Wednesday, December 15, 2004
Stat of the Day, Jess Settles Edition
To paraphrase another blogger's recent paraphrase of an idea from Moneyball - a useful statistic is one that helps clarify things we think we know, point out things we didn't know or were wrong about, and provide context for debates. Today's idea started under the first category, as I wanted to clarify a conjecture of mine. (By the way, I recommend this interview with Moneyball author Michael Lewis, especially if you're into investing. It's a five-parter, so you have to scroll down to the bottom and click on Part I to start at the beginning.)
I was looking ahead to next year thinking that Michigan State, Illinois, and Wisconsin each have rosters that feature seniors or players who could turn pro, or both. Meanwhile, Iowa plays mostly juniors and should have every significant contributor back next year, and I suspect they'll be at least as competitive as they are this year. I have no reasonable means of quantifying a player's likelihood to leave early for the NBA, but measuring a team's collective age doesn't seem too difficult.
The simplest way to measure age would probably be to assign weights to each class (e.g., 1 = frosh, 2 = soph, etc), and average for the number of players on the roster. That's a little too simple since it doesn't account for playing time, meaning a guy playing 5 mpg would factor in just as much as a guy playing 35 mpg.
Here's the approach I took. I assigned each class the following numerical value - frosh = 1, soph = 2, junior = 3, senior = 4. Easy enough. Then I multiplied each player's total points scored by his class value. Call the new number weighted points (WP). I divided the sum of a team's WPs by the team's total points to arrive at an index that approximates the relative age of a team, based on the scoring contributions from each class. A score of 1.00 means a team gets all its points from freshmen, while 4.00 signifies all scoring from seniors.
Here are the calculations for Iowa to get you started.
Jeff Horner, Jr, 3 x 145 pts = 435 WP
Pierre Pierce, Jr, 3 x 143 =429
Adam Haluska, So, 2 x 120 = 240
Greg Brunner, Jr, 3 x 120 = 360
Erek Hansen, Jr, 3 x 74 = 222
Doug Thomas, Jr, 3 x 43 =129
Mike Henderson, So, 2 x 42 =84
Carlton Reed, Fr, 1 x 34 = 34
J.R. Angle, Fr, 1 x 8 = 8
Alex Thompson, Fr, 1 x 8 = 8
Jack Brownlee, Sr, 4 x 3 = 12
Justin Wieck, Jr, 3 x 2 = 6
Seth Gorney, Fr, 1 x 1 = 1
(Total WP) / (Total Points) = 1968 / 743 = 2.65
Sorry for the sloppiness, but I think you get the idea.
Since this stat purports to measure "age," I've chosen Jess Settles as its figurehead. Yes, Father Time himself, the former sixth-year senior who was harrassed with chants of "GRAND-PA SET-TLES!" in a game at Michigan State.
Jess Settles Sidenote - He definitely fit the classic "Iowa's Favorite Son" mold, didn't he? By that I mean a former farm/small town kid turned successful athlete - guys like Robert Gallery and Tim Dwight. (OK, Dwight is neither farm boy nor small town Iowa, but Iowa City classifies as "small town" for anyone who's spent time outside of Iowa, and he's the first guy I think of for most popular Iowa athletes of recent years. Never knew he has a yoga studio in town.) Just curious - Jared Homan, dairy farmer turned NBA prospect, meets the criteria, so why doesn't he get any affection?
Without further ado, here is the Big Ten, as arranged from oldest to youngest according to the Jess Settles Index (JSI).
- Wow, Penn State is playing some pups out there. Their list of 20+ mpg players includes 3 freshmen, 2 sophomores, and 2 juniors.
- 4 of Wisconsin's top 5 scorers are seniors, but their rating drops a little since most of their other players are freshmen and sophomores.
- Michigan State has 3 seniors playing at least 22 mpg.
- Illinois's top five is comprised of 2 seniors and three juniors. An NBA defection or two could really improve Iowa's prospects next year.
- Michigan has 11 guys who average at least 10 mpg (there would be fewer without the pile of injuries), and none are seniors.
- I also measured age by weighted minutes played, but the results weren't significantly different.
Of course it's too early in the season for this stat to be very conclusive, as many teams are playing deep into their lineups while they feast on cupcake opponents. Team points will be concentrated among fewer players as we get further into the season. I'll be updating the Jess Settles Index in a post toward the end of the year (if I remember, of course).
Also, I'm kind of curious to see if there's any correlation between team age and performance statistics, such as winning percentage, offensive/defensive efficiency, shooting percentages, etc. I've put up with the maxim of "you need veteran leadership to win in the tournament" without any skepticism for far too long. It's about time I actually looked into it. I've been working on database to compile numbers for teams and players nationwide for a while, and a productive winter break might give me a chance to explore this idea as well as several others.