/cdn.vox-cdn.com/uploads/chorus_image/image/1506513/20120320_kkt_al6_487.0.jpg)
There are a few areas where a college program should be able to control its own destiny; the goal is that any ranking of a good women's college basketball program be limited to those areas, and nowhere else.
It should put a good product on the court. It should be able to draw fans. It should get benchmark wins. It should do well in the post-season. It should get the most well-known recruits. And there is no way a college program should be able to put a brave face on things - either it succeeds in these areas, or it fails.
The problem with any metric, however, is that once people know that a metric is used to evaluate something the first response is find a way to bend that metric to their own purposes. In the HBO hit series "The Wire" they called it juking the stats. In order to boost the rate of solved crimes, police departments look at unsolved felonies and rewrite them as misdemeanors or get the complainants to drop the charges. In order to boost money to schools, schools round up truants - but only during the days when the state takes attendance. In order to boost test performance scores, schools teach to the test. If it isn't an eponymous law, it ought to be: "If success depends on reaching a benchmark, the social pressure to find loopholes will end up with the benchmark redefined or reduced to meaninglessness."
It would seem that the stats one uses to measure a women's basketball program can't be juked. A win is a win and a loss is a loss. Either you have a good enough record of success to be invited to the post-season, or you don't. But a recent article by Luke Winn at Sports Illustrated discusses how some smart coaches are doing just that.
RPI
When the NCAA Selection Committee gets together to determine which teams get invited to the Big Dance in March, they are given something like a cheat sheet on each team. (A mock-up of cheat sheets can be found on wbbstate.com.) The sheet categorizes all of the team's wins by RPI, or "Ratings Percentage Index". The most basic measure of RPI is as follows:
RPI = (win percentage* 0.25) + (opponents win percentage * 0.50) + (opponent's opponents win percentage * 0.25)
In short, a team is judged based on winning games, playing opponents who also win games, and playing opponents who play other teams who also win games. The triple-weight of RPI should cover any attempt to evade the system by playing patsies.
A digression: the roles on a coaching staff are loosely delineated as follows. There is a head coach who is usually in charge of Xs and Os. There is a recruiting coordinator who is charge of bringing talent into the program. And there is someone whose job it is to schedule games. This is sometimes a lower ranked coach, and the process of finding games takes place in a rather ad hoc manner. There are message boards - I won't tell you where - where coaches post looking for opponents. Good programs sometimes have to offer cash incentives (it's perfectly legal) to those Bluebell Girls Academy schools that are going to come to Big State U and get thumped for two hours. It's a good deal for everyone - everyone has a place to play, lesser programs get rewarded and deals can sometimes be arranged for home-and-away series that fill a hole not just in this year's schedule but the following year's as well.
The best schedulers, however, know how to work the RPI game - they know how to force that mathematical formula to sing sweet music to the NCAA or WNIT selection committee. How do they do it?
1) They figure how many conference wins they're going to need to be taken seriously. That can't be helped. You need to know how good you are now before you can begin tinkering with perception. Connecticut and Baylor don't need any RPI help.
2) Note that this is a three part formula. You might not be able to master all of the parts of the formula, but you might be able to swing two out of three. You obviously need to win games, so you can't give up part one. So let's abandon one of the other two parts.
2a) In this case, we'll give up the second part in favor of the third. A good scheduler will schedule opponents that he thinks will have a lousy year but will have a great strength of schedule. The best teams to schedule are the doormats in power conferences. They will have a high "opponent's opponents win percentage" and you have a good chance of beating them - you might even want to take the chance of playing them at their home.
Obviously, you're not looking to play teams in your own conference. So who were the best bets in enemy conferences to play?
