Take It Away, Steelers! Part 2

I ran across the following quote in an article by John Dudley of the Erie Times-News:

Two statistics in particular illustrate why the Steelers have failed to dominate. Their margin of victory is only 2.8 points per game. In each of their past three Super Bowl seasons, they won by an average of at least 7.8 points per game.

And their defense, although top-ranked, hasn't made big plays. Through six weeks the Steelers rank dead last in the league in takeaway-giveaway ratio at minus-10. The offense has turned the ball over 12 times and the defense has forced only two turnovers, one interception and one fumble. They have ranked lower than 24th in turnover ratio only once in franchise history, when they finished 27th in 2006.

Last season the Steelers took the ball away 35 times, with 21 interceptions and 14 forced fumbles. I decided to find out how many points those turnovers had resulted in, and compare it to how many points they lost on offensive turnovers (18 total last year.) My hypothesis is that the "missing" 5 points per game so far this year can be attributed to the lack of points achieved via turnovers.

This post will compare Games 5-7 of 2010 and 2011. There will be a final reckoning after the Steelers are finished for the year. After Part 1 comparing the first five games of both season, which can be found here, there were a couple of comments asking some additional questions about the data, and I added an addendum to the first post to look at those factors. 

In this post I have included the additional analysis —the point swing that takeaways or turnovers represent, and whether the resulting points, if any, can be blamed on the defense (or be expected from the offense in the case of a turnover.) The latter will be noted by giving the Probability Point for the field position at which they began, giving the end result, and grading the offense or defense with a minus (below the line,) 0 (did their job, or not really their fault) or + (better than expected.) If this doesn't make sense, the rather long-winded explanation is in the addendum post (link above.)

Game 5 (2010) vs. the Browns was Ben's first game back after his suspension. The defense decided to throw a little "welcome back" party for him via a couple of INTs, one by Lawrence Timmons, the other by Ryan Clark. Harrison threw in a FF for good measure. Keyaron Fox also recovered a muffed catch by the Cleveland punt returner. The new, kinder Ben decided to welcome rookie Joe Haden into the league by gifting him a pick.

Rookie QB Colt McCoy began the game, his first ever in the NFL, with a 3:30 minute series culminating in a throw to Ryan Clark, who picked up another 11 yards. Ben got the ball on the PIT 29, and 4:46 minutes later, with the snap at the CLE 14, Ben throws the ball to Joe Haden.  The Steelers D, who was probably pretty annoyed by this point, held the Browns to a FG.

The Harrison FF was early in the 2nd quarter, when Joshua Cribbs took the snap in a wildcat formation. It was, however, recovered by CLE, but Cribbs was out with a concussion, the first of two Harrison was held responsible for.

After that both teams kept their noses, although not their jerseys, clean until the crazy incident in the 4th quarter where Sepulveda had to re-punt twice because of penalties on Anthony Madison. The third time was a charm, as the 34-yard punt was muffed by Stuckey and Keyaron Fox grabbed it at the CLE 13. This time the offense was more grateful, and converted it to a TD.

A couple of series later Timmons intercepted Colt McCoy at the CLE 23 and ran it back to the 18. Ben converted it to a TD 3 plays later.

TA 1: PIT 45 (2.25) + PIT 29 (1) = 3.25 Point Swing

      Probability Point: 1    Result: TO    Offense -

TA 2: CLE 13 (-.2) + CLE 13 (4) = 3.8 PS

      PP: 4    Result: TD    Offense +

TA 3: CLE 15 (1) + CLE 18 (3.85) = 4.85 PS

      PP: 3.85    Result: TD    Offense +

TO 1: CLE 14 (4) + PIT 35 (2.9) = 6.9 PS

      PP: 2.9    Result: FG    Defense 0

Turnover differential for 2010 Game 5: +2

Turnover point differential:  +11

Final Score: Steelers 28 Browns 10

Game 5 (2011) vs. the Titans was, finally, a triumph of the offense. Neither team had a turnover until the end of the 2nd quarter, when Ben was intercepted by Courtland Finnegan. Fortunately the Titans only had 8 seconds left and didn't manage to do anything with the ball.

The sole Pittsburgh turnover came near the end of the 3rd quarter. After Chris Johnson ran a yard for a TD, TEN recovered an onside kick. This apparently really annoyed the defense, because in a fit of temper Brett Keisel tipped Matt Hasselbeck's very next pass, and LaMarr Woodley grabbed the ball to put the offense back on the field near midfield. About 6 minutes later Shaun Suisham kicked a FG.

But Tennessee wasn't finished, and blocked Sepulveda's punt 4 minutes later. Courtland Finnegan ran 30 yards for a TD, which was nullified by penalty. So Matt Hasselbeck got the ball at the PIT 37. (I'm not counting this as a TO, because it was Tennessee's ball - they just got it at better field position than they probably would have otherwise.)

The Titans got the ball at the PIT 37. Cameron Heyward forced a fumble during his first sack, but TEN recovered. Three snaps later Heyward again grabbed Hasselbeck, who incurred an intentional grounding penalty, thus denying our rookie his sack. Hasselbeck then converted a 3rd and 19 into a TD, which was fairly annoying.

Three plays later Ben had another TD and the game was for all intents and purposes over.

TA 1: PIT 45 (2.25) + PIT 46 (1.65) = 3.9 Point Swing

      PP: 1.65    Result: FG    Offense +

TO 1: TEN 27 (3.3) + TEN 34 (1.2) = 4.5 PS

      PP: 1.2    Result O   There were 8 seconds left in Q2, so the D doesn't get a       lot of credit for this

TO 2: Recovered onside kick: This is hard to calculate, but let's assume that it would have been a TB, so:

PIT 20 (.3) + PIT 47 (2.25) = 2.5.5 PS

      PP: 2.25    Result: TO    Defense +++ : )

Turnover differential for 2011 Game 5: -1

Turnover point differential: +3

Final Score: Steelers 38 Titans 17

Game 6 (2010) at Miami wasn't the 2010 Steelers' finest hour. They left with a win, thanks to a couple of lucky breaks.

It didn't exactly start with a bang. Emmanuel Sanders fumbled the opening kickoff, and it was recovered by Miami. They quickly converted it to a FG, and the Pittsburgh offense finally took the field. On the third play from scrimmage Ben was sacked, fumbled, and Miami recovered. Dang. It was looking like it was going to be a long day.

But the defense finally started to even things up in the 2nd quarter when Bryant McFadden forced a fumble during the tackle of Brian Hartline and James Farrior recovered it. A couple of minutes later Ben threw a pass to Hines Ward, who ran it into the end zone.

Ben fumbled again in the 4th quarter as he was being sacked, but an alert Maurkice Pouncey grabbed it. Then at about 2:30 left to play came the incident that the Miami faithful are probably still bitter about. Ben ran the ball into the end zone from the MIA 2, and it was initially declared a TD. The Dolphins challenged that the ball actually made it in and the ruling was reversed. Ben had lost the ball after he hit the ground, and since it was initially ruled a TD the officials hadn't bothered to sort out the pile that ended up on top of it. Although it looked extremely probable that a Dolphin had gotten it, you couldn't see that for certain from the replay, and so the ruling was that Pittsburgh retained the ball. The FG Jeff Reed kicked was the game winner, as it turned out.

The Dolphins got the ball at 2:26, but a James Harrison interception a few plays later put their hope of a comeback to rest.

TA 1: MIA 38 (1.2) + MIA 34 (2.9) = 4.1 Point Swing

      PP: 2.9    Result: TD    Offense +

TA 2: MIA 33 (1.1) + MIA 32 (3.1) = 4.2 PS

      PP: 4.2    Result: this ended the game

TO 1: PIT 22 (.3) + PIT 22 (3.7) = 4 PS

      PP: 3.7    Result: FG    Defense 0

TO 2: PIT 21 (.3) + PIT 13 (4) = 4.3 PS

      PP: 4    Result: FG    Defense 0

Turnover differential for 2010 Game 6: 0

Turnover point differential: +1

Final Score: Steelers 23 Dolphins 22

Game 6 (2011) vs. Jacksonville was another turnover-free game. Despite getting 3 sacks no Steeler forced a fumble. One Jaguar forced a fumble out of Ben that Trai Essex recovered.

Turnover differential for 2011 Game 6: 0

Turnover point differential: 0

Final Score: Steelers 17 Jaguars 13

Game 7 (2010) at New Orleans was filled with weirdness in the stands as the Guinness Book of World Records officially recorded it as the largest costume party ever.

The game itself was another wash in terms of turnovers, but the result wasn't as satisfactory as the previous week at Miami, with the Steelers losing by 10 points. It was also the Battle of the Red Flags, and Sean Payton won that one. His challenge that the Randle El catch-and-run was a TD resulted in a reversal, but Tomlin's challenge that a Saints punt returner who fumbled was not down by contact when the ball came loose was upheld.

Nothing further occurred until nearly the end of the 2nd quarter, when a deep Drew Brees pass was intercepted by Ike Taylor. It was all for naught, though, as the resulting Jeff Reed 51 yard FG attempt was no good.

In the middle of the 3rd quarter Sanders fumbled yet again at the end of a kick return, but Isaac Redman alertly grabbed it. Ironically, the kicker, Garett Hartley, was injured whilst tackling Redman.

