It may seem silly to talk about the Steelers’ 2012 season two months after the team played its final game. However, the question of how Pittsburgh played last year relative to its expected performance is still relevant because it has a bearing on how the team is likely to fare in 2013.
Every year in the NFL, there are teams that either San Diego Chargers" href="http://www.pro-football-reference.com/teams/sdg/2012.htm">under or Baltimore Ravens" href="http://www.pro-football-reference.com/teams/rav/2012.htm">overachieve when compared to the predicted win totals generated by various statistical models. The reasons why almost always boil down to luck. Some teams get easy schedules and rack up wins against weaker opponents, others don’t. Though the odds of recovering a fumble are essentially 50-50, some teams recover far more than their share in a given season. Others aren’t so lucky.
When looking for teams that are likely to make a big jump or big fall the next year, it is important to identify those who either did better or worse than they should have. Teams that are either particularly lucky or unlucky in a given season are not guaranteed to regress or improve the following year. But they are the most likely candidates to post a drastically different record.
For example, Green Bay, Baltimore and San Francisco all won two or more games more than what Football Outsiders’ analysis predicted in 2011. Though all three made the playoffs this year, each won about two games less than in the previous season.
A .500 team last season, the Steelers were certainly a disappointment to fans who expected them to reach the playoffs for the third straight year. But how did their 2012 record compare to what it should have been? What does that mean for 2013?
As alluded to above, various analysts who study pro football have come up with statistical models that can be used to predict every team’s expected wins and losses in a given year. The most basic of these simply look at every club’s point differential over a 16-game season.
A team that scores more total points than its collective opponents tends to have a winning record overall, whereas a losing team generally gets outscored by its foes. The clubs with the five highest average margins of victory last year were the Patriots, Broncos, Seahawks, 49ers and Falcons. Unsurprisingly, all five made the playoffs.
By that measure, Pittsburgh won about as many games as expected in 2012. The difference between the points the team scored per game and the points it conceded in every contest was 1.4. Meaning the Steelers just barely outscored their opponents in each game, which is about what one would imagine an 8-8 team would do.
However, because a 16-game season is a pretty small sample size, this method of determining expected wins is somewhat crude. A couple of blowout wins or losses can skew a team’s point differential in such a way that it no longer matches up with its win total.
For example, the 11-5 Colts were outscored by their opponents by nearly two points per game. This is because Indianapolis won seven games by six points or less and lost games to the Bears, Jets and Patriots by 20, 26 and 35 points, respectively. And this turns out to be one of many indications that the Colts were not as good as their record suggested last year.
The other problem with using point differential to determine how many games a team should have won is that it does not adjust for the aforementioned luck associated with the quality of a team’s opponents. A team with a large point differential may have beaten poor clubs but struggled against good clubs. Though this would lead to a winning record, it is hard to argue that that team is better than competitors with tougher schedules.
Similarly, a franchise that bloated its win total on losing teams is not necessarily a high-quality outfit. And that tends to get reflected over time. Because schedules change from year to year, a team that piles up wins against bad opponents is likely to suffer a painful reality check in subsequent seasons.
Pro-Football-Reference.com attempts to adjust for this through its Simple Rating System (SRS), which adds the average total point differential of all a team’s opponents to its average margin of victory or defeat. So a team that plays a tough schedule gets a boost, and clubs that face weaker opposition week after week are downgraded.
On this scale, the top five teams (the Patriots, Seahawks, 49ers, Broncos and Packers) again made the playoffs. The bottom five—the Eagles, Titans, Raiders, Jaguars and Chiefs—didn’t get within sniffing distance of the postseason. Because the Colts had the league’s weakest slate of opponents last year, they drop to the 24th in the SRS rankings.
The Steelers also played an easy schedule in 2012. But unlike the Colts, who squeaked out wins against bad teams, Pittsburgh lost Pittsburgh Steelers" href="http://www.pro-football-reference.com/teams/pit/2012.htm">four contests and had only three double-digit margins of victory. As a result, their SRS falls to -0.7—just below the league-wide average of zero. This indicates that an 8-8 finish is a little better than what the Steelers should have expected to have in 2012.
Other analysis controls for strength of schedule by determining what a league-average team would look like and then comparing each team to it.
For example, Advanced NFL Stats calculates each NFL team’s generic winning percentage (GWP) for the year, which estimates that club’s chances of beating said league-average opponent on a neutral field.
Like SRS, GWP correlates pretty well with wins. The teams with the highest GWP in 2012 were Denver, San Francisco, Seattle, Green Bay and New England. Coming in at the bottom of the league were Oakland, Tennessee, Arizona, Kansas City and Jacksonville. The Colts’ GWP indicates that they fattened up on bad teams, as they would have defeated a league-average team only 44 percent of the time.
According to Advanced NFL Stats’ analysis, Pittsburgh would have beaten that same league-average team on a neutral field about 51 percent of the time last year. This means the Steelers were about exactly as good as an eight-win club should have been.
In a similar fashion, Football Outsiders imagines a league-average team and compares each franchise’s performance to it. That site’s metric for rating teams looks at every play in every game in a particular year and evaluates how well a club did, considering the expected outcome for an average team, the situation and the opponent. The resulting statistic, Defense-adjusted Value Over Average (DVOA), can then be matched up against a team’s win-loss record to get a sense how the club did.
DVOA comes up with the same five best teams as GWP, though in a slightly different order. The clubs with the highest value over that of an average unit were the Seahawks, Broncos, Patriots, 49ers and Packers—teams that averaged almost 12 wins apiece last year. The teams below average were also a familiar group: Eagles, Raiders, Titans, Jaguars and Chiefs. Those five averaged 4.5 wins in 2012.