School | Conference | Record | SOS |
UCLA | Pac-12 | 14-16 | 11 |
Illinois | Big Ten | 11-19 | 22 |
Missouri | Big Twelve | 13-18 | 26 |
Boston College | ACC | 7-23 | 34 |
Marquette | Big East | 14-17 | 35 |
Auburn | Southeastern | 13-17 | 40 |
Xavier | Atlantic 10 | 8-20 | 56 |
Northeastern | Colonial | 8-22 | 59 |
Portland | West Coast | 12-20 | 84 |
Indiana State | MVC | 15-16 | 92 |
Western Kentucky | Sun Belt | 9-21 | 98 |
East Carolina | CUSA | 12-19 | 110 |
Buffalo | Mid-American | 4-22 | 116 |
St. Francis (PA) | Northeast | 14-16 | 119 |
Colorado State | Mountain West | 13-17 | 126 |
CS-Bakersfield | Independent | 7-22 | 131 |
Charleston | Southern | 7-23 | 133 |
East Tennessee St. | Atlantic Sun | 8-22 | 137 |
North Dakota State | Summit | 11-20 | 149 |
Hawai'i | WAC | 11-19 | 152 |
Milwaukee | Horizon | 9-21 | 154 |
Dartmouth | Ivy | 6-22 | 167 |
Loyola (MD) | Metro | 13-17 | 168 |
Arkansas-Pine Bluff | SWAC | 1-29 | 174 |
UC-Riverside | Big West | 5-24 | 192 |
Sacramento State | Big Sky | 13-18 | 194 |
MD-Eastern Shore | MEAC | 12-18 | 202 |
Texas-San Antonio | Southland | 8-21 | 215 |
Binghamton | America East | 13-18 | 219 |
Gardner-Webb | Big South | 4-25 | 226 |
Colgate | Patriot | 8-22 | 250 |
Austin Peay | Ohio Valley | 9-22 | 259 |
Texas-Pan American | Great West | 13-17 | 286 |
2b) But even better opponents to schedule are teams that have an inflated RPI. Luke Winn calls it the "exploitable gap". How does a team have an inflated RPI? These are teams that are forced to play a great percentage of their games against weak teams - they are basically the power teams of the weakest conferences. They have high win percentages - and thus excel in the first part of RPI - but against weak competition.
A smart scheduler will look at who is going to excel in a weak conference. (This will be explained by saying that "we played eight teams that went to the postseason last year" but they all came from the Patriot League, MEAC, Northeast, Big South, and Ivy Leagues.) Find teams that can pick up 20 wins in a weak conference, and play those teams. That's an even better boost than 2a) is. You can win the game if you're smart, and your opponents win percentage will be good as well.
3) One thing that you never want to do is play weak teams. These are teams with an RPI at less than 250 - they stink and their opponents stink, too. These are the empty calories of RPI, and the NCAA selection committee will not look kindly on these junk food games. According to Luke Winn, a college team is far better off scheduling home games against non-Division I teams than against teams ranked at below 250 in RPI - the reason being is that only Division I teams are counted in RPI and losing to a good DII team won't kill your RPI.
Let's take a look at the bottom ten power conference teams in RPI last season. Did they follow Rule #3, or not? The NCAA Selection Committee will be looking at non-conference games, since a Big East team can't control for having to play Pittsburgh twice a year.
Final RPI | 1-50 NC | 251+ NC | NCAA seed | |||
Clemson | 245 | 1 | 3 | none | ||
Virginia Tech | 223 | 2 | 1 | none | ||
Indiana | 217 | 2 | 3 | none | ||
Boston College | 203 | 3 | 1 | none | ||
Pittsburgh | 197 | 1 | 2 | none | ||
Mississippi | 190 | 2 | 7 | none | ||
Seton Hall | 181 | 2 | 2 | none | ||
Alabama | 175 | 1 | 5 | none | ||
Wisconsin | 165 | 2 | 1 | none | ||
Arizona | 162 | 1 | 4 | none |
Note that in the majority of cases, these low power conference teams made the mistake of scheduling too many weak sisters. Ole Miss scheduled *seven* non-conference opponents with an RPI of below 250. Even if they had done well in 2011-12, that schedule would have doomed them in the eyes of the NCAA selection committee.
Now go over to the other side of the table and look how the big girls do it:
Final RPI | 1-50 NC | 251+ NC | NCAA seed | |||
Baylor | 1 | 4 | 2 | 1 | ||
Stanford | 2 | 5 | 1 | 1 | ||
Connecticut | 3 | 6 | 1 | 1 | ||
Notre Dame | 4 | 8 | 2 | 1 | ||
Tennessee | 5 | 8 | 0 | 2 | ||
Maryland | 6 | 3 | 3 | 2 | ||
Texas A&M | 8 | 6 | 0 | 3 | ||
Kentucky | 9 | 4 | 3 | 2 | ||
Duke | 10 | 6 | 2 | 2 | ||
St. John's | 11 | 5 | 1 | 3 |
The best teams avoid scheduling any weak teams. There is nothing to prove for Tennessee scheduling the likes of NJIT. Tennessee and Texas A&M had going for them the fact that they didn't even play a team with an RPI below 200 last year.
Can we figure out who is best at scheduling?
Theoretically, we could run the numbers - if we had a computer programmed with the right data - and see if program out there knows how to beat the system.
