For the past several years, I have been running several staff Survivor Pools. For those who are not familiar with the concept, it is pretty simple - pick a winning NFL team each week. The catch is that you cannot use a team more than once all season. Lose and you are out. Hence - "Survivor", or "Eliminator" pools.
I've done some prep work each year in getting ready for the coming NFL season. I have outlined in the past my strategy, and I will be revamping it once again this year with new ideas. I like to read plenty of other people's thoughts on football and strategy, but one article caught my eye and sparked an idea for this year. Dave Larkin (right here at Footballguys) wrote an interesting article about the win totals that are published each year for NFL teams, and that gave me a few thoughts. What if I took those numbers and tried to forecast the best matchups throughout the season? Could I find the easy games to pick, forecast the land mine weeks and maybe find a few diamonds in the rough? I think so, so this is my first attempt at doing just that. Stick with me as I go through this, because I think the chart at the end of the article could be worth the price of the Insider Pro membership all by itself (yes, I'm boasting, but I think it is really, really valuable).
So here was my plan - break down the entire NFL schedule and use Las Vegas to help me out. We do this all the time in fantasy, daily or season long. The guys that run the sportsbooks in Las Vegas are sharpest of the sharps, so why not use their numbers to start our analysis? I assigned every NFL team a number equal to their Over/Under line for wins for the coming season, and then I just compared each matchup for all 256 games. Now, there are a few wrinkles, mostly due to suspensions and changes since the numbers came out, but this is a reasonable start of a baseline to figuring out a plan for the Survivor season.
A few questions arose as I looked at the numbers - what is a "good" matchup? If a team expected to win 10 games faces a team expected to win only three, of course that is a good one - but what about "9 vs. 7" or "8 vs. 6"? Great questions, so the first thing I did was to just see what the numbers came out to be, and then tried to draw a few lines in the sand. After looking at all 256 matchups, I did make one adjustment for each team's number, and that was to add 1/2 to their total if they were the home team. You can make an argument as to what that number should be (and I encourage you to play with it and see if there is a better number), but 1/2 a win seemed reasonable. Below is the chart that resulted, with some highlights to point out the big disparities.
The 2016 NFL Schedule Using Team Win Differentials
Note that all numbers are absolute differences, so that would explain why Cleveland and San Francisco get a lot of highlights all year long. Both the Browns and the 49ers (along with the Titans) look to be fantastic targets for Survivor matchups this year, which should surprise no one. The question becomes when to use teams against them, and when to look elsewhere.
The chart above highlights quite a few things for me. First, only 42 of 256 games (16%) have a margin of four or higher. Those are the games we really want to target, if possible - but we cannot pick against Cleveland every week. After all, the Browns will win a few games, and we would very likely be reusing several teams if we did just that. The games at "3.5" (32 contests) and "3" (49 matchups) account for just over 31% of the schedule, so we have eliminated over half of the games on the schedule. Combining the chart above and the rules I outlined for Survivor pools shoudl result in a very solid list for the coming season.
Visually I like this chart because I can easily see which weeks are going to be tough to navigate. For example, Weeks 12 and 13 have zero green squares, so these are going to require more planning and homework. This is also the case for Week 4 and Week 8 - a week most would have guessed given that six teams are on byes. Knowing that these four weeks are going to be challenging gives us a great head start on picking key teams to reserve for those tougher slates of games.
One last point about the numbers that I chose to highlight (game differentials of 2.5 or more) and the 0.5 margin for a home team - there are more algebraic equations that can be used from Las Vegas to see if these numbers really do make sense. We have several unknowns here in August, but we do know what the sportsbooks have told us about the season-long aspect of teams via the win totals and also the opening lines for Week 1 matchups. By taking those 16 games for Week 1, I was able to compare the current spread of each game to the team total win differential combined with my 0.5 addition to the home team. Adding up the totals of the 16 favored teams and the point spread (including the oddball game of New England being an underdog despite having a higher expected win total than Arizona) and the ratio of the two numbers is about 2.5 (26 in game differential, 65.5 in point spread). Even if I exclude that Patriots-Cardinals matchup, the ratio is not much different (2.1). So why does any of that matter? I am glad you asked. If you agree with my premise that you want to target games with a team win total differential of three or more, that translates to a point spread between 6.3 and 7.5 points. By using another conversion table, again supplied by our friends in Las Vegas, an NFL game with point spread of 6 or 7 points implies a money line of -300 to -360 towards the favored team. That translates to a likelihood of winning the game around 70-75% of the time. That's exactly what we were looking for - games with high probabilities of winning for the targeted team.
I hope that you found this analysis and table useful, and I welcome all comments and feedback on the results.
Good luck this year.
Questions, suggestions and comments are always welcome to firstname.lastname@example.org.