All of charlespig's recent recruiting number-crunching got me inspired to do something I've been wanting to do for awhile--make a graph charting the historical in-conference performance of all the Big Ten men's basketball teams.
The graph below displays a 5-year central moving average of each team's winning percentage in regular season conference games (no BTT games included). What that means is that the winning percentage displayed for any particular year on the graph is actually showing the combined winning percentage over the target year, the 2 years before, and the 2 years after. So if 1995 is the target year, line on the graph displays the combined winning percentage for 1993-1997. Obviously for the years after 2011, the data is not complete. The 2012 figure is calculated using 2010-2013, and the 2013 figure is calculated using 2011-2013.
Why use only regular season conference games? Simply because a team's regular season conference schedule is a larger and more consistent sample of games than its pre- or post-season schedules.
Why use a central moving average? In any one season, there are going to be random results that are probably not good indicators of changes in team quality. Take Purdue's 2007-8, 2008-9, and 2009-10 seasons, for example. Over those 3 seasons, the team posted conference records of 15-3, 11-7, and 14-4. Did the team really get worse during the 2008-9 season? Probably not. And the graph shows this nicely. In fact, it suggests that the 2007-8 and 2008-9 seasons were in fact steps on the road from mediocre to great conference performance. The fact that the 15-3 season came before the 11-7 one is likely just accidental. Using a central moving average captures this nicely.
I think of it this way: Including the 2 previous years is a proxy for information about program strength--coaching, reputation, location, facilities, and current roster advantages, for example. Including the 2 subsequent years folds in information about future potential--coaching changes or improved recruiting, for instance. Adding these bits of info smooths out the ups and downs in the graph and probably gives a fairer picture of where a program is at any given point in time.
So without further ado, here's the graph. I know it looks like a mess, but I couldn't think of a good way to simplify it. Just try following a single team's trajectory first, and you'll probably get a feel for it. Most of the colors on the graph match at least one of the team's colors. [Click on it for a larger version.]
1. The graph shows increasing parity in the league. To take the most obvious example, from 1985-1992 Northwestern won a total of 13 conference games. In the past 8 seasons, they've won 43. Wisconsin, who used to be a perennial doormat, has become one of the conference's elite, and Minnesota, though still consistently mediocre, has ticked up slightly from where they started in the late 80s.
2. The graph also clearly shows periods of dominance: IU in the late 80s, Purdue in the mid-90s, MSU in the years around its 2000 championship, Illinois in the years around its 2005 national runner-up finish, and, since then, a neck and neck race between Wisconsin, MSU, and OSU.
3. The graph divides teams into useful groups, I think. And these groups are different than what you would come up with if you just looked at last season's performance:
The Current Elite: Michigan State, Ohio State, Wisconsin
I'm sure no one would dispute the first two. Wisconsin's style of play and lack of postseason success might make some (not me, I should note) question their inclusion here, but going by the metric of conference regular season performance (for reasons I articulated above), there's no way you can leave them out. Over the last decade, they have consistently been among the conference's very best.
Somewhere in the middle: Indiana, Michigan, Illinois, and Purdue
This is where the graph defies "what have you done for me lately" rationality. If you were merely going by the results of the 2012-3 season, you certainly wouldn't group these four teams. IU won the Big Ten and held the #1 ranking for much of the season. UM played in the national championship game. Purdue and Illinois...didn't.
But IU and UM are coming out of long periods of mediocrity (IU's was of course exacerbated by the Sampson misadventure). They have not had the kind of extended success that would put them in the elite group. Purdue and Illinois are both backsliding from periods of great success, though Purdue's was of course more recent than Illinois'. But taking a broader view of their performance leads me to group them together.
UM and IU appear to be on the upswing, yes, but that doesn't always continue. Their ascendancy could fizzle out, like Minnesota's did in the mid-90s and Iowa's did in the mid 2000s. Purdue and Illinois appear to be on a downward slope, but they might recover.
Bottom Feeders: Minnesota, Iowa, Penn State, Northwestern
Again, if you were looking at last year alone, Iowa and Minnesota would certainly not be in this group. But the graph suggests that (over the last 30 seasons, at least) neither program has ever been able to pull itself into the elite range. Of course, that doesn't mean they won't be able to in the future, but it does give reason for skepticism.
In conclusion, this is not meant to be a definitive statement on program strength, but I think it does provide some insight. I dislike thinking that is too responsive to recent events and prefer looking at the big picture if possible. This graph is an attempt to do that. And it's fun to see how the fortunes of programs have waxed and waned.
[One final note. I didn't make IU pink just to be a jerk, but there are simply too many red teams in the conference. Wisconsin got the bright red, Minnesota got their maroonish color, OSU got gray, and since cream wouldn't show up on a graph, I just decided to give IU a red and white mix, hence pink (actually salmon, but whatever). Don't hate me, Hoosiers.]