:

Team	Opponent	Spread	TR Odds	Average Pick%	Better Future Games
Houston	vs Indianapolis	-9.0	79%	4.4%	0
San Diego	vs Minnesota	-9.0	77%	7.3%	1
Arizona	vs Carolina	-7.0	71%	14.0%	0
New England	@ Miami	-7.0	74%	15.5%	11
Cleveland	vs Cincinnati	-6.5	68%	12.3%	1
Kansas City	vs Buffalo	-6.0	69%	22.3%	0
San Francisco	vs Seattle	-5.5	73%	1.3%	0
NY Jets	vs Dallas	-4.5	63%	0.9%	11
Philadelphia	@ St. Louis	-4.5	68%	10.1%	3
Green Bay	vs New Orleans	-4.0	59%	1.5%	15
Denver	vs Oakland	-3.0	64%	1.5%	0
NY Giants	@ Washington	-3.0	57%	2.9%	8
Atlanta	@ Chicago	-2.5	51%	1.1%	12
Baltimore	vs Pittsburgh	-1.0	50%	0.1%	13
Jacksonville	vs Tennessee	-1.0	53%	0.4%	1
Tampa Bay	vs Detroit	-1.0	52%	1.0%	10

Average Pick% = the average percent of the public that is selecting each team, based on the Yahoo! and ESPN public picking data.
Better Future Games = the number of future games where the team is expected to have higher win odds than the current week, based on our NFL Survivor Tool.

Teams We Already Picked

None (hooray for Week 1!)

Best Options

This week is an interesting one. In case you haven’t heard, Peyton Manning is officially out for Indianapolis’s Week 1 game against the Texans. As a result, the Colts have moved from slight underdogs (+1.5) to large ones (+9.0). That massive line change has made the Houston Texans the biggest Vegas favorites (tied with the San Diego Chargers).

The Texans also have the highest TR Win Odds, have no better future opportunities for use, and are being picked by only 4.4% of the public (only the 7th-most popular favorite). This game hits all three check marks: high likelihood of winning, low public popularity, and low future value.

That said, this is about as risky of a 9-point favorite as we can think of. The 7.5-point change in the line is almost entirely based on Peyton Manning being out, but there’s not a whole lot of hard evidence about how good the Colts are without Manning. If the Vegas line is an accurate gauge of the Manning-less Colts, then Houston is a great pick. If you don’t trust the line, and would prefer to go more conservative this week, there is another pretty decent choice.

The San Diego Chargers are tied with the Texans as the biggest favorites of the week, according to both the Vegas spread and our win odds predictions., and they have little future value (one great opportunity — Week 12 at home against Denver). They are being picked by a slightly larger portion of the public (7.3% vs. 4.4%), but the differences isn’t huge. A 7.3% public pick rate is still pretty low, and you’ll still have some big upsets to root for. We slightly prefer Houston given the current lines, but if the Texans line moves back towards the Colts, the Chargers are a good replacement pick.

The Arizona Cardinals and Cleveland Browns are also possible choices this week, though less attractive than the Chargers or Texans. With spreads of -7.0 and -6.5, respectively, and TR Odds of 71% and 68%, they are relatively safe choices, and have little future value. The main negatives for these teams are that they have lower win odds but higher public pick percentages than both San Diego and Houston.

If you’re in a massive pool, and you’re the gambling type, we could see taking a flyer on the San Francisco 49ers. The Niners’ spread is a bit low for our liking, but our win odds project them to have a slightly better chance than the spread indicates. Plus, a minuscule 1.3% of the public is picking them, and they appear to have no better opportunities in the future.

Stay Away

We’re not sure why the New England Patriots are the second most popular pick this week. Despite the high win odds, we’d advise you stay away, since they’re so popular and have so much future value (for example, waiting to use them at home against Miami rather than on the road).

The spread and win odds for the Kansas City Chiefs do not justify them being the most popular pick. Yes, they don’t have a ton of future value, but we at least think that Week 10 versus Denver will be a possible opportunity. Plus, our models think the Vegas line is shaded a bit more towards the Chiefs than it ought to be. It’s fairly common for the public to overrate last year’s playoff teams at the beginning of a new season; that might be the case here.

Official Pick: Houston Texans over Indianapolis Colts

While there are several possible picks, the data definitely indicates we should choose Houston. They have the biggest spread, the highest TR win odds, and little future value. The main source of risk here is that the large line may not be an accurate assessment of Houston’s true win odds, due to the lack of info about the Peyton Manning-less Colts.

However, this situation isn’t too different from other Week 1 lines — many teams undergo important offseason personnel changes. We looked back through our database to see if large Week 1 spreads (7 points or more) were less accurate than large lines later in the season. What we found was that large Week 1 favorites went 14-5 (74%), while large favorites from other weeks went 492-123 (80%). That difference is negligible — if just one more large Week 1 favorite had won their game, the two win rates would be even.

Given the above info, we trust the line enough to make Houston over Indianapolis our official pick. If you’re still not convinced, we’d recommend San Diego over Minnesota as an alternative.

If you have any questions, or need advice about your specific situation, feel free to chime in below. At this point, you won’t have too good of an idea about who your opponents might be picking, so it’ll be tough to do too much in the way of game theory. But if your pool has weird rules, lay them out for us, and we’ll tell you what we’d do in your situation.