Well, you've done it, Brian. You've created "the single greatest thing of all time." Now even Double-Stuff oreos pale in comparison.
Woah! Penn State and Florida State's offenses are like clones of each other. Freakish. Virtually unstoppable at short distances, pretty much average at longer distances.
But when it comes to usage, they're like the polar opposite of Florida State - a team that used its RBs *effectively*. Florida State: 3rd and 10 almost twice as much as normal. Penn State: 3rd and 10 much less than normal.
Cool graphs. Your next task is adjusting them based on strength of opponent. Matrix deconvolution, baby.
Great job Brian. Excellent work.
Effing awesome.
Thanks for this, Brian.
Prepare for 200 links to this thing.
I'm gettin' no real work done today... that's fershure! Brian: is it possible to make the refresh event execute when the radio button is clicked (switched), rather than when they lose focus? A very, very low-level enhancement, no doubt, but I know your standards are high.
Eight kinds.
I know you probably have a billion requests for graphs or analysis with all this data available... but here's another.
One thing that I'd like to see, speaking from a strictly Michigan perspective, is this sort of graph limited to just running plays, and broken up by running back.
Ie. What percent of 3rd and 1s did Mike Hart convert, what percent of 3rd and 2s, what percent of 3rd and 3s, etc. and then another graph (or an overlay) of Kevin Grady at 3rd and 1, 3rd and 2, etc.
My suspicion is that Kevin Grady is as good as hart at the 3rd and 1s or 2s, but from 3+ Hart has the knack to find the chains.
Maybe this wouldn't be interesting to anyone else though... I also thought someone elses suggestion of showing the breakdown by pass versus run would be fascinating.
Freakin. Sweet.
Check out Bowling Green. Over 40% on 3rd down efficiency from 13 yards on out.
Thanks, Brian. This totally rocks.
The NC State offense's 3rd down efficiency graph is appalling...but not surprising.
Excellent work. This is the kind of thing you may want to put up somewhere as more than just a blog entry that fades off the front page after a week.
Absolutley unreal. This is my new favorite toy.
Collecting this data and then inputting it must have been a pain, but it was worth it.
Looking at Texas' conversion rates boggles the mind.
This is inexpressably cool. Wow.
Check out Texas and SOuthern Cal. As you would expect, both are unbelievable on offensive 3rd down conversions, but look at their defenses. Texas is in the green all the way, and Southern Cal is atrocious on 3rd down. Pretty interesting.
Miami's offense... yuck! No wonder Coker jettisoned his staff. I can only guess at how the OC would address this -
"Okay, boys, listen up, 'cause this is important - If it's second down, and you have the ball, and you can see that you're not quite gonna reach the marker, for God's sake - backpedal about three yards!"
Check out the freakish green mountain for Wisconsin on third and long. Unreal.
I have a new-found respect for John Stocco. And they did it against a decent schedule (39th in SOS as per Sagarin).
Happily (for me at least!) the Badgers were just three for eleven on 3rd-and-7 to 13 against Penn State, with four sacks and an interception.
Brian: Sorry to nitpick, but on the "raw numbers" graph, you might want to change the red to green and blue to red to stick with the "green=good and red=bad" theme.
James:
That Bowling Green graph has got to be low statistics. There's just no other explanation.
Bowling Green's 3rd down distribution graph kinda confirms this. They're well below the average for number of 3rd and 12-15 they faced.
wow.
I second what Orson said. Wow.
Coolness in the extreme. First-rate work, Lawgiver.
By the way, the over/under on the number of hours that will pass between right now and the time Heismanpundit calls Brian a know-nothing self-loathing grad student is: seven.
Pat--
True. I hadn't checked out the third graph when I posted. BG was still 8-16 from 3rd and 12 to 20. Small number or not, that's still damn good.
Colorado: I'll look into it. It's an IE peculiarity. re: colors. Will do. Good suggestion.
biak: One thing that I'd like to see, speaking from a strictly Michigan perspective, is this sort of graph limited to just running plays, and broken up by running back.
Ie. What percent of 3rd and 1s did Mike Hart convert, what percent of 3rd and 2s, what percent of 3rd and 3s, etc. and then another graph (or an overlay) of Kevin Grady at 3rd and 1, 3rd and 2, etc. This data is probably too sparse to be meaningful, especially with Hart's abbreviated season. I do think looking at the various distances gained by pass vs. run is interesting and doable. It's on the list.
