Wednesday, July 29, 2009

Dove Entrails, Tea Leaves and Returning Minutes

Looking Back to Look Forward
About this time last summer I posted a couple an entry listing returning minutes for each Big East team. A day or so later I broke those returning minutes down by experience level (roughly 1, 2 or 3 years of DI ball) in a post that speculated about the potential impact of differing experience levels on the corresponding team's ranking within the conference. Rob Lowe over at Cracked Sidewalks worked the "experience levels" along with 4 other factors into a series of posts that looked at teams most likely (according to his calculations) to finish in the conference's lower division and in the upper division. To Rob's credit he called the top 2 and the last place teams on the numbers. He also placed 3/4 of the bottom quartile (seeds #13-#16), 1/2 of the top quartile (seeds #1-#4) and 1/4 of the 3rd quartile (seeds #9-#12). I envy his accuracy and look forward his projections this preseason. Rob has developed a good model, his next step (what I look forward to this next month or so), is the refinements he will no doubt employ as he looks at the teams going into this season.

I was not especially comfortable with my "experience-level" refinement. A too-specific sampling can skew results in (random sampling) polling data. Overvaluing experience from seniors (or juniors) could pose a problem. I decided to "load the shotgun" and see how 2009 winning percentages (overall winning and Big East regular season...) correlated to the returning percentages of sixtenn different statistical categories. It was an interesting exercise. Don't expect a "greyCat's Predictive Model Based on Defensive Rebounding" post anytime soon. Nine of the more interesting categories...

Notre Dame47.055.719.753.261.655.552.6
Seton Hall75.883.983.684.073.583.072.7
St. John's93.496.797.997.194.492.595.1
South Florida70.969.867.571.663.373.579.8
West Virginia65.966.741.663.581.259.788.7

I have listed the table alphabetically, giving no bias for teams with higher percentages of returning minutes (Min%), points (%Pts) or rebounds (%Reb), nor against those with a smaller percentage of turnovers (%TO) nor blocks (%Blk). Every team however does not value each category equally. Notre Dame's Mike Brey has to be concerned that, given the large role the 3 point attempt assumes in his offense, that the Irish will only return 19.7% of their 3FGMs from 2009, even though it was admittedly, a down year for Notre Dame's 3 point shooting. Looking over the percentages while anticipating the state of individual teams, does raise a few questions. Cincinnati for example, struggled to score consistently last season, but the Bearcats do return 77.5% of their points. Stud recruit Lance Stephenson will not come into an unsettled offense system and be expected to fill a point vacuum created by the departure of a Bearcat scorer; he will step into a fairly well defined system, one with an existing array of scorers. He will not have to carry this team, but hit enough shots to make his teammates better. He can use his shooting to help set up his teammates. And they can score when the opportunity presents.

Returning Minutes -- the Outliers
If returning minutes cannot accurately predict where a team will finish relative to its conference mates, returning minutes, especially those "at the edges", should be able to tell us something about how a team will do relative to it's previous record. Rob suggested something of this in his posts last summer. Maybe it's less about how Team X will do relative to Team Y, and more about how Team X will do this season, relative to how it did last season. While more about that in another post, those teams that are returning a relatively large (or relatively small) percentage of their minutes have drawn my attention...with a few comments. St. John's should return at least 90% of their minutes (among other categories -- see green highlight in table above). In two previous instances (Connecticut in 2008 & Villanova in 2009) where the team returned at least 90% of their minutes resulted in an increase in the number of wins (in the conference between 4 and 6, and OOC). Good news for Johnnies fans no doubt, but with a grain of salt -- Norm Robert's 2009 squad returned 20% more minutes over his 2008 squad, but improved their conference wins by only 1 (5 to 6).

The Cincinnati and Connecticut squads of 2007 returned a small fraction of their minutes from the previous season -- 20.5% and 32.1% respectively -- and their 2007 records reflected the shift. The Bearcats notched 6 fewer wins (dropping from #8 to #16) while the Huskies won 8 fewer games (14 down to 6), dropping from a #1 seed in the NCAA (and BET) to failing to make the (non-BET) post season in 2007. Connecticut, Marquette and Providence will return < 40% of their minutes next season (red highlight in table above), the lowest for Big East teams since 2007, and low enough to surely cause their fans some concern going into the season.

Observations and Questions
1. If Connecticut fans are not especially concerned about the lack of returning 3s (Coach Calhoun's offense never emphasized 3s, the Huskies' point distribution typically has fewer than 27% coming from 3FGAs), the lack of returning rebounds (orange highlight above, 38.1%) -- especially defensive rebounds (not shown 36.8%) -- and blocks (orange highlight above, 26.4%), have to be troubling as Coach Calhoun's teams are defined by their defense, especially rebounding and blocks. Charles Okwandu, along with incoming Ater Majok and freshman recruit Alex Oriakhi will have very big roles to fill without the benefit of an "orientation" season.
2. For the second straight season Georgetown returns a low percentage of the previous season's 3 point conversion, a troubling development for a team that relies on perimeter shooting via the Princeton Offense. Identifying and developing a pair of consistent outside offensive threats to replace the now 2 seasons departed Jon Wallace and recently departed Jessie Sapp and DaJuan Summers will, most likely, be a priority. Omar Wattad transferred out while redshirt sophomore Nikita Mescheriakov was disappointingly inaccurate from the 3 point line last season. The staff most likely will look to sophomore Jason Clark to step up, while giving incoming freshmen Vee Sanford and Hollis Thompson long looks.
3. Though generally considered undersized, Villanova has consistently held it's own on the boards. The trio of Dante Cunningham, Shane Clark and Dwayne Anderson will be missed, especially on the offensive glass (not shown, 38.4% returning). Junior Antonio Pena, an especially adept rebounder on both the offensive and defensive boards, will return, along with redshirt freshman Maurice Sutton, transfer Taylor King and the incoming duo, bf/c Mouphtaou Yarou and pf Isaiah Armwood. Like Connecticut, the offense begins with the defense. Forecasters who see Pena on the bench early drastically undervalue the junior's rebounding contributions. But Pena and whatever combination of the others gets to the floor will have some very large rebounding stats to cover.


Rob Lowe said...

greyCat, you are too kind! I keep meaning to start up on this year's predictions (and to figure out how to improve on what was done last year). Thanks for the prod. I sure did miss the mark on Nova's prediction.

By the way, your WUG coverage was great.

Unknown said...

Darn good stuff greyCat. And now I also can't wait to see Rob's predictions.