Tuesday, August 23, 2011

Dove Entrails, Tea Leaves, Returning Minutes & Scoring...and Pythagoreas

Projectioning the Future by Looking Back?
The Big East conference rosters are still shaking out, but most questions at this point concern entering freshmen, not returning players. The Pythagorean Winning Percentage was employed as an analytic tool by a number of bloggers in the runnup to the 2010 and 2011 seasons with varying degrees of success. The Vegas Watch guys (the previous linked article) kind of winged it after running the numbers, I have used a slightly different approach. I pulled the Pythagorean Winning Percentage (PWP) as computed by Ken Pomeroy, computing the conference-wide average for PWP, and from other sources I pulled the 2011 stats for all players. I then computed averages for (all tracked player stats, among them...) returning minutes & points scored, and then mapped the teams into one of four quadrants based on whether they had an above/below average PWP (overall PWP for the season) and above/below average returning minutes. I performed the same for PWP and returning scoring.

The method was a useful, but hardly a perfect predictor for the 2010 and 2011 seasons. For the curious, click here to see the 2010 chart. [Note -- that chart was current through early August of 2009. After I posted the analysis, Rutgers dismissed Cory Chandler, and removing his 2009 contributions from Rutgers' list of returning minutes remapped the Scarlet Knights into the red quadrant. West Virginia reinstated suspended guards Joe Mazzulla and Truck Bryant before the beginning of Fall Practice. The addition of the two guards' 2009 contributions repositioned West Virginia farther into "the above average" quadrant, reinforcing perceptions that the Mountaineers would be one of the co-favorites to win the conference in 2010.] and here to see the 2011 chart (a quiet preseason for conference roster moves).

Thoughts on the 2010 & 2011 Charts
Of the three squads which mapped into the "above average" quadrant (Above Average PWP, Ret Mins) on the 2010 chart, two -- Georgetown and West Virginia -- logged better 2010 conference records than 2009, posting a net gain of 6 wins. Villanova, the third "green" team, posted the same record record (13-5) and finished #4 again. One team in the "below average" quadrant (Providence) took a giant step back in 2010 (4-14) from their 2009 (10-8) record, while the other, Rutgers, earned two more wins and moved up one spot (#15 to #14) in the conference standings even as the Scarlet Knights sidestepped a crashing Providence.

The quadrants going into the 2011 season showed Pittsburgh, Georgetown and Villanova in the "above average" quadrant. Pitt bettered their already high standings by going 15-3 (2 wins better than 2010). The Hoyas pushed, posting another 10-8 conference record (and an #8 seed in the conference tournament), while the Wildcats took a large step back, posting a -4 conference wins over their 2010 record. All three teams earned NCAA bids in a post season that saw a record 11 Big East teams earn bids. Villanova joined with Notre Dame as the only schools from the "above average" who failed to equal their previous season's performance.

Of the three "below average" quadrant (less than average PWP and less than average returning minutes) teams, South Florida crashed as suggested, posting -6 conference wins from 2010 to 2011. Rutgers replaced their coach, and made no progress (granted it was a more difficult go in conference play), earning an identical 5-13 record and #13 seed, while Providence also held steady with a 4-14 record and #14 finish. The Friars fired their coach at the end of the season.

Among the programs with winning PWPs, but below average returning minutes (there were six teams), Syracuse, West Virginia and Marquette gave back wins, between them seven total, while Notre Dame, Connecticut and Louisville gained wins -- seven actually (no attempt here to suggest a correlation between the two groups) -- with the Irish producing the biggest gain with 4 wins, but the Huskies won the war, taking the National Championship even though they finished #9 in the conference. Of the four teams that plotted into the "low PWP/high Ret. Minutes" quadrant, St. John's fired their coach and improved dramatically (+6 wins), DePaul fired their coach and had a push (one win each of the 2010 & 2011 seasons). Seton Hall fired their coach and lost ground (-2 wins over 2010), while Cincinnati did not fire their coach and picked up 4 wins (chemistry and maturity perhaps?).

The conference average for returning minutes is down from last season (56% from 62%), as is the PWP (82% down to 66% -- additional evidence that the conference was more challenging last season?); If the PWP drop is predictive, the 2012 edition of the Big East will be a bit easier than the 2011 edition, whether the 2012 mid/end of season aerials will show a different, a greater distribution of teams to the "high or low" quadrants...take with a grain of salt. Out of curiosity, I decided to plot the teams using Pythagorean Winning Percentage (PWP -- overall again) and scoring, that is, percentage of returning points. Of the 16 teams plotted, only one (Connecticut) mapped to a different quadrant..."The Value of Kemba Walker"?

