As some of you may recall, while previewing Michigan State, I took a look at how well both MSU and Notre Dame have been utilizing their time of possession. After all, it's one thing to possess the ball for any period of time, but it's quite another to actually do something productive with the ball while you've got it. And sure, there's plenty of relatively difficult to quantify benefits to simply "matriculate the ball down the field" to engage and defeat the opposition in the "field position battle," but it's also really awesome to score points while you're doing that. In fact, some might even argue points matter more than anything. Yep, can you believe that? So out of curiosity I've decided to continue tracking this metric of quality time-of-possession under then column "Time Well Spent" as an experiment and thus far the results are at best interesting, and at worst completely indicative of nothing.
Simply, "Time Well Spent" is the total time an offense spends in possession of the ball while on a drive that ultimately scores points.
So here's what we've got so far for the Irish's Time Well Spent.
| Opponent | Time of Possesion | Time Well Spent | % Well Spent |
|---|---|---|---|
| South Florida | 28:54 | 6:49 | 24% |
| Michigan | 37:01 | 16:51 | 46% |
| MSU | 27:32 | 13:16 | 48% |
| Pitt | 31:00 | 4:54 | 16% |
| Totals | 124:27 | 41:50 | 33.6% |
Such a curious thing. Easily the Irish's "best" win of the 2 thus far was against Michigan State, beating the Spartans by 18 points in a series where the Irish usually barely win by more than a score of late. And, somewhat sensibly, the Irish also enjoyed their best Time Well Spent of the season, nearly spending half their time with the ball on drives that ultimately scored points.
But in a complete turnaround from the previous week, the Irish's Time Well Spent plummeted last weekend to it's worst number yet, 15.81%, more than 8 percentage points worse than the USF game, despite only killing drives 2 times versus Pitt with turnovers as opposed to 4 times (plus that last T.O. on the final kickoff) against USF. And yet last week the Irish still pulled away with a win.
Even more baffling, against MSU, while the Irish spent nearly half of their Time of Possession on scoring drives, one of the Irish's touchdowns came on a 0 minute "possession" while returning a kickoff for a touchdown. And against Pitt, the Irish had one 14 second drive when Jonas Gray ran the ball 79 yards for a score.
At this point, I've got my doubts about this little experiment, but I'll continue to pursue it throughout the season and post updates on my findings on semi-regular intervals. Once the first BCS rankings are published, I'll take the top-5 programs and calculate their game-by-game TWS metrics to see if any patterns exist there and compare those numbers to the Irish.
And now, Don Henley...
By Jeremy September 29, 2011 - 10:59 pm
I think these need to be balanced somehow with a similar defensive stat to make sense (or recognize they are being viewed in a vacuum). Our two best performances, we scored pretty well but not great. The difference being the defense stuffing MSU and having no way to stop Michigan. Similarly, with Pitt, we spent a lot of time getting nothing and only scored 15 points, so the stat seems to be making sense.
Thumbs Up/Down:
1 
0
By Matt Q. (DMQ) September 29, 2011 - 11:03 pm
Yeah, I think there’s something to that. Something about the delta between ND’s TWS and their opponents’. Will certainly explore that.
Thumbs Up/Down:
0 
0
By The Subway Domer September 30, 2011 - 5:30 am
Good job on all of this Matt. I’m curious as to what a full seasons worth of data may tell us.
Thumbs Up/Down:
0 
0
By Erik '04 September 30, 2011 - 9:47 am
Agreed. Four data points isn’t enough to see a trend either way, especially when it’s really just 2 data points for each Win/Loss category.
I think this would be a great metric to combine with MOE, as I bet you would see a large MOE would have a low TWS, and vice versa. Or, if you have a large MOE and also a large TWS, then you can see your team must be good at overcoming mistakes such as penalties.
Thumbs Up/Down:
0 
0
By Matt Q. (DMQ) September 30, 2011 - 10:09 am
Reading my mind this time. Definitely going to layer it with MOE.
Thumbs Up/Down:
0 
0
By NDtex September 30, 2011 - 8:15 am
You read my mind on this one. There definitely needs to be some kind of defensive stat as well. That should balance out things like the Michigan game in which we continued to give up scoring drives.
Thumbs Up/Down:
0 
0
By trey September 30, 2011 - 9:36 am
Some kind of healthy balance between very simple and whatever it is NDTex is doing with his ranking formula.
Thumbs Up/Down:
0 
0
By NDtex September 30, 2011 - 1:03 pm
I’m doing science: http://cn1.kaboodle.com/hi/img/c/0/0/9c/6/AAAADHAxxPAAAAAAAJxkJA.png?v=1260927750000
Thumbs Up/Down:
0 
0
By TXIrish2 September 30, 2011 - 12:55 pm
Seems like if the stat is supposed to be a measure of offensive production, it should be normalized by the yards-to-endzone/redzone/FG range at the beginning of each possession.
For example, if we got the ball at the opponents 20 and didn’t get any points, that’s much worse than if we got the ball at our own 20 and didn’t get any points.
Thumbs Up/Down:
0 
0
By kyndfan September 30, 2011 - 1:26 pm
I still say, um game withstanding, that eating the clock in the second half to protect a lead is productive time even if you don’t score. I.E. msu
That said, can we go back to joking on PU now? It’s really fun.
Thumbs Up/Down:
0 
0
By trey September 30, 2011 - 10:43 pm
Kinda like kicking a one legged mongrel pup, eh? Where is Mike Vick when you need him?
Thumbs Up/Down:
0 
1