Great stuff, Brian. What other brilliant statistical analysis devices do you plan to unleash on the unsuspecting college football populace?
This really is great stuff Brian. I agree that there is a ton to learn about a team based on what they are presented with on 3rd down and how they handle it. My only concern is whether focusing on 3rd down misses some of the benefits of a really high-octane offense like Texas or USC or Spurrier's Florida that doesn't bother with third down as much as a methodical, plodding offense like Ohio State that makes its living on third down. Conversely, what about a big-play defense that's willing to cede some long drives and big plays, but has a great turnover ratio. I can't think of anyone now, but maybe some of the old Miami Ds. Do the numbers show anything odd there?
Am I the only one who lacks the brain capacity to follow the graphs? I seriously feel stupid. Maybe I just am stupid. I suppose that's a very real possibility.
What don't you understand, Y2K? I will endeavor to answer any questions.
It's interesting that the notable red spot on the Florida State Offensive Efficiency Graph corresponds directly with the 3rd and 10 that they faced so often. I would imagine that that weighting would tend to push FSU back near overall conversion % of HTAT and that's considerably below where most 'nole fans would be satisfied.
Do you have the ability to see conversion success rate by type of play (i.e. pass or run)? It would be entirely too much information at each distance, but say at intervals like 3rd & short (n7)?
I'm curious as to play-calling tendencies for one thing...but I also want to see if the conventional wisdom (run if short, pass if long) has more or less success than innovators.
Oh, and this is good stuff, by the way.
i bet if you graphed ohio state's graduation rates, you would see the exact opposite of their efficiency graphs...only multiplied by 10.
Dude, that rocks. Well done.
Math make brain hurt.
Great stuff, Brian. Seeing other school's data made me keep flipping back to UM's stats. I think this shows one thing that we all basically knew all along: We were decidely average this year. One particular issue that now stupifies me is why our D was so sub-par in short yardage situations. Given our tendencies, that should have been our strength. I'm willing to bet we were ahead of the curve in '03 and '04. Stripling??!!
So now what?... We all need to start coming up with dependant variable candidates for Brian to collect data on.
Good point Cblue. 04 data should be interesting. We were running 3-4 exclusively (or did we mix it up on third and short) and i can only guess that 4-3 would be better against the short yardage run, so if o4 was better the we have some serious stuff going on.
Great work LAWGIVER, continue the "ass kicking" (how many kinds is it anyway?)
Dependent variable candidate = Possesions resulting in a touchdown. What were the third down effiency measures as well as, the first-to-first down and second-to-first down conversion rates on those drives relative to others ending in punts or field goals. We would have to filter out all returns (kickoff, punt and interception)for touchdowns, of course.
What's really scary is how good OSU was on 3rd and Short. With 1 yard to go, they were at 54% (it looks like) to 67% on average? That is some drive killing goodness.
Brian, this stuff is really awesome. It would be interesting to see The Ohio State through the Penn State game and post-Penn State. They were a completely different team after leaving Happy Valley. (Illustrating, once again, the challenges involved when analyzing records if one does not account for strength of schedule.)
Continually cited is the spread's difficulty in short yardage and goal-line situations, but one would figure this goes hand in hand with increased efficiency in the middle distances the classic dink-and-dunk spread offense is designed to get on every play.
I have more than a passing interest in this subject, as Penn State will reportedly meet with Indianapolis Colts staff during the off-season and install the spread. (Colts Assistant Head Coach/Quarterbacks Coach Jim Caldwell is a former PSU assistant.)
I think a more telling measure for a spread offense is their first-down efficiency. That is, the percentage of the time they gain 5+ yards on the first play of each series.
One particular issue that now stupifies me is why our D was so sub-par in short yardage situations. Given our tendencies, that should have been our strength. I'm willing to bet we were ahead of the curve in '03 and '04. Stripling??!!
Well, Massey's move inside probably didn't help things, but I think it's a combination of Michigan's staunchly conservative approach in '05 and the inability of either Burgess or Graham to shed a blocker and make a hit at or behind the LOS.
Kevin: The post-PSU numbers for the OSU offense would probably be astronomical, since they played Illinois, Indiana, MSU, NW, and ND after that. Michigan's defense was the best they faced over the entire second half of the season and it wasn't close--nor is was the M D particularly good anyway. The problem is that I'm already stretching the data about as far as it can go using 12 games... six would be iffy.
I have always been interested in a rough gauge of how many points on average a turnover leads to for the opposing team? Can you calcuate some overall statistic to say on average, a turnover leads to X points? Sort of analogous to baseball and how many runs (or fraction of a run) does an error lead to.
Nice Work. Very illuminating.
Smoke 'em if you got 'em,
The Doctor
Michigan State had a good offense?!?!?! They had the worst offense in the Big 10 last season. I am confused.
Commenting by HaloScan
