Predicting Providence's Season

Victories will be hard won for the Friars this season. After another four-win conference campaign and the loss of two of the team's top five players, including All-Big East performer Marshon Brooks, Providence should fall comfortably among the league's bottom four for the third straight season.

I'm not able to produce the basketball statistical equivalent of voodoo economics to dispute the consensus, but I do offer some perspective on the upcoming season supplemented by a few juicy stats.

One statistic I'm going to reference a lot -- and have discussed before -- is Poss/100. It is a simple way to measure how much a team relied on a player. It multiplies the percentage of minutes played by the possessions used by a player (more on possessions). In this feature, I'll try to determine how difficult it is to replace a player with that influence. Brooks, of course, was one of the most instrumental players to his team in the Big East last year, and Poss/100 expresses that.

Big East leaders in Poss/100, All Games (2010-11)

Player Team Min Poss Poss/100
Kemba Walker Connecticut 0.924 0.314 0.290
Marshon Brooks Providence 0.909 0.272 0.247
Dwight Hardy St. John's 0.852 0.256 0.218
Ben Hansbrough Notre Dame 0.871 0.25 0.218
Vincent Council Providence 0.903 0.237 0.214
Corey Fisher Villanova 0.831 0.257 0.214
Maalik Wayns Villanova 0.750 0.277 0.208
Austin Freeman Georgetown 0.841 0.245 0.206

Only Player of the Year Kemba Walker had a higher Poss/100 last season. Some household names, including PC's own Vincent Council, fill out the list. These are the only eight Big East players who accounted for at least 20 percent of their team's production last season, and Council and Maalik Wayns are the only two who return this season.

There have been 10 teams in the 16-team Big East who have had to replace a player as influential as Brooks. It's not surprising that those teams, as a group, got worse, but they didn't see the severe drop-off that one may expect.

Big East highest individual Poss/100 lost (2005-06 to 2009-10)

Player Team Season Poss/100 EM1 EM2 Diff BEW1 BEW2 Diff
Quincy Douby Rutgers 2005-06 0.297 -4.7 -17.7 -13.0 7 3 -4
Dominique Jones South Florida 2009-10 0.282 -6.5 -13.2 -6.7 9 3 -6
James Holmes South Florida 2005-06 0.261 -13.4 -13.6 -0.2 1 3 2
Jerome Dyson Connecticut 2009-10 0.260 -1.4 5.8 7.2 7 9 2
Dar Tucker DePaul 2008-09 0.256 -21.3 -15.9 5.4 0 1 1
Randy Foye Villanova 2005-06 0.254 8.9 6.6 -2.3 14 9 -5
Jerel McNeal Marquette 2008-09 0.245 8.3 4.8 -3.5 12 11 -1
Will Walker DePaul 2009-10 0.241 -15.9 -21.2 -5.3 1 1 0
Greg Monroe Georgetown 2009-10 0.238 8.6 -1.6 -10.2 10 10 0
Lazar Hayward Marquette 2009-10 0.236 4.8 3.0 -1.8 13 9 -4

EM refers to a team's efficiency margin, how many points better or worse than its opponents that a team is per 100 possessions (BEW refers to conference wins). Taken in a large sample, EM is the best indicator of how good a team is. The figures for EM above consider only conference play. This table shows that -- on average -- Big East teams that lose a very influential player tend to be about three points worse in EM and about a win and a half worse in conference record.

The post-Quincy Douby Rutgers team and the post-Greg Monroe Georgetown team had the fourth and fifth largest one-season declines in the 16-team era. Those two plus the post-Dominique Jones USF club account for the entirety of the 30-point EM decline in the table above. The other seven teams basically broke even despite losing a highly influential player.

That table was perhaps intellectually interesting, but it doesn't tell us much about what we can expect from this year's Friars. Those teams were not very similar to last year's Providence team in terms of either how good they were or what else they lost apart from the top player.

Let's take a look, then, at teams that have been about as good -- or bad -- as 2010-11 Providence. Last year's Friars team was the school's worst in the expanded Big East, with a -11.2 EM in conference play. Here's the subset of Big East teams within three EM points of Providence in the last six years.

Big East teams between -8 and -14 in efficiency margin, conference play only (since 2005-06)

Team Year Lost EM1 EM2 Diff BEW1 BEW2 Diff
St. John's 2005-06 0.15 -8.2 -8.7 -0.5 5.0 7.0 2.0
St. John's 2008-09 0.15 -10.8 -3.2 7.5 6.0 6.0 0.0
Cincinnati 2008-09 0.20 -9.7 -4.8 4.9 8.0 7.0 -1.0
Cincinnati 2006-07 0.27 -13.3 -4.5 8.8 2.3 8.0 5.8
Seton Hall 2006-07 0.29 -8.9 -6.7 2.2 4.5 7.0 2.5
South Florida 2008-09 0.33 -12.9 -6.5 6.4 4.0 9.0 5.0
St. John's 2007-08 0.33 -11.0 -10.8 0.3 5.0 6.0 1.0
Rutgers 2008-09 0.44 -13.8 -15.9 -2.0 2.0 5.0 3.0
South Florida 2005-06 0.46 -13.4 -13.6 -0.2 5.0 3.0 -2.0
St. John's 2006-07 0.47 -8.7 -11.0 -2.3 7.9 5.0 -2.9
South Florida 2006-07 0.55 -13.6 -10.9 2.7 3.4 3.0 -0.4
South Florida 2007-08 0.63 -10.9 -12.9 -2.0 3.0 4.0 1.0
Providence 2010-11 0.32 -11.2 ? ? 4 ? ?
South Florida 2010-11 0.21 -13.2 ? ? 3 ? ?

Prior to Providence and USF last season, 12 teams had efficiency margins within three points either way of last year's PC squad. On average, teams that were quite poor though not utterly decrepit (six other teams have had worse EMs than anyone on this list) one year, were -- on average -- 2.1 points in efficiency margin and 1.2 conference wins better the following year regardless of what personnel those teams lost from the previous. Most of that is simply regression to the mean -- it's statistically difficult (though Rutgers, DePaul and USF have tried) to be that bad in consecutive seasons.

That's good news, I suppose, but the potentially better news comes in that new column called "Lost." In simple terms, Lost is the percentage of possessions a team loses from the prior season. The average Lost figure for teams on the table above is .35 (these figures have been normalized for each season). Teams that lost less than 35 percent of its possessions actually faired significantly better than teams that lost more, which makes sense. The seven teams below the average saw an improvement of 4.2 points in efficiency margin and 2.2 conference wins, about double the total of the group as a whole. At .32, Providence would fall into the more favorable range, although just barely.

There are always other factors at work beyond just noting how good a team was one year and what it lost, but this is a simple way to get an idea of where expectations should be. The formula I use to project how a good a team is -- based on its Lost percentage and its two previous EMs -- peg Providence at -7.8 EM and five-to-six conference wins (estimate is 5.6). That's certainly an improvement, albeit a modest one and one that is likely to come primarily from regression to the mean. There is, of course, plenty of variance in my projection, but the chances of Providence finishing with an EM better than -1 (a figure that would put them around the NCAA Tournament bubble) are just 2.5 percent.

In other words, there's a good reason for PC fans to focus on more than just the final score this season.

If you'd like to discuss this story or have any other Friars-centric stats thoughts, you can PM me at UltimateCranston.

Scout Friars Top Stories