O-Score: Final Rankings
In case you missed it, here are the first five parts to this series:
Some final thoughts on our variables before we look at the final scores:
– Turnovers: It seems like common sense that a team that turns the ball over a lot is unlikely to score a lot of points – but that’s not necessarily the case. In fact, a team’s turnover rate is only very weakly linked to their points scoring. The 1977 Cleveland Browns, 2001 St. Louis Rams and 1993 Houston Oilers all had offenses which scored 10%+ more than league average but are all bottom 2o (since 1970) teams in turnover ratio. However, of the 50 teams which turned the ball over the least, only 5 had worse-than-average scoring offenses (measured by points per drive). The conclusion here seems to be that great scoring offenses who turn the ball over a lot aren’t hurt too much since they are so efficient at turning their other drives into points. Offenses which turn the ball over a lot tend to be mediocre or worse at scoring points, but some of that is probably due to a lack of talent. After all, how many great QB’s turn the ball over a ton? How many high-quality RB’s fumble a lot? A few, for sure. But in general, elite scoring offenses are laden with elite talent – the type of which doesn’t turn the ball over and can overcome the mistakes they do make. The negatives effects of offensive turnovers are more likely to appear in their team’s defensive stats.
– Yards Per Drive: Again, the takeaway point here is that great scoring offenses generate a lot of yards per drive. Teams who have high points-per-game or points-per-drive tend to have high TD%. Because TD drives are (usually) longer than drives which end in field goals or punts, a high YPD is very common amongst high TD% teams. Since 1970, there have been 571 offenses which have been below-average in terms of YPD. Of those, only a mere 83 (14.5%) put up a PPD better than league average.
-TD% – A variable which is closely tied to Points Per Drive. Touchdown drives maximize the number of points your team scores (i.e. it’s unlikely you will be outscored in a game if you score a TD on every drive, assuming a PAT conversion, your opponent would either need an extra drive or to make a 2 point conversion). Furthermore, drives which end in TD’s obviously cannot end in turnovers or punts – both of which can lead to an opponent scoring a non-offensive TD.
– Yards Per Point – At the risk of sounding redundant, great scoring offenses almost always have high YPP rankings. However, there is a large “middle class” of YPP offenses which under-perform their expected PPD ranking. This could be due to a number of variables (shaky field goal kicker, red zone troubles, turnovers, etc). It’s rare for a below average YPP offense to have an above average PPD (about 15%). This stat can be skewed a little by a team having an elite defense or special teams – creating an unusually high number of short fields for the offense.
-Points Per Drive – This is the most important measure of an offense. It tells us what the offense does with what they are given. It’s a much better way to measure an offense than Points Per Game – as teams with terrible defenses will have fewer offensive opportunities. The problem with using PPD as the sole measure of offensive efficiency is that it doesn’t the whole story. Imagine these two scenarios:
Team A: 8 offensive drives, 8 FG’s = 3 PPD
Team B: 8 offensive drives, 4 TD, 2 punts, 2 turnovers = 3.5 PPD
While team B scored 4 more points (assuming XP and not 2 point conversions), the 2 turnovers and 2 punts are more likely to lead to their opponent scoring than Team A’s 8 kickoffs (even more the case with the new kickoff rules). Therefore, while team B scored more points, they are more likely to give up more points – assuming league average defense and special teams for both Team A and Team B. Another way to think of this is that Team B’s offensive points are worth less than Team A’s
So, in coming with my final scoring system, I took these five variables and weighted them in this order:
Points Per Drive
Yards Per Point
Yards Per Drive
All variables were taken as a percentage of league average. For example the 2007 New England Patriots had a YPD of 43.58, which was 40% better than league average. So for my rankings, I assigned them a YPD value of 45. The 2007 San Francisco 49ers had a YPD of 20.98, 33% worse than league average, so they get a value of -33.
The coefficients by which I weighted each variable were determined largely by a series of linear regression models. If you’re unfamiliar with regression analysis, it’s a way of mathematically determining the effect a variety of different variables have on another variable. The goal was to determine which offense gave their team the best overall chance to win, assuming a league average defense. Because the math is as boring as it is complicated for most, I’ll skip it for now (and revisit it perhaps in a future post for the math geeks out there).
The other, smaller, part of the coefficients is much less scientific. There’s a point to which stats unfortunately can’t explain everything. There’s simply no way to know how many times a drive which ends in a turnover would have otherwise ended in a score. Furthermore, there is very limited data (in terms of years) available on things like points of turnovers and red zone scoring %. There’s also stuff which isn’t reflected in stats, such as “can this offense effectively run out the clock when they’re winning, even if they’re not a high-scoring team” (YPD is probably the closest we can come to figuring that out statistically). To this end, I put a little additional weight on TO% and YPD and a little less weight on TD% (which is partially overlapped by PPD anyway).
Here are the top and bottom 20 offenses of all time. For rankings of every team since 1970 click here. Keep in mind that an OSCORE of 0 would be exactly average
|1||2007||New England Patriots||65.48|
|3||2010||New England Patriots||50.26|
|5||1994||San Francisco 49ers||49.22|
|6||1993||San Francisco 49ers||49.19|
|8||1992||San Francisco 49ers||48.86|
|9||1982||San Diego Chargers||47.26|
|11||2000||St. Louis Rams||45.25|
|12||2006||San Diego Chargers||43.11|
|19||2002||Kansas City Chiefs||38.33|
|20||1973||Los Angeles Rams||37.69|
|1170||1973||San Diego Chargers||-35.44|
|1172||1997||New Orleans Saints||-35.55|
|1173||1990||New England Patriots||-36.01|
|1180||1976||New York Jets||-38.76|
|1184||1976||Tampa Bay Buccaneers||-40.56|
|1185||2009||St. Louis Rams||-41.96|
|1189||1977||Tampa Bay Buccaneers||-61.32|