Chicago: I didn't enter the data by hand. I read and parsed it from the NCAA's website... I ain't that crazy.
Matt: a caveat on USC's numbers: only eight games actually got in the DB because USC failed to report several of its games.
Matt: a caveat on USC's numbers: only eight games actually got in the DB because USC failed to report several of its games.
Yet more proof that Pete Carroll is eeevilll.
I'm hard-pressed to find an uglier chart than the Temple defense. So that's what 45.3 points allowed per game looks like!
Brian - I knew you didn't, but even so that's a lot of data to restrict and manipulate. Just thinking about trying to make it work in Excel gives me hives.
Difficult or not, excellent work.
Brian, great job.
Any chance the scale could be locked at 0-100? It's a little disconcerting and deceptive when one chart goes to 70 and another goes to 90 or 100.
I've already gushed about the amazing coolness. I agree with Kevin on the 0-100% thing.
Also, shouldn't these default to Michigan rather than Air Force? Or would we rather not look at those after this season?
By the way, the over/under on the number of hours that will pass between right now and the time Heismanpundit calls Brian a know-nothing self-loathing grad student is: seven
He knows I can't compete with his riveting coverage of two sets of idiots and their billboards.
This is amazing stuff, this third down conversion chart.
I mean, what you are saying by these charts is that:
1. Teams convert more often on third and short.
2. Teams convert less often on third and long, or longer.
Absolutely brilliant!
You probably could have saved a lot of time and energy by just providing the following links:
Third Down Conversion Offense
Third Down Conversion Defense
But, I guess some people need pictures to understand some concepts.
Brian, perhaps you should consider labeling these with the year. For instance, the default chart would read "Third Down Efficiency: 2005 Michigan Offense."
"By the way, the over/under on the number of hours that will pass between right now and the time Heismanpundit calls Brian a know-nothing self-loathing grad student is: seven "
Who had the under? Not sure what time the post went up, but I saw it six hours after the over/under was called. Brian, that's two postings from him about you this week. Impressive...
By the way, outstanding work. I am highly encouraged by efforts like this...takes me back to my college days.
Um, Mr. Heisman, sir, the point is not just that teams convert more often on third and short. The point is to see whether a specific team converts more often on third and short (or third and medium or third and long) than the national average. And then the poing of the second graph is to show how often a team is put into a position of converting third and short or third and long. Just linking to the overall conversion percentage wouldn't do the trick. Seeing that your team has an overall 34% 3rd down conversion rate doesn't tell you too much. Seeing that it's half as good as the national average in short yardage situations and twice as good in long yardage situations tells you a hell of a lot more. Especially if you also see that the team is in third and long twice as often as the national average.
If every singe person who's posted is gushing about how great these graps are, why would you assume that you're the smarty pants who sees that Brian has posted something obvious? Perhaps before making an ass of yourself you should look at the damned graps.
Someone's a little cranky because Florida was worse than average in 3rd and long despite their obvious advantage in scheme. Compounding the problem is that their scheme put them in more 3rd and 12-16 positions than average.
I would have thought Urban's triple wing-T shotgun play action max protect option would have minimized 3rd and long instances and maximized conversion rate when faced with them.
Louisville, Boise, USC and Utah were all better than average in 3rd and long.
Of course, I would have been able to tell you all that just by looking at the provided link and seeing they rate 58th in the country in efficiency, 3 spots behind Army.
"1. Teams convert more often on third and short.
"2. Teams convert less often on third and long, or longer."
Let me guess; you had to repeat several grades, right?
HE IS MAN!!!
As a Notre Dame fan, I have to tell you your website is one of the greatest football blogs I've seen. You are on a pedastal with The Blue-Gray Sky in my eyes. Keep up the great work.
By the way, I do notice one small cosmetic error with your code...
Your initial graphs are Michigan graphs, meanwhile your "Select a Team;" says Air Force. You could fix this by having it say "Select a New Team:" or by making the graphs start off with Air Force, etc. You get the idea.
Brian, did you see a mention of this post on Bruce Feldman's ESPN blog? Now He Is ManPundit will really be jealous.
Yes, Kevin, I obsessively track all referrers... thanks for the heads up in any case.
Good work, Brian.
Colorful = Glorious.
Very cool...very very cool
Is there an easy way to see the number of third downs forced/faced by a defense and the number of times an offense went to third down vs. the averages? that would be interesting as well.
Not that anyone here wants to hear this, but look at Ohio St. they were as almost as good as Texas
Commenting by HaloScan
