Otherworldly Mike Scott performances and a suffocating Pack Line defense led the 2011 basketball team to it's first NCAA Tournament trip in five seasons.1 Fans hung on every baseline jumper, embraced the occasionally grueling pace, and sweated out some untimely injuries down the stretch.2 The season was a significant step in the right direction despite ending with a deflating loss to an under-seeded Florida team.
Excitement is building quickly for the continued progress of the Bennett regime.3 However, the team will face it's fair share of obstacles in an attempt to sustain last year's success. The aforementioned undisputed core of the offense is now in the NBA, aged guard Sammy Zeglinski is pulling down double-digit rebounds in Iceland, and the best defensive center is ... out of eligibility. Counteracting the attrition is the 20th ranked,4 five member recruiting class that will be forced into early significant minutes. Virginia fans face difficult questions. Should we temper our expectations for this season? Is a return trip to the Tournament even remotely within reason?
To predict the outcome of the 2012 season, and answer those burning questions, I devised: The BENNETT5 System.
[The BENNETT System]
BENNETT is a method for predicting the final regular season record based on past Bennett-led teams. The thought is, with a relatively consistent pace and strategy, using the pool of Washington State and Virginia basketball seasons under Tony will establish much stronger trends than, for example, just using historic Virginia season-to-season performances. Below is a chart of points scored, points surrendered, record, and strength of schedule rank6 for each of Bennett's six head coaching seasons.
BENNETT splits offense and defense, using past trends to predict points scored and points surrendered per game in 2012. The difference in point margin per game over the course of the season can then be used to predict a final record.
Before we launch into explanations and calculations, I wanted to outline two particular concerns with this system. First, the sample size of six seasons is not exactly robust; trends definitely exist, but their predictive value can certainly be questioned. Second, at Washington State Bennett had been some level of assistant for the three years prior, and didn't need to undo any foot-stomping or death stares inflicted by a previous coaching regime.7 Significantly less upheaval in his first season of coaching the Cougars could alter the magnitude of season-to-season improvement trends.
Despite some inevitable reasons to doubt the pinpoint accuracy of BENNETT, we'll valiantly press on with the final record prediction. The remainder of Part 1 will discuss the method of predicting points scored per game. Part 2 will outline the prediction for points surrendered per game, ending with a final season record.
[BENNETT Step 1 - Offense]
Scoring points remains critically important despite our repeated refusal to bend to the pacism8 rampant in college basketball. The departures of five-year players Scott and Zeglinski highlight a critical variable in the success of college basketball teams. Teams with the most offensive success intuitively feature the highest percentage of returning players. But let's take this one step further, tracing offensive success as a function of the percentage of returning points scored.9 The chart below shows the percentage of returning points scored versus the overall points scored per game for each season of Bennett's head coaching career.10
For each the Washington State and Virginia sets of seasons, outside of Bennett's first season with the Cougars, the points scored per game increase with each increase in percentage of returning points scored.11 The relationship here seems quite strong. I then plotted these numbers and fit a curve to the points to create an equation relating these statistics.
Virginia returns players scoring 51.1% of their 2011 points. Plugging this number into the graph equation gives 60.8 points scored per game in 2012.
This low figure doesn't bode particularly well for a strong final season record. Check back later for the defensive analysis and the thrilling conclusion of the season preview.
1 For the sake of consistency, I'll refer to seasons by the year in which they began. So this coming season is 2012. Really didn't want to type each season as "2012-2013."
2 Some that appear to still be a problem
3 Especially since I often find that football success and basketball anticipation are inversely related
4 Per ESPN
5 Basing Expected Net Non-losses on Experiential Tony Triumphs
6 from Ken Pomeroy
7 *cough* Leitao *cough*
8 Pace in 2011: 61.3 possessions per game, good for 334th in the nation
10 I was going to factor in Strength of Schedule, but a few issues convinced me not to: (1) all of the SOSs were relatively close, and thus wouldn't dramatically affect the final point figures, (2) the minimal effect that they would have is within the error of the BENNETT system as a whole, so it wouldn't really be affecting the accuracy of the predictions, (3) the preseason strength of schedule would likely be poorly representative of the end-season SOS, so it would really just be skewing the overall system
11 Aside: although Bennett has undoubtedly done a great job position the team for success, the firing of Leitao was some odd timing. Some rankings felt that the Hoos schedule in his last season of coaching was top five in difficulty (strength of schedule). As seen in the chart, he also would have been returning nearly 80% of his offensive output in the season following his firing. Seems like the normal ups and downs of coaching to me. Other factors, such as Leitao's renowned temper, were at work; and maybe the Athletic Department has their eyes on Bennett for a while.