Per drive statistics - Printable Version

+- The CardBoard (https://thecardboard.org/board)
+-- Forum: C-House! (https://thecardboard.org/board/forum-4.html)
+--- Forum: The CardBoard (https://thecardboard.org/board/forum-5.html)
+--- Thread: Per drive statistics (/thread-10209.html)

---

Per drive statistics - BostonCard - 10-13-2015

The discrepancy between Utah's yards per play and their wins/losses made me wonder if there is something about that metric that was flawed, or if the Utes were just lucky to win games they really shouldn't.  For that reason, I decided to look at the per/drive statistics.  I like per-drive statistics in that in any game, each team will have roughly the same number of drives (usually within +/- 1, barring successful on-sides kicks).  Moreover, per drive statistics reward consistency in offense (or defense), in that an offense that consistently gains 4 yards each play will sustain longer drives than an offense that gains 16 yards on one play, and then gets three consecutive plays that gain 0 yards.  Per drive stats also incorporate drive killing penalties or the ability to draw drive sustaining ones (or, on defense, the converse).  Thanks to the FBS Drive Stats site (http://fbsdrivestats.com/categorized-rankings-2015/), this is easy to obtain.

So, here are the per drive statistics:
Average drive starting position, expressed in yards to go until the goal line (first offense, then defense, then the difference):



Interesting that we lead here by a very large margin, despite not having great special teams or many turnovers (both of which would result in good field position vis a vis our opponents).  I think it is largely due to the fact that we rarely go three and out (only 13.3% of the time while our defense forces our opponents to go three and out 31% of the time.  Sadly, this may also reflect our coach's penchant for punting from the opponent's side of the field.  Regardless, we are winning the "field position" game, by a large shot.  By the way, contrary to my expectations, Utah is merely average here.

Next is yards per drive (offense, defense, and difference):


Stanford and USC are tied atop the conference in net yards per drive, but, of course, we have the advantage in starting position.  Utah, despite its poor yards per play differential, does well here, and is able to sustain drives much better than its opponent, probably because their strong running game means they are able to gain the first down more consistently than their opponent.

Finally, points per drive (offense, defense, and net):
Unsurprisingly, given our best in conference field position advantage and tied for best in conference yards per drive advantage, we are tops in the conference in points per drive.  We easily lead the conference in points per drive scored (the only team above 3), which means we would have the top scoring offense in the conference if we had as many drives as the other teams in the conference.  However, because we don't use a tempo offense, we have fewer drives than most.  USC and Utah are tied for second in net points per drive, and interestingly, for Utah, it is because they are the best team in the conference at preventing their opponents from getting points from each drive (they are third in the conference in defensive yards per drive), so it is fair to say that they epitomize the idea of "bend but don't break".

I draw a few conclusions from the analysis.  First, we really are quite good, and arguably the best team in the conference.  Second, USC is dangerous, and you might not have expected them to lose twice given their statistics.  The underlying drama in the USC locker room has likely produced more inconsistency than you would otherwise expect, with an ability to absolutely trounce a team one week, and then lose inexplicably the next.  Third, Utah is probably better than it's per play statistics indicate.  I suspect that its consistency on offense means it is able to sustain drives better than you might expect based on its per play stats, and its bend-but-don't-break defense gives up yards (especially per play) but few points, helped in part by turnovers.  Contrary to my expectations, though, Utah is pretty average in field position.  Fourth, as I suspected, Cal is a pretty average team; it is reasonable to expect them to struggle with Stanford, USC, and UCLA (and possibly ASU).  Fifth, Oregon really is as mediocre as its record has indicated.  It's in the bottom third of the conference in points per drive (0 net), and only has a 2.2 yards per drive advantage over its opponents.

BC

---

Re: Per drive statistics - thunder_chik - 10-13-2015

Hope we don't get too cocky over all these great stats.  Thanks for sharing... did you crunch all these numbers yourself?

---

Re: Per drive statistics - stupac2 - 10-13-2015

(10-13-2015, 05:22 PM)thunder_chik link Wrote:Hope we don't get too cocky over all these great stats.  Thanks for sharing... did you crunch all these numbers yourself?

Fans getting cocky doesn't affect the outcome of the game, and I doubt players/coaches are aware of these stats.

---

Re: Per drive statistics - BostonCard - 10-13-2015

(10-13-2015, 05:22 PM)thunder_chik link Wrote:Hope we don't get too cocky over all these great stats.  Thanks for sharing... did you crunch all these numbers yourself?

(10-13-2015, 04:25 PM)Boston Card link Wrote:Thanks to the FBS Drive Stats site (http://fbsdrivestats.com/categorized-rankings-2015/), this is easy to obtain.