Hints About 2012
If the trend for teams in the upper right quadrant (Villanova and Notre Dame excepted) is an improved (or at worst, the same) record over the previous season, the six teams (a very large number for that quadrant, by the way) are in for a collective good year. Indeed, that Pitt, Louisville, Cincinnati and Syracuse are on track for good seasons does not contradict the consensus of preseason projections so far. Marquette and UConn are surprises...for only a moment. Buzz Williams has been consistently under valued by the press and fans from schools other than Marquette. I expect the Warriors will earn another NCAA bid next March, marking every year in Williams' tenure that they have danced. Despite some early off season waffling from Coach Jim Calhoun, the coach with the conference's second longest tenure will be back for his 26th season. If last season's run to the National Championship rekindled the coach's enthusiasm, the recent retirement of Athletic Director Jeff Hathaway did nothing to dampen it. The loss of Kemba Walker noted, Jeremy Lamb's performance in the U19 World Championship Tournament in July should give UConn fans reason to believe their team will have scorers to step in.

Rutgers and Seton Hall have mapped to the lower left (below average) quadrant. If the trend (Rutgers in 2010 excepted) holds, those two will continue to struggle in the conference's lower division (if not the bottom quartile) for another season. What to make of St. John's and Georgetown? Villanova's season-ending swoons in 2010 and 2011 caught most of the media (and fan) attention. Few beyond the Hoya Faithful have noticed that Georgetown suffered through a 5-6 ending in 2010 that included a first round loss in the NCAA Tournament, and that 2011's end was an even more disheartening 1-6 skid that included a first round NCAA exit for the second consecutive post season. The Johnnies' remarkable Big East season cloud the memories of the struggles in Coach Lavin's earliest days as the Red Storm mentor. The veteran coach had an experienced and very hungry group of seniors to work with, a luxury he won't have in 2012. Late 2011 post game pressers touched repeatedly on prospects in the year after Dwight Hardy and DJ Kennedy, and even the coach acknowleged 2012 would be one long learning curve. Mid/late conference season may be an interesting time for Red Storm fans, but the early parts of 2012 will probably have a few hard-to-watch affairs.

Programs in the "Above/Below" Quadrants
How does Notre Dame, which has mapped into the below average for returning minutes continue to maintain an above average PWP? Coach Mike Brey seems to have a constant stream of transfers to mix with red shirts and true freshmen nearly every season. Ben Hansbrough, Scott Martin and (eligible in 2013) Garrick Sherman out of Michigan State. While a few schools have been adverse to accepting transfers (not to mention JUCOs), having an experienced player practice for two semesters with the squad (even though they have less eligibility than a true freshman) before taking the court can give the squad a strong replacement with no noticible on-the-court learning curve. Although the 12 programs mapped into the lower right quadrant (below average returning minutes, above average PWP) in the 2010 and 2011 seasons have fared well in conference play, which may bring comfort to members of the Nova Nation and fans of the Mountaineers. Both must replace minutes and scoring from classes graduated in 2011. While West Virginia can count on rising seniors and a steady supply of transfers and JUCOs to fill out the squad. Villanova however, has a squad heavy with juniors and freshmen, and on the heels of last season's late season struggles, may find it harder to maintain a winning program in 2012.

The upper left quadrant has been something of a coach's coffin corner. Of the six programs that mapped to that quadrant in 2010 and/or 2011, four were about to replace, or had just replaced their head coaches. Cincinnati's Mick Cronin dodged that bullet, but will South Florida's Stan Heath escape a second time (the Bulls were there in 2010)? Providence has mapped into the quadrant in 2012, but the Friars replaced Keno Davis in the 2011 off season, which means five of the seven programs to map there have replaced their coaches since 2009.

Things We Know We Don't Know
Villanova and Georgetown's experiences in 2010 and 2011 suggest that chemistry may be undervalued when looking at a stats-based projection. Fans overvalue the impact of incoming freshmen, and the media does too, but with virtually no statistical history with the team, freshmen are ignored in these charts -- immunizing these charts from the hype. Yes there are impact freshmen (Carmelo Anthony, Luke Harangody, DeJuan Blair), but enough to affect more than one or two conference teams in a season? I feel largely the same about JUCOs, though they do bring more collegiate-level experience with them, they tend to play support, rather than impact roles, for any destination team. Maybe this season, with St. John's having immediate need roles for Nurideen Lindsey and God’s Gift Achiuwa (having lost over 95% of their minutes from 2011...), those two may have an impact approaching the level fans anticipate annually in other programs. Properly assessing the value of transfers however, has proven to be the bigger challenge. Unlike the true freshmen and JUCOs -- most recognize (on some level) they have a learning curve to overcome -- the impact of red shirt freshmen and transfers are unrepresented by charts like the ones used here over the past three off seasons. Those players often find their way into the regular rotation early, because the coaching staff has had a year to observe and work with the player. Coaches develop confidence in the player which is not measured/reflected in a statistical database. Teams this season that will put potential impact transfers out on the floor include South Florida (Victor Rudd) and Marquette (Jamil Wilson -- Buzz Williams again).

No comments: