I thought that myself when I read that article that talked about a Data Scientist(tm)
i am very proud not to have taken unnecessary shots at the south in this post
'de-moh-NAY!' s'il vous plait.
The NCAA has published its 2013 data submitted by member institutions for the purposes of Title IX compliance. You can download the spreadsheets from ope.ed.gov.
Politics refresher: Title IX is a gray area topic since it is political but affects college sports which this blog is about. This is a feel thing: it is logical to point out that a male wrestler's experience will be more similar to that of any female basketball player than Derrick Walton's, it is politics to label that "reverse discrimination."
Quinze, seize you: Generally BCS teams spent between 37% (Stanford) and 77% (Oklahoma State) less on the women's sports than the men's. Michigan spent about $7.00 on the fellas for every $3.00 on the gals, a ratio near the top. BCS schools, private schools (who didn't used to have to comply) and Southern schools tended to higher disparities; among the 15 lowest women-to-men expenditure ratios all but three (Minnesota, ND and Pitt) were in the Confederacy. The Dept. of Education doesn't regulate an annual expenditure ratio between men's and women's sports, but they look at them as part of the nebulous compliance system.
|Avg Expenditures by Conference
(in millions) 2012-13
Building Lies. Weirdly, expenses appear more normal than the revenues, which get downright weird. A few examples (for reference, Michigan's men's hockey team reported revenues of $3.2 million, the 4th-most in that sport):
- Stanford's women's basketball team, which was a 1 seed that lost in the Elite 8, reported $16.5 million. The next-highest is Baylor's ($5.9 million), Vandy ($5.6 M), Tennessee ($4.9 M) and UConn ($4.7M)
- Clemson's women's diving reported revenues of $406k. Only two other schools reported any revenue for that.
- TCU said they made $3.4 million from horseback riding and $416k from women's rifling.
- Southern's women's soccer team, which didn't make the tournament field, reported $3.1 million in revenue, which is more than their football team and almost as much as all of their men's sports combined.
- Robert Morris's women's hockey team reported more revenue ($1.1 M) than its men's team ($997k).
- Michigan's men's lacrosse team led the country in revenue: $2.4 million
- Wisconsin's women's ice hockey reported $7.6 million; their men's team reported just under $12 million (double what next-highest, Minnesota, made).
Michigan's the rare school that doesn't pretend its opulent escalator entrance was built for the women's gymnastics team. [MGoBlue.com]
Wisconsin's hockey numbers might be a clue as to how these schools are getting their numbers. The Badgers recently built a practice facility adjacent to the the Kohl Center with donated funds; the women's team plays their game there. Stanford got a massive donation' last year from its version of Ross and built a multi-sport athletic facility with his name on it. Michigan appears to have funneled some of their Big House improvement through lacrosse.
It appears what's happening is when a donation is put toward a building project the schools tend to split that between whichever teams use it. End result: teams that funded major construction projects ended up with the highest ratios of $$ spent on women versus men.
Biggest liars? There's no way to figure out the accounting for these things but it's obvious some programs play with the books more than others. TCU is pretending they built a $3.4 million storage shed for saddles and bridles that the football team just happens to use as an indoor practice facility. They also upgraded the ROTC rifling range, which they attributed to the women's team. They're a private school that
to be a women's college and is still 57% female [ED-S: apologies—you have no idea how many people I've repeated that factoid to over the years] that spent the last decade trying to become a BCS program, which explains the fiscal acrobatics.
[After the jump, comparing expenses to recruiting and performance]
Last week I started playing with Lemming's recruiting information to see how national recruiting had changed over the last few decades. Too late for that article I realized I could actually take this study back more than a century using lists of historical lettermen published by various teams in their annual media guides. Using the same regional breakdown I tried to get data for the same six I used in the old decimated defense series—Michigan, MSU, Ohio State, Penn State, Notre Dame, and Alabama—but only three had accessible info.
These are based on letterwinners, not entire rosters, so scholarship freshmen on the scout team aren't counted. It still gives us a picture:
Vestibulum ut maior
First the goofy things. The breaks in Alabama's lines are 1919 and 1943, when the Tide didn't field teams because too many players were fighting World Wars I and II, respectively. That big spike for Bama right before WWII is because they were used as an officers training base just prior to the war. Michigan got this same temporary—by 1943 those officers had shipped overseas—bump, but not as many registered since most came from the Midwest (e.g. Elroy "Crazy Legs" Hirsch from Wisconsin).
Second the obvious things. Notre Dame's claim to be a "national" program is defended by the data. From the 1920s through the '70s their teams were fielded from between 30% and 60% by players from outside of the Great Lakes States (plus Iowa). Since the '80s about 65% is the new normal. Michigan and Alabama tracked pretty similarly as major state schools who kept to their home regions up until Bo arrived. Beginning in 1969 Michigan began a trend upward that finally settled over 30% of players from outside their region. Alabama remained a predominantly regional program until Nick Saban arrived.
I zoomed in on the years since 1964 so we can better appreciate how coaching changes affected the programs:
Tangere facere magnum
Ewww don't touch me. I trust you know who Michigan's and Alabama's coaches were in 1964. The late '60s were the last time Michigan was as regional as Alabama. That now seems in danger of happening again. Last year Michigan dipped to 25.67% and Bama peaked at 25.33%.
Michigan graduated five players from outside the Midwest (Dileo, Gallon, Gibbons, Lewan, and Qwash), and two more (Ash and Furman) aren't returning for fifth years, but they brought in nine this year (Peppers, JBB, Winovich, Pallante, Watson, Speight, Mone, Cole and Canteen) so I expect their total to climb a little for 2014. It's too hard to say what Bama's number will be since they still need to cut 10 or 11 players before fall.
Championships win defense. Carr's 1997 championship, Parseghian's 1966 one, and Saban's 2009 title were all followed by steep increases in national recruiting; of the 14 titles in that chart two (Holtz's in 1988 and Stallings's in 1992) were accompanied by drops in the % of roster made up of extra-regional players. Michigan's historical peak is 2000, three years after their only TV-era national championship and one year after their Orange Bowl victory. The great Hoke classes of the last two years were very local.
[After the jump, I test another culprit]
Yesterday Ace posted a link to the full Lemming recruiting rankings from 1990 to 2004. Just perusing the list is pretty interesting, since accessible recruiting data on a national scale otherwise only goes back as far as the Rivals and Scout databases. Since nobody likes to make their information easy to get at, it'll take some time for this all to be processed.
But for a first stab I did find something I can pull relatively easily from both Lemming's sheets and modern data: where players come from, and where they went. Lemming only had data on where recruits were from going back to 1999. Since it was easiest to grab a Top 400 from 247, I took theirs too, but they run out of rankings before 2008 so there's a gap. It won't matter for this. I broke the nation into regions that quasi-match the traditional conference footprints:
And here's the % of high school recruits that each contributed to Lemming's (on the left) and 247's (on the right) lists:
Who's been telling you that demographics are responsible for the SEC's rise? It's not there. The Big Ten's traditional footprint was providing 15% of the nation's talent in 1999 and the SEC was around 35%; today it's almost the exact same.
[After the jump: regional retention]
One Play. I got really into this piece by Brhino where he went over Michigan football seasons going back to the "Year of Infinite Pain" (i.e. 2005) to point out games where one play may have meant the difference. Interesting way to reassess how we view the seasons. For example this year's team was a couple things going right away from 11-1 and a BCS bid (MSU would have still have won the Bo Division, with BCS eligibility riding on that), and a couple of things going wrong away from 4-8. I chart:
Bicking makes it cligger.
Quibble: I may be stretching "one play" too far, but Football Armageddon had that late hit out of bounds by Crable on 3rd and 15. OSU scored on that drive to go up 10. Who knows if Michigan can drive the ball the same as they did on the next possession. NFL win probability calculator says OSU was 79% to win if Crable doesn't make that hit, and 91% after the call. Fan brain says Michigan would have drove for the victory, beaten essentially the same Florida team they played the following year, and cured cancer.
Trend Lines. If you're into seeing how the rest of the Big Ten progressed on offense as this season did, dnak followed up last week's Michigan chart with some for the rest of the field. MSU is a young offense slowly growing up, Michigan's is one coached by insane people.
Hypothesis: UConn was just a bad game and the coaches over-responded to it, putting themselves behind the 8-ball the rest of the season. I submit as a different model Penn State, which had more than a few personnel shortages but big talent in places and stuck with their scheme all year, seeing noticeable progression but no spikes until the last game.
Goals! The Corsi Charts have been shelved for the moment so MGoBlueline can do those goal analysis things I like much better anyway:
Compher wins the faceoff, which is huge. Even more important, however, is that DeBlois is able to tie his man up. This allows a clean tap across from Compher to Guptill…
The OSU weekend's tallies at the link.
Charity. Tomorrow is Adopt-a-Shelter and both sites still could use some volunteers. K.o.k.Law had a tailgate at the house next door to MGoPatio, the cause being to fix up the house of an 11-year-old with a terrible illness.
Weeklies. Turnover Analysis talked about the Countess interception, which had a negative result of half a point. I still would have taken it; I think Furman stood there amazed for a second while OSU's receivers recovered, and if he hadn't he was in the exact right spot to make a key block. Turnover margin has been steadily climbing to the good since UConn, though the offense going into a shell to make that happen might have negated the good that's done. Inside the Box Score. Best and Worst talks about Ben Gedeon. LSA's usual stuff.
[Jump to learn an important lesson about swearing.]
I think I made this same chart in 2008 #1: Gulo Gulo Luscus is joining Off-Tackle Empire and other folk in looking back at recruiting and OL development time to explain why the problems. His method is to look at average age of the starters versus the performance of the unit as measured by YPA/RPI. Cool chart:
If you can't find Michigan it's because we're buried under the Gophers. TSS's response should be part of the read.
