Furthermore, the points and yards-per-game measurements actually seem contrary to the play-to-play metric that other performances are measured by in the NFL. Running backs and quarterbacks are measured greatly by averaged yards-per-play and pass-attempt – not just total yards - and so should defenses and offenses be measured by their average outputs per play as well.
As a result, I tried a points-per-play metric to rate defenses on a more level playing field. This allows each team to be measured by the same principle – how well did they do on average per-down-played. This creates an equal measurement, with one exception – offensive turnovers -- which I'll address in a moment.
First - the results:
In 2005, the Pittsburgh Steelers were second only to Chicago in points-per-play at .258 - Chicago had .195, Indy was 3rd at .259, Seattle 4th at .260 and Denver 5th at .262 (Cleveland was 9th at .295, Baltimore was 12th at .30 and Cincinnati 23rd at .359).
In 2006 so far, Pittsburgh plummeted to 22nd at .368. Denver is 1st at .194, Jacksonville second at .226, Chicago third at .230, NE fourth at .244 and Baltimore fifth at .257 (Cleveland was ninth again at .312, Cincinnati 21st at .366).
Pittsburgh fell by .110 points-per-play - more than any other NFL team. Indy, Seattle, Carolina, TB and Washington also plummeted greatly in 2006, while NE improved the most, followed by Minnesota Baltimore, Oakland.
As there are approximately 60 defensive plays per game, that roughs out to almost seven points-more-per-game swing from 2005 to 2006.
Yes, these are the root of some issues. Turnovers sometimes create poor field position that allows for easier points. However, Pittsburgh rarely held other teams to field goals after turnovers. Take into account Oakland, Cleveland and Chicago - who have only, respectively, 1,2 and 4 less giveaways than Pittsburgh in 2006. Yet, their points-per-play ratios are far better (Oakland .46 better, Cleveland .56 better and Chicago 1.38 better.
Minnesota, with seven less Turnovers than Pittsburgh, is .80 better.
How do These Rankings Differ from Points-Per-Game?
In some cases the differences were minimal, in others great. Pittsburgh was fourth in yards-per-game in 2005 and second in points-per-play, for example. But some defenses – especially ones whose offenses were on the extremes of being very poor or good - were greatly undervalued in the points-per-game rankings. Miami rose from 18th to eighth when evaluated by points-per-play in 2005, Indianapolis rose from 11th to third, Denver went from 15th to fifth, Seattle from 17th to fourth, etc. Good offenses often allowed their defenses to pay prevent-style defenses by the fourth quarter, meaning that they let up many yards – but points? Likewise, poor offensive squads kept their defenses on the field longer. More plays may equal more points – but do they equal more points-per-play?
On the other hand, some teams fell dramatically, like Baltimore (fifth to 12th), Tampa Bay (1st to eight) and Arizona (eight to 28th) went ranked instead by points-per-play. These are often teams who's offenses may have done well in the time-of-possession game, thus protecting their defenses from playing too many plays and thus letting up loss overall points and yards. But, on a points-per-play measure, these defenses were actually less effective than many that let up more points overall.
So, this indicates that there are certainly flaws with the conventional measurements if we're evaluating defenses at their base function – stopping other teams from scoring points.
What does this mean as it relates to the Steelers?
Focusing on Pittsburgh, it helps show that Pittsburgh is letting up more points-per-play in 2006 by a huge margin over 2005, and its not just because of the turnovers.
Something else is going wrong on the defensive side.
Pittsburgh is fourth-worst in the NFL in allowing passing touchdowns – 14 so far in nine games (the Redskins are last in the NFL with 17, Denver first with six).
In 2005, Pittsburgh let up a total of 15 passing touchdowns – the Steelers are already just one off that mark. At this current pace, they will let up 24 passing touchdowns. Only St. Louis, San Francisco and Tennessee let up more than that last year (Tennessee was the NFL's worst with 33).
Instead of limiting teams to field goals, Pittsburgh is allowing teams to capitalize on long drives and turnovers with touchdowns.
Going into the Cleveland game, Pittsburgh was tied for fourth-worst in the NFL in letting up passing plays of over 20 yards – 30. Last year, Pittsburgh was 14th in the NFL in 20+ yard plays let up, with a total of 38. They had almost reached that total. At this current pace, with seven games left, Pittsburgh will allow 53 20+ yard plays in 2006. To put that In perspective, San Francisco led the NFL in such receptions in 2005 with 61 (interestingly enough, New England was second-worst with 56).
Again, before Sunday's game, Pittsburgh was "middle-of-the-road" with a yards-per-catch average of 11.7. Baltimore is the NFL's worst, at 13.5 – Chicago and Indianapolis are the league's best, at 9.8 and 10, respectively.
This is not that far off from Pittsburgh's 2005 average of 11.0. And this final statistic may hold another the piece of evidence for what's going wrong with Pittsburgh's defense. It helps show that, considering the vast increase in long receptions over 2005's number, that the other completions against Pittsburgh's defense must be shorter ones, in order to account for the fact the overall yards-per-catch average is approximately the same.
What does this indicate?
Short or long - this then seems to be the strategy adopted by offenses facing Pittsburgh in 2006. Teams are rarely attempting the 10-15 yard pass against Pittsburgh. Instead, they are either throwing underneath for short gains or going over the top for big ones. The yards-per-catch average combined with the number of big passing plays in 2006 helps makes this clear.
In the past, teams would test the middle of the field more. Now, teams are utilizing this two-layer approach and foregoing the riskier 10-15 yard pass. Basically, the philosophy has been that as Pittsburgh blitzes, either the deep pass will be there or the short pass will. The in-between passes are what suffer via the blitz – they are easiest to cover as players drop into those shorter blitz zones and the hardest to complete in the face of a blitz, so it makes little sense to get baited into trying them.
In other words, quarterbacks and coordinators are realizing that the blitz either gets there or it doesn't. If its successful, the drop-off, shorter pass works. If not, the deep pass will be there.
As a result, more running backs and tight ends are catching short passes against Pittsburgh as the short option – ten such receptions by New Orleans, six by Denver, eleven by San Diego, five by Atlanta, etc.
And when teams go deep, quarterbacks are often finding the time to make those deep completions. As a result of the increased number of deep passes, you're seeing teams throw more interceptions versus Pittsburgh this year (twelve already this year compared to 15 in 2005). But these are often less risky ones, as they are further downfield.
This is evidenced in part by the fact that the nearly every defensive back has equaled or beaten their interception total from 2005 already.
It's a difficult challenge for LeBeau. Every year sees a new adjustment to the 3-4 and new challenge, and this year has been no different. Teams are playing smarter against Pittsburgh's 3-4 defense. The latest challenge for LeBeau is how to get opposing offenses and quarterbacks to abandon their approaches and, instead, force them to re-discover the 10-15 yard passing game. That's how Pittsburgh's blitzing defense succeeds – that is, until it need no longer rely on the blitz.