Using Predictive Stats to Anticipate the 2016 Season

Dec 27, 2015; Detroit, MI, USA; Detroit Lions head coach Jim Caldwell during the first quarter against the San Francisco 49ers at Ford Field. Mandatory Credit: Tim Fuller-USA TODAY Sports
Dec 27, 2015; Detroit, MI, USA; Detroit Lions head coach Jim Caldwell during the first quarter against the San Francisco 49ers at Ford Field. Mandatory Credit: Tim Fuller-USA TODAY Sports /

There are so many stats in today’s sports leagues that sometimes it is hard to keep up with the ones that actually matter.  In 2012, Bill Barnwell (then of Grantland) put together a list of the four best predictive and widely accepted stats that have come into the recent realm of knowledge.  Three of those stats are predictable (in a sense), and can be applied to each team the league.  The Detroit Lions are no exception.  To visualize the upcoming season, take a look back on the last two seasons within these parameters, as a general comparison.

Record in One-Score games: (10-5)

The Lions were 6-1 in one score games in 2014, and 4-4 in 2015.  Typically, teams like the 2014 Lions do exactly what the 2015 team did.  Winning close games is simply not a something that happens to even great teams year after year.  Games decided by seven or fewer points are largely clinched by one or two plays, often of which are fluky or laden with human error (As we all know too well).  In fact, most teams in the league hover around .500 in games decided by less than one touchdown.  Exceptions occur, like always, such as the New England Patriots in their back-to-back Super Bowl years.  The Patriots went 10-2 in one score games in 2003 and 2004.  However, teams usually revert back to .500 in any given season.  For instance, in the 2011 season, groups of teams that won more than 75% of their one score games combined to finish at 50% the next year.

What does this record mean to the Lions, though?  Well, in the two years before Jim Caldwell, the Lions had a record of 6-14 in games decided by seven or fewer points.  It would seem as if the “Football Gods” were averaging these few years out.  In terms of the 2016 season, this would mean that we are likely to see another average or perhaps below average season from the Detroit Lions in games decided by a touchdown or less.  Considering the past two years were at or above the .500 mark, unless the Lions are one of the few outliers in every season, the team is likely to repeat last season’s output.

Football’s Pythagorean Theorem/Point Differential: (-3 points)

Football in general is tough to predict.  In fact, football is tough even to draw conclusions from,

The NFL cannot draw definitive conclusions from their 16 games, which is easily the smallest amount in major professional leagues

given the microscopic sample size compared to other leagues.  The NBA and NHL play 82 games, the MLB plays 182, and most soccer leagues around the world play each team in their league twice.  The NFL cannot draw definitive conclusions from their 16 games, which is easily the smallest amount in major professional leagues.  However, a stat that holds most of the predictive power over the league is that of Point Differential.  Teams that have winning records usually have a high point differential, in that, they score more points than they give up.  In contrast, teams with poor records have poor point differentials.  However, there are always teams that over-perform and those that under-perform.

In 2014, the Lions were one of the teams who over-performed their expected wins by way of point differential.  To see how this Pythagorean method works, check out Pro Football Reference.  Essentially, plug-in the amount of points scored and the amount of points given up into this equation, and the result spurts out the amount of games any given team should have won.  In 2014, the Lions scored 321 points and allowed 282.  The Pythagorean Theorem tells us that 2014 team, who won 11 games, should have won 9.2 games.

Usually when a team over-performs their Pythagorean method by nearly two games, the same team declines in wins by an average of 1.8 the next.  In 2015, the Lions scored 358 points and allowed 400.  The same equation shows the Lions as a 6.95 win team, nearly their exact amount of wins that they ended the season with.  The difference in wins from 2014-2015 was four, which overshot the estimated average decrease in wins by two.