I don't think I was getting too cocky; I certainly didn't collect the numbers to show how great we were.  My primary aim was to try to explain the discrepancy between Utah's per-play stats and their record.  Their per-play stats predicts a team that would be 2-3; their record, of course, is 5-0.  I think their per-drive stats go a wasys to explaining that; you might expect them to be 4-1 based on their drive stats, and of course all records do have some element of fortune and playing a "good game" against a tough opponent and having your B or C game crop up against an overmatched opponent that you can beat anyways.

Honestly, while I wasn't surprised we came out near the top, I was surprised per-drive metrics showed Stanford as the top (or one of the top-two) teams in the conference.  Let's hope it stays that way after Thursday.

BC

---

Re: Per drive statistics - washingtonismoney - 10-14-2015

Here's a related concept: http://www.footballoutsiders.com/fei-ratings/2015/fei-week-6

---

Re: Per drive statistics - CTcard - 10-14-2015

Been meaning to add a thanks to BC for putting this up. I have been meaning to get around to looking at the drive stats.

(10-13-2015, 04:25 PM)Boston Card link Wrote:The discrepancy between Utah's yards per play and their wins/losses made me wonder if there is something about that metric that was flawed,

If you are trying to ask a question as broad as "How good is a football team", every metric is flawed. The major, and obvious, issue with yards per play is that it vastly favors a 3 play 70 yard TD drive over a 12 play 70 yard TD drive. I see no reason to do that for questions associated with "how good is a team/offense/etc." .

And yeah, I know you more or less mention this.

Quote:Average drive starting position, expressed in yards to go until the goal line (first offense, then defense, then the difference):

Code:

           LOS    LOS    LOS advantage
Stanford    68.2    75.4    7.2
UCLA    Â        68.3    73.0    4.7
ASU    Â        69.2    72.4    3.2
Oregon       66.6    69.6    3.0
Washington    68.0    70.8    2.8
WSU    Â        68.2    69.7    1.5
Utah    Â        70.4    71.0    0.6
Low APR U    68.8    68.4    -0.4
Arizona       70.9    69.3    -1.6
$C          74.6    71.2    -3.4
Colorado    70.6    65.6    -5.0
OSU    Â        73.4    65.2    -8.2

Have to say I haven't thought a lot about average starting position. Maybe I should. Though I do note that the best to the worst starting positions (for the offense) only varies by 5 yards. Not a ton.

Quote:Next is yards per drive (offense, defense, and difference):

Code:

       YPD    YPD    YPD net
$C          41.0    30.5    10.5
Stanford    40.8    30.3    10.5
UCLA    Â        35.7    27.8    7.9
Utah    Â        36.6    28.9    7.7
Low APR U    37.5    31.7    5.8
ASU    Â        30.0    27.0    3.0
Oregon      34.4    32.2    2.2
Arizona      37.7    36.0    1.7
WSU    Â        36.7    35.1    1.6
Colorado    28.4    31.3    -2.9
Washington    25.2    29.6    -4.4
OSU    Â        28.5    34.4    -5.9

Even here the best to the worst don't vary by as much as I would have guessed. 12.5 yards on offense, 9.0 yards on defense. I guess that's a reminder of the fine line between failure and success.

Quote:Stanford and Under Stanford Control are tied atop the conference in net yards per drive, but, of course, we have the advantage in starting position. 

Strictly speaking, for the purposes of this stat, don't we have the disadvantage of field position?
Even a TD drive, on average, can only cover 68 yards, whereas $C's can cover almost 75.

Quote:Finally, points per drive (offense, defense, and net):

Code:

        PPD    PPD    PPD net
Stanford     3.11    1.76    1.4
$C                 2.95    1.61    1.3
Utah    Â         2.67    1.38    1.3
UCLA    Â         2.84    1.60    1.2
WSU    Â         2.61    2.00    0.6
ASU    Â         2.16    1.68    0.5
Low APR U     2.42    2.02    0.4
Washington     1.58    1.46    0.1
Oregon       2.52    2.51    0.0
Arizona       2.8    3.0    -0.3
Colorado     1.95    2.33    -0.4
OSU    Â         1.78    3.02    -1.2

The site is quite specific about what they consider a drive. I couldn't see such details on points. Do they count defensive scores in these numbers?

Quote: for Utah, it is because they are the best team in the conference at preventing their opponents from getting points from each drive (they are third in the conference in defensive yards per drive), so it is fair to say that they epitomize the idea of "bend but don't break".

Interesting.
An aspect of Utah's success I had not noticed.

Quote:I draw a few conclusions from the analysis. 
First, we really are quite good, and arguably the best team in the conference. 
Second, Under Stanford Control is dangerous,... 
Third, Utah is probably better than it's per play statistics indicate. 
Fourth, as I suspected, Low APR U is a pretty average team;
Fifth, Oregon really is as mediocre as its record has indicated. 

First - I hope so.
Second - Will be interesting to see if $C still wants to play hard.
Third - One way or another, that was clear. This says something as to why/how.
Fourth - Yeah ...
Fifith - ... well nothing really to add here.

---