Ron Utah has the beginnings of his own UFR database. With Brian's permission I'll make mine available to anyone who asks for it—don't want to host it online but I'm willing to share copies if you say what you're going to do with it. Only concern there is those numbers are very much in context and the n's are quite small so it's easy to abuse. Anyway, Ron's thing tracked performance of different formations in the UConn game. Tiny sample sizes are going to hide most good info but we can whittle that down personnel instead of formation, since these two:
…have more in common with each other than either has with this:
. By personnel:
|Type||Runs||Rush YPA||Passes||Pass YPA||Total YPA||Pass%|
|Empty TE (1TE, 4WR)||0||-||1||6.0||6.0||100.0%|
|Big (2RB, 2TE)||4||-1.3||1||12.0||1.4||20.0%|
|Goal Line (2RB, 3TE*)||3||6.7||0||-||6.7||0.0%|
* Michigan brings in an OT (Magnuson) instead of a third tight end
Still too small a sample size . I'll check in the next bye week and see what these look like when we're not just zooming in on Gardner's terrible awful no good very bad day. The ratios look normal but given the relative paucity of plays with a fullback (and many of those probably in short situations) it's hard to judge those at all.
Goal line—6.7 YPA—we must be manballius!—looks great because of two Gardner boot-to-runs (it would be epically higher if he'd seen Funchess wide open downfield on one of them), but the other play from that formation was a near-crippling fumble-defensive TD.
On the other hand: 12 rushes for 8.3 YPA from three-wide is not a red herring: that is Gardner's rushing ability and the fact that Michigan is far better at getting linebackers out of the box by making them split out with a slot receiver than having a tight end forcibly remove him. Note when it counted on that late 4th and 2 they went with a 3-wide formation and let Gardner pick his spot.
Our Demographic. I made this to explain our numbers versus other media outlets to a potential advertiser:
She's probably on the phone with WTKA now.
There was a bit of trivia going around basketball circles last week that teams named after birds all won their first round games. That's true, but including Kansas makes that a bit of a factoid: a Jayhawk ain't no kind a' bird!
The veracity of this internet knowledge is unverified, but the term comes from just before the Civil War and one of many incredibly stupid compromises they tried to come up with over the great uncompromisable thing. When they couldn't decide whether Kansas would be entered as a free state or a slave state, Congress decided to leave that answer up to whoever could get more settlers in there (caveat: settlers must be alive enough to vote). In the most violent fan poll until the invention of the internet, people from both sides, but mostly northern anti-slavery folk (cause there was more of us) poured into Kansas.
The pro-slavery people, rather than graciously accept defeat, formed up bandit brigades along the Missouri border, hence the term "Border Ruffians," and tried to harass, rob, and murder their way to an electoral victory. In response, the new homesteaders formed up their own guerrilla groups. Possibly naming themselves after popular Revolutionary John Jay, these makeshift counter-terrorism bands became known as Jay-hawkers. They were celebrated across the freedom conference footprint, but quickly became just bandits for the other team instead of protectors.
Just as "Wolverines" became popularly associated with the hardy, stubborn, fierce Michigander regiments who took that as their mascot in the Civil War, the Kansas Union regulars popularized "Jayhawks" but real jayhawkers continued guerrilla attacks on rebel camps and Confederate homesteaders. The opposite were the "Bushwhackers," former border ruffians who robbed, burned and murdered under the rebel cause. And if you were an indiscriminate band of lawlessness in Kansas, you were a "Red Leg."
So what I'm saying is if there is some sort of ornithological protection spell over the 2013 NCAA Tournament, Kansas isn't under it.
How it works:
- I put up a winnable prize that consists of a desirable good.
- You guess the final scores of the designated game, and put it in the comments, preferably in the format of [M's Score]-[Opponent's Score]. First person to post a particular score has it.
- If you guess either game correctly, we contact you. If not, go to (5)
- The desirable good arrives at the address you give us.
- Non-winners can acquire the same desirable good by trading currency for it.
- Seriously, you don't have to actually guess a basketball score to get this shirt. You can buy it.
About Last Time:
How is it there's no footage on the internet of a Wolverine devouring a rabbit? Get on this David Attenborough. The Glove guessed 74-55, which is as close as anyone got to the 71-56 final score of the SDSU (NTSDSU nor TSDSU) game.
This Week's Game:
#4 seed Michigan versus #1 seed Kansas for a trip to the Elite 8.
And the Prize:
Stick around. We may learn something.
Fine print: One entry per user. First user to choose a set of scores wins, determined by the timestamp of your entry (make it easy on me and write your score in digits with a hyphen between them. Deadline for entries is sometime within 24 hours before the start of the game—whenever I can get online in that time and lock the thread. MGoEmployees and Moderators exempt from winning because you can change scores. We did not invent the algorithm. The algorithm consistently finds Jesus. The algorithm is banned in China. The algorithm is from Jersey. The algorithm is not just a shooter. The algorithm always fouls Cody Zeller. The algorithm can’t explain why Big Ten officials think it’s their duty to help Bo Ryan. The algorithm spent 10 years as the Indiana of basketball, if that makes sense. This is not the algorithm. This is close.