What does this mean for 2016?  Well, the 2015 team won exactly as many games as what their points total said they did.  When teams do such, they average no changes in their win percentage the next year.  Of course, with player, coaches, and in the Lion’s case, General Manager changes; teams can drastically over or under-perform those averages.  However, when making a prediction for next season, it would be wise to remember this Theorem.

Turnover Margin: +1

Mandatory Credit: Jeff Hanisch-USA TODAY Sports
Mandatory Credit: Jeff Hanisch-USA TODAY Sports /

The predictive stat that most fans know and trust the most is turnover margin.  This one is as simple as can be; teams with larger positive margins perform better in the win column than team with large negative margins.  A team with a -10 turnover margin (like the Jacksonville Jaguars in 2015) routinely find themselves with a poor record, while a team with a +10 turnover margin (like the Cincinnati Bengals in 2015) commonly find themselves in the playoffs with fantastic records.  However, as a predictive stat for win-loss records, this particular one does not have many uses.

For one, turnovers are largely random occurrences.  Fumble recoveries are literally a 50/50 chance for either team and interceptions vary every season depending on a multitude of factors.  Of course, quarterbacks that have low interception ratios usually keep them low, but random developments such as dropped or tipped passes can greatly affect such margins.  In the same essence, running backs that fumble in bulk usually lose more fumbles than normal, but because the football is oblong and misshapen, players cannot recover fumbles with regularity.

Another example, for instance, in 2014, the Oakland Raiders had a turnover ratio of -15, and the New Orleans Saints had a margin of -13, which were last and 2nd to last respectively.  In 2015 though, both teams improved to +1 and +2 in their ratios, equal to 16th and 17th in the league.  The Raiders went from 3-13 to 7-9, while the Saints stayed at 7-9 both years.  However, in the Saints’ case, they also decreased their point differential by 45 points from 2014-2015, which would make sense of their turnover margin increase, yet identical record.  In general, though, teams that improve their turnover margins generally see an increase in wins as well.

The 2014 Lions had a +7 turnover margin, and the 2015 Lions had a -6 turnover margin.  Of course, the 2014 team finished with 11 wins, while the 2015 team finished with seven.  Drastic differences in wins such like this are not highly based on stats like turnover margin.  Good teams typically stay in the top half of the league in turnover margin, and bad teams habitually stay in the bottom half.  The Lions have been just about smack dab in the middle of the league in the past few years, which lends little credence to predicting a turnover margin.

Keep in mind, however, the recovery rate of fumbles.  Considering fumble recoveries are 100% random, teams will generally end up recovering around 50% of the fumbles they force.  In 2015, though, the Lions were 31st in the NLF in recovery rate, at 40%.  If the team is just a bit luckier in 2016, they could easily best the -6 margin from last season.


By all values, be it Pythagorean Theorem, turnover margin, close games record, or even just the eye test, the Detroit Lions were far worse in 2015 than in 2014.  This does not come as a surprise to anyone who follows the team.  What these stats say about 2016 is a different matter altogether.  Their Pythagorean equation predicts another mediocre season, likely between 7-9 and 8-8.  Their record in close games predicts a few more heart-breakers than in recent seasons.  Their turnover margin predicts an average team in the same vein.

As it happens in nearly every offseason, fans of teams get excited about their players, new systems, coaches, and overreact to how good or bad they could be.  Looking at stats like the ones above can give fans more of a base to predict the season, and the stats are saying the 2016 season for the Detroit Lions could very well be a very “blah” 8-8.  I do not personally like to predict a season when it is still more than two months away, but I do like to set reasonable expectations as to not get caught in a 2012 Detroit Lions situation.  Every statistic listed above predicted a worse season, by 2-3 games for the 2012 season, after a return to the playoffs the year before.  The newly minted playoff team drastically under-performed even by the decrease in expected wins the next season, and reverted back to norm in 2013.  It is helpful to manage expectations for the upcoming season, whether or not to take them into consideration is up to you the fan.

Next: Detroit Lions' All-Fantasy Football Team