Passing Inefficiency

Omahadomer has finally put some numbers behind my dead horse beaten argument that Notre Dame threw the ball far too much that last few years despite a sieve of an offensive line. If you add sacks in, Notre Dame completed… just 48% of passes attempted in 2007. More than half of our attempts went for zero or negative yards and 56 of those went for negative yards and another 9 gave the ball to the other team. What was even more shocking to me was how dreadful 2006 was in adjusted yards per attempt. Add back in sacks, and Quinn’s completion % was 57% in 2006. Here’s OD’s analysis.

The “raw” yards per passing attempt is simply yards gained passing divided by passing attempts. An average figure for a college team is usually about 6.9 yards per attempt.

These figures are always higher than average yards per rushing attempt, which might lead one to wonder why teams ever run the ball. But running is a lower-risk proposition (lost fumbles on true running plays occur with only about 1/3 the frequency of turnovers on passing plays if one includes fumbles on snaps), the risk of a zero gain is lower and yards per passing attempt overstates the net benefits of passing.

In 2007 N.D. averaged 5.2 yards per passing attempt, which was one of the worst in the nation. But really it was worse than that. In 2006 N.D. averaged 7.3 per passing attempt (which was good) and in 2005 N.D. averaged 8.7 per passing attempt, which was excellent.

However, even those big differences understate how much better the passing attack was in 2005 and 2006 than it was in 2007. Sacks are really passing attempts too, so they should be counted as passing attempts and the negative yardage subtracted from the passing total. Moreover, interceptions should be counted as about negative 50 yards. Of course, not all interceptions are created equally. Some are basically harmless (e.g., a Hail Mary at the end of the half that’s intercepted instead of being knocked down), once in awhile they’re actually helpful (e.g., on 4th down the defender reflexively catches the ball instead of knocking it down which would actually result in better field position) and sometimes they’re positively devastating (e.g., an interception returned 100 yards for a touchdown). But on average they deprive a team of a chance to advance the ball and at least to punt and change field position. So let’s use negative 50 as a rough approximation.

So let’s calculate a “net” yards per passing attempt as follows: (gross passing yardage – sack yardage – (interceptions x 50 yards))/(passing attempts plus sacks).

In 2005, Quinn’s true yards per passing attempt was 7.3 yards per passing attempt and in 2006 it was 5.7. I’m not quite ready to say that these figures are the equivalent of rushing the ball for 7.3 or 5.7 per attempt, but they do suggest that a team that can put up numbers like that legitimately might favor the pass.

In 2005, N.D.’s “true” yards per rush was about 4.6 because that’s what N.D.’s tailbacks who saw significant action (Walker and Thomas) averaged between them. So the 2005 might have rightfully been one where the play calls should have favored the pass.

In 2006 it was a more even proposition because N.D.’s tailbacks who got more than a few carries(Walker, Aldridge and Thomas) averaged about 4.9 per carry.

Now, if we turn to 2007, it’s actually hard to see why N.D. rationally tried to throw the ball at all, except perhaps to keep teams from just playing the rush. N.D. averaged a pitiful 2.5 per passing attempt (there was no meaningful difference between Clausen and Sharpley; Clausen averaged 2.5 and Sharpley 2.6). However, while N.D.’s rushing attack was not as good per carry as it was in 2005 and 2006 it didn’t see nearly the collapse that the passing game did. The five N.D. tailbacks who got carries last year (Aldridge, Allen, Hughes, Thomas and Jabbie) averaged just a hair over 4.0 per carry.

It’s probably not news to anyone, but N.D. really would have been much better of running the ball more last year. I, for one, however underestimated just how much better off N.D. would have been being a run heavy team. However, if sacks are counted as passing attempts, N.D. actually attempted to pass on 54% of the plays from scrimmage, which was similar to prior years under Weis (56-44 passing in 2006 and 50.3-49.7 passing in 2005).

I hope the basic message of it from last year has gotten home to Weis and Haywood. I expect that N.D. will be more proficient passing the ball than it was last year. But N.D. needs to commit to the run and probably be a run-heavy team for next year. Unless the “true” yards per passing attempt at least doubles to the low 5’s, N.D. is likely to be much better off keeping the ball on the ground for a large majority of the plays.