During the middle of the 4th quarter the defense gave the offense one more chance to make things right. Bryant McFadden sacked Drew Brees and forced a fumble that LaMarr Woodley recovered. Two plays later Heath Miller fumbled the ball when trying to get a few extra yards. He already had 19 YAC on a 2nd and 8, and should have left well enough alone. NO recovered and four minutes later Drew Brees converted it to a TD.

The Steelers got the ball back with 2:37 to play, trailing by 10 points, but Ben threw a short pass to Wallace that was picked off by a DB cutting in front of him, and the game was over.

TA 1: NO 8 (-.4) + NO 36 (2.9) = 2.5 Point Swing

      PP: 2.9    Result: missed 51-yd. FG    Offense -

TA 2: PIT 20 (3.8) + PIT 27 (.9) = 4.7 PS

      PP: .9    Result: TO    Offense -

TO 1: PIT 41 (1.3) + NO 45 (1.6) = 2.9 PS

      PP: 1.6    Result: TD    Defense -

TO 2: PIT 44 (1.6) + PIT 31 (3.2) = 4.8 PS

      PP: 3.2    Result: end of game   

Turnover differential for 2010 Game 7: 0

Turnover point differential: -7

Final Score: Steelers 10 Saints 20

Game 7 (2011) at the Cardinals was historic for this season, in that it represents the very first game in which the Steelers had a positive TO differential. Woohoo!

Ryan Clark's interception of Kevin Kolb on the very first series of the game resulted in a touchdown pass to Heath Miller five plays later. Way to go, Ryan! (And Ben and Heath, of course.) As usual, Ben had a couple of near-interceptions, but those don't count any more than the ball that hit Troy Polamalu right in the chest, because he dropped it.

TA 1: ARI 27 (.8) + ARI 31 (3.2) = 4 Point Swing

      PP3.2    Result: TD    Offense +

Turnover differential for 2011 Game 7: +1

Turnover point differential: +7

Final Score: Steelers 32 Cardinals 20

So how did my theory fare when we look at the next three games? The Steelers managed a net 5 points from TOs/TAs during the first four games of the season last year, although they were already +7 in turnover differential. In the next three games the net was an additional 5 points, for 10 total so far, and the TO differential was +9.  

If we look more closely we can see that once again in every win except one in the next three games, the games in which the Steelers had a positive turnover point differential were won by more than 7 points. Games with a neutral or negative point differential resulted in either a win by less than 7 points or a loss. The sole exception is the Titans game this season. The TO differential was -1, but the Steelers won by 21 points. This means that over the course of 14 games this hold true in all but two instances.

It will be interesting to see if this moves toward the mean. Presumably it will, if the Steelers keep winning without generating a positive TO differential. The question is, can it be done on a consistent basis? Of course, they could be winning games by less than a TD, in which case the principle would still hold.

At this point my theory doesn't look too promising. The turnovers generated by the defense in Games 5 - 7 in 2010 resulted in a total point swing In the next three games of 26.4. The offense did a much better job of capitalizing on them with Ben at the helm, and converted them to 21 points. However, they gave up TOs resulting in a total point swing of 22.9, and the opposing teams garnered 16 points from them.

Oddly, the TO point differential was actually larger this season during Games 5 - 7, with a net +5 last season and +10 this season. A good trend, in my opinion! It doesn't help my theory much, though.

This prompted me to challenge the initial assertion that the Steelers are winning by a margin of just 2.8 points per game, as opposed to a margin of 7.8 points per game last season. It finally occurred to me is that this isn't really comparing apples to apples, because you're comparing six weeks of games to, presumably, 19 weeks (or 16 if the 2010 figure only includes the regular season.) 

Therefore I decided to calculate the margin of victory for just the first 7 games of both seasons. Last week it was indeed a 2.8 MOV for this year's club after Game 6, but they have moved up to 4.1 after last Sunday. After game 6 the figure was a 9.6 MOV for the 2010 Steelers, but after Game 7 it moved down to 6.45. (The MOV incorporates how much each team lost by as well, and of course Game 7 was the loss to New Orleans last season.) So I'll be including each year's MOV at that point in the season at the end of each set of games.

And while I'm at it, this year's TO/TA ratio is now -9. That's nothing to write home about, but again it's a trend in the right direction. After 7 games the 2010 Steelers were at +9. In a way, it's astonishing that this difference of 18 has so far resulted in the same W-L figure of 5-2 after Game 7.

The Steelers are moving on from the huge hole that Game 1 put them in, both in the TO category and the MOV. Let's hope it continues.  I'm not writing the 2011 Steelers off until such time as they have reached a point where it is mathematically impossible to make the playoffs. May that day never arrive. Don't break my heart, guys!