DVOA also suggests that Colts are in for a rude shock in 2013. Despite winning 11 games last season, Indianapolis was 16 percent worse than a league-average team.
The Steelers turn up in the middle of the pack. Pittsburgh ranked 18th in the NFL last year with a DVOA of -1 percent. The slightly negative score was mainly due to the team’s easy schedule. A steady diet of bad teams boosted the Steelers’ non-adjusted Value Over Average (VOA) to 5.5 percent. Adjusting for subpar performance against the dregs of the league, however, puts Pittsburgh’s overall performance below the league average.
This indicates that the club was a bit lucky to get to 8-8 last year.
That assessment is also reflected in a different Football Outsiders estimate of Pittsburgh’s expected win total. To make it, the site took each club’s DVOA and adjusted it to give more weight to certain critical game situations. That metric was then used to generate an expected record based on a league-average schedule and rate of fumble recoveries. By this calculation, the Steelers should have won 7.4 games in 2012, meaning they overachieved slightly by winning eight.
What if teams are broken into their various parts and analyzed based on how their players did individually? After all, what is a team if not the sum of its collective parts?
Sites like Pro Football Focus look at teams in exactly this way. They analyze every play of every game and assign each player a numerical grade for his performance on every snap. Those scores can then be compiled to give an overall mark for his season. Using the notion mentioned above that a team is the sum of its parts, adding up the grades for the year for each of the team’s players then yields a total rating for the club.
Based on this system of measurement, the five best teams in the league last year were San Francisco, Denver, New England, Seattle and Cincinnati. The five worst were Kansas City, San Diego, Indianapolis, Jacksonville and Arizona. Though the other four finished at the bottom of the standings in 2012, the Colts would appear to be out of place, given that the team made the playoffs.
Pittsburgh’s overall Pro Football Focus score put the team 20th in the league, suggesting that the Steelers had a slightly better record than their performance merited. The franchise with the 20th-best record last year was Tampa Bay, which went 7-9.
The problem with this method of evaluating teams is that it lacks context. Not only does it not account for the opponents a club plays, it also gives no indication of how much should be expected from each player. Without defining a baseline, how can one determine whether players—and by extension, a team—over or underachieved?
Pro Football Focus addressed this recently by calculating every NFL player’s Performance-Based Value (PBV). The site’s analysts took every player’s 2012 grades and assigned a corresponding salary based on the player’s rank at his position. Comparing that number to his actual salary cap hit gives an indication of whether the club got its money’s worth out of him. Or, to put it differently, whether he underperformed or overachieved.
Summing up the PBVs of all members of the team and comparing it to the franchise’s overall payroll allows analysis of whether the collective group earned its salaries last year. Which is a useful way to assess whether a team exceeded or fell short of what was expected of it.
Though Pro Football Focus has not yet completed its assessment of all 32 NFL teams, the results so far seem to match up well with what one might expect. New England and Denver, both of which won about one less game in 2012 than Football Outsiders’ model predicts, had total PBVs that exceeded their 2012 payroll by more than New England Patriots" href="https://www.profootballfocus.com/blog/2013/02/25/performance-based-value-new-england-patriots/">$30 Denver Broncos" href="https://www.profootballfocus.com/blog/2013/02/28/performance-based-value-denver-broncos/">million. This lends credence to the notion that both teams were unlucky not to win more games.
Conversely, the Super Bowl champion Ravens won one more regular season game than expected in 2012. Their PBV was Baltimore Ravens" href="https://www.profootballfocus.com/blog/2013/02/26/performance-based-value-baltimore-ravens/">5.3 million less than their 2012 payroll, which indicates that the team did indeed overachieve in getting to 10 wins.
The Steelers got similar value from their roster last year, finishing with a total PBV that was Pittsburgh Steelers" href="https://www.profootballfocus.com/blog/2013/02/26/performance-based-value-pittsburgh-steelers/">2.7 million less than its total salary cap hit. This suggests that, like its division rival, Pittsburgh overachieved slightly last year.
What does this mean for the Steelers heading into 2013? On one hand, the news is not very good. At best, Pittsburgh finished with the record it should have. At worst, the team was an 8-8 outfit that should have gone 7-9.
Though a difference of one win is not significant enough to worry about, the numbers do not indicate the Steelers were a good team that played a tough schedule without getting many bounces. The team’s opponents were among the worst in the NFL, and Pittsburgh recovered a little more than the half of the fumbles that hit the ground during their games.
So how did the Steelers go from being a slightly overachieving 12-4 team in 2011 to a club that was lucky to reach .500 in 2012? What the stats mentioned above do not account for are injuries. The team’s players missed a total of Cowboys use health as an excuse?" href="http://www.dallasnews.com/sports/databases/20130109-nfl-team-by-team-injuries-can-cowboys-use-health-as-an-excuse.ece">52 games in 2012, the 12th-most in the NFL last year. Only eight Steelers appeared in all 16 games, the 10th-fewest in the league.
And it wasn’t just the sheer volume of injuries. It was also who got hurt.
Quarterback Ben Roethlisberger and safety Troy Polamalu missed a total of Pittsburgh Steelers" href="http://www.pro-football-reference.com/teams/pit/2012_injuries.htm">12.5 games and were not 100 percent when they returned to the field. A healthy Polamalu increased the Steelers’ probability of winning each game by .12 in 2011. In 2012, he was only able to boost the team’s chances by .05 per game. His replacement, Will Allen, had a decent year, but only added .03 to the probability that Pittsburgh would win games in which he appeared.
As a result, the average team that Pittsburgh put on the field in 2012 probably wasn’t capable of winning more than seven or eight games, and that is what should give hope for next year. If the Steelers get some luck with injuries in 2013, they should expect to have a winning record. Provided they perform as they should, of course.