One hopes this two year trend is a “passing” one.


7 Responses to “Passing Inefficiency”

  1. Anonymous Says:

    Great analysis, but don’t forget that ND was behind in most games — this often forces more passing to get bigger plays and manage the clock. Also in most cases the opposition’s defense knew this and took advantage — along with our inexperienced line and quarterback
    I agree though, a more experienced line and seasoned running backs should help favor our running game this year — this should open up the passing game for a stronger and more accurate Claussen. I don’t think we’ll see anything close to those passing stats this year.

    Keep up the great work!

  2. Anonymous Says:

    Jimmie does not have to throw the ball more this year.He does have to accomplish 2 things this year.

    1. less interceptions 2007 7 TDS,6 INT.

    2. 2007 245 ATTEMPS, 138 COMPLETETIONS,56.3% ,1254 TOTAL YARDS. Average yards per pass 9.0869565.

    3. Throw more completeions and average at least 15-20 yards per pass,you cannot score very many TDS
    you average 10 yards or less per pass.

    4. We will be more succesfull if we run the ball more .Lets try to do all our scoreing on 3downs instead of 4 . That would eliminate less fatigue on our offensive line .

    5 The kicking game has to imrpove offensively as well as a defensively

    6. So far only 1 bowl prediction

    Konica Minolta Gator Bowl
    Jan. 1 / 1 p.m. ET Jacksonville, Fla. CBS
    Tie-Ins: ACC No. 3 vs. Big 12/Big East/Notre Dame
    Way Too Early Projection: Miami vs. Notre Dame

  3. jim / Redondo Beach Says:

    …weis turning the offense play-calling to someone that has never before performed that critical function…jeeze…but as the offense was 119th of 119 Division 1 teams in 2007…I look forward to an improvement…how could we not?…or…

  4. Anonymous Says:

    I believe with staistics you can prove just about any point you want. Sacks should be counted the way the NFL does (passing not rushing). With that said it would be nice to see a 55/45 run pass ratio. Also running is more of an attitude which ND really needs this year. It is time for ND to stuff the ball down some opponets throats and go for the jugular. When running game works it opens up play action passing.

  5. Anonymous Says:

    Thanks for the analysis. I would think too that, as either yards per pass attempt or yards per carry increase, you are likely to see a reciprocal increase in the other.

    Like you, I’m hopeful that in 2008 the Irish will open the pass up with a run-first attack. It should vastly improve Jimmy’s numbers.

  6. Joseph Says:

    From an economic perspective, the key is to match the marginal benefit of another passing play with the marginal cost (the cost of giving up the yards earned by running). The key is NOT to match the averages until avg yds per pass attempt are the same as avg yds per rush attempt.

    As an example, assume a pitcher has just a fastball and a curveball. Even the curveball is more likely on the average to generate outs, the fastball is thrown with higher frequency. Why? Because the more he uses the curveball, the less effective it becomes. Whenever a pitcher relies on his secondary stuff too much, it becomes ineffective.

    Similarly, the more a football team passes, the less effective each pass becomes. The best way to gain total yards is to find an ideal mix between running and passing, even though the avg yds per passing attempt, no matter what statistic you use, will be higher than the yds per run. Look at the avg yds per pass and run in the NFL’s successful teams to get a good idea about what an ideal ratio might be.

    From an armchair quarterback’s perspective, it certainly seems like the Irish relied on the pass too much. We became predictable, which is part of the reason why the pass rush was so effective. A better O-line and stronger run game should help raise the avg yds per run, the proportion of running plays called, and therefore the avg yds per pass as well.

    -Joe (a Notre Dame Econ/PLS major)

  7. Anonymous Says:

    Moving the chains for first downs is the key. Numerous times over the last 3 years, ND has had 3rd and 1 or 3rd and 2 and failed to pick up the key first down with a run.

    -50 on INT seems a bit much, perhaps it is closer to -35.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: