In 2011, I mentioned the idea that the NFL is that it is a league is entirely predictable, despite the incessant claims that “you just can’t figure this league out”. In fact, the contrary is true. Yep, by simply using history as our guide we see that the NFL is entirely, almost eerily, predictable. And with that I unveil Colin’s official “NFL Season Prediction Rules” to help guide you in making season predictions as well as help you hammer out some profitable season win totals.
The first step is looking at history in the following areas – overall home wins, playoff turnover from year-to-year and number of teams that experience either the euphoric feeling of 4+ additional wins or the day after a bender feeling of 4+ losses. The table to the right lists regular season home team wins, new playoff teams and teams with +/- 4 wins dating back to the 2003 season. I use the 2003 season because that is one year after the NFL re-aligned the divisions to the current format. This is also when the NFL was fully transformed from a league where defense and running the ball were staples of success to “you have to be able to throw the football to win”. And I throw out the first season of the NFL re-alignment due expansion and new division match-ups for numerous teams.
Based on the data above we can be fairly certain of a few things for any NFL season – 1. The home teams will win somewhere between 140-152 games; and will more often will fall in the 143-147 range; 2. We will have at least 6 new playoff teams each year; and 3. There will be at least 10 teams that have a difference of wins from 2010 of four or more. I can also add one more nugget – at least one team will go from last to first in their division. That has happened at least once every year since 2002.
The data tells is that the NFL is certainly predictable, however, therein lies the problem, while being predictable at this summary level, the NFL is capriciously un-predictable at the detail level. Yup, that’s the problem we know 6 new teams will make the playoffs from year to year, however, I cannot make a wager that there will be “six new playoff teams”. We actually have to figure out the specific teams and make the wagers accordingly! Fortunately for all of us, I am giving out the exact steps to predicting where each team will fall.
Steps for NFL season predictions:
- Determine the teams that have the potential to fall into +/- 4 categories. For starters, we can obviously throw out any record that would cause a team to go over 16 wins or under 0 wins; Next the likelihood of a 0/1/15/16 win team is about 1 in every three seasons, so unless that is due, we can throw those records out as well. Finally, a 14 win team happens about once a year, as does a 2 win team. Therefore, I have to take into consideration all previous season’s 10 and 6 win teams make either the leap to 14 or fall off the map to 2. Even with that concession, the list is dramatically reduced.
- Group the teams into categories where they most likely will fall, either +4 or -4 wins. This is much harder than it sounds but using Bill Barnwell’s2012 well written piece on “new school stats“, we can begin to see where teams might go south and what teams are looking up. I also incorporate the new schedule, how fragile is a team, both mentally and physically and pre-season hype. Of course, this is also where personal opinion plays a huge part as well.
- Download the complete NFL schedule into Excel or a database system. Start with the teams that you feel will most likely end up with +/- 4 wins. Run through their schedule, giving a win or a loss to each game. Remember at this step that it is very easy to take wins for granted, yet each year we see places where teams trip up. Keep in mind “look ahead” situations, non-conference games that overall mean less in the standings, games before a bye, third road game in a row, west traveling east and vice-versa and occasionally you just have to assign a loss because a team is due for one.
- Apply wins and losses to the remaining teams, remembering that those teams have to fit within the parameters of already set teams from the above steps.
- Final sweep to make sure I have followed the rules and fix teams that are seem slightly “off”. We call this step “Colin’s personal touch”. Yup, this is where the whole thing goes up in smoke!
- Once you’ve made it through the entire schedule and you have your full complement of +/- 4 teams, check the standings to make sure you have six new playoff teams and at least one “worst to first” team. If not, it’s back to the drawing board with determining wins and losses for the +/- 4 teams.
- Updated 2014: Normalize home wins to somewhere between 148 and 154 wins; there is a little bit of flexibility here, but not much. A tip here, is if you have division teams splitting, split them so both road teams win. You’d be surprised how often that happens.
- Update for 2016: Home wins in 2015 dipped to the lowest total since 2006 at 138. In 2007, the total bounced back to 147 and I expect 2016 the total home wins to be back in the target range of 143-148 (9 out of last 13 seasons). I am going to make a slight adjustment to the number of new playoff teams. From 2003-2010, the average new playoff teams averaged 7; conversely, from 2012-2015, the average is 4.5; with 2011 being 6. I believe the dip is due to a couple of things – 1. the practice restrictions in place recently give the more well run franchises a significant advantage (Seahawks, Pats, Packers, Steelers); and 2. the current batch of hall of fame quarterbacks (Brady, Roethlisberger, Rodgers).