For example, take step 2b) above - playing against good teams from bad conferences. I chose the teams from the SWAC, Ohio Valley, Atlantic Sun, Patriot, MEAC, Big South and Independents that all had winning records. The teams are listed with their record and final RPI.
Howard (24-9, 55)
Florida Gulf Coast (29-3, 60)
Hampton (26-5, 63)
Liberty (24-9, 95)
Tennessee-Martin (23-9, 104)
Eastern Illinois (22-9, 121)
Stetson (23-11, 123)
Coppin State (20-12, 132)
American (23-8, 138)
Florida A&M (22-8, 146)
Mississippi Valley State (15-13, 150)
High Point (20-13, 153)
Holy Cross (19-15, 159)
Seattle (20-12, 160)
SIU-Edwardsville (18-12, 166)
Winthrop (18-13, 176)
Lehigh (17-13, 186)
Navy (18-14, 188)
Campbell (18-12, 189)
Alabama A&M (17-12, 196)
There are some good teams here - Florida Gulf Coast went to the NCAA Tournament - but undoubtedly all of them suffer some degree of RPI inflation. Were there any teams that scheduled these opponents regularly?
There are two teams that scheduled at least three of these opponents. One team is from a power conference and finished with a 23-10 record last year. Its final rank in RPI was #29, but it finished in seventh place in its conference.
The other team finished with an RPI of #46. However, it came from what we would call a high mid-major and finished in fourth place, which is usually where high mid-majors get overlooked for the post-season. This team made the mistake of scheduling three sub-250 RPI non-conference opponents, just like the one above it.
The first team played Tennessee-Martin, High Point, and Alabama A&M all last year.
The second team played Florida Gulf Coast, Liberty, Lehigh, and Navy.
Give up? The first team was Vanderbilt, which made it to a #7 seed in the NCAA Tournament and got as far as the second round. The second team, despite its #46 RPI, didn't get a nod for the NCAA tournament. That team was Richmond, which was one of several Atlantic Ten teams to get a WNIT bid and made it as far as the second round.
The theory in practice
Men's college basketball has a wealth of statistics that are not available to women's college ball. The reason is that men's ball is a much bigger game in terms of money, and stat crunchers don't see it as cost-effective to devote time to the women's game. Ken Pomeroy puts out his own rankings of college teams which recalculate non-conference strength of schedule and - in the words of the SI article - "which creates a Pythagorean winning percentage based on opponents' adjusted efficiency". Coaches who subscribe to Pomeroy's stat service can see which teams are weaker than their RPIs indicated, and step into the "exploitable gap".
Of course, you'd have to have some luck on your side. With reciprocal agreements, you might be locked into playing certain teams a couple of years from now. If you're St. Joseph's, you'll be required to play all the Big Five Philadelphia teams. And if your forecasts of the winners of the lower conferences are wrong, you could set yourself up for a lot of pain.
Even so, scheduling might be more of a science than just calling someone and saying "wanna play"? When you're a college basketball program, your survival depends on exploiting every available opportunity. A good schedule might mean the difference between getting an NCAA bid or not, or between getting a sweet #4 seed as opposed to a more dangerous #5 seed. As every college head coach knows, even the little things mean a lot.
#26: St. John's: Joe Tartamella takes over in Queens for Kim Barnes Arico, his very first head coaching job. Previous HC took Red Storm to first-ever Sweet Sixteen, a tough act to follow. Sandra Udobi-Ofogu, signed by Arico, will be there this season along with four other starters.
#27: Gonzaga: The best mid-major program in the United States. Kelly Graves loses Courtney Vandersloot, #3 pick in WNBA Draft, might find it hard to return to national polls. Zags are 42-2 in West Coast Conference over last three years. Three straight years of at least the Sweet Sixteen - the team that no one wants to play.
#28: Iowa: Hawkeyes make it to NCAA but fall short of a 20-win season and get a first round exit. Wing Alexa Castanek (class of '13) has already verballed to Iowa, had work done on her knee in June. Bad mojo, two players ended careers early with injuries last year, C Bethany Doolittle arrested in June for public intoxication and providing false information. Iowa women have a new basketball website.
#29: Florida State: Sue Semrau decided to stay in Tallahassee. A disappointing 14-17 year for the 'Noles. Cierra Bravard is gone and playing in the Liga Feminina in Spain. Even with Semrau in for the long haul, selling anything but football at Florida State is going to be difficult and I don't see her there past April 2015.
#30: Middle Tennessee State: Four consecutive NCAA appearances for Blue Raiders. MTSU gets commitment of G China Dow and picks up 6-6 post JUCO player Jeniece Johnson, and none of their starters graduated in the off-season.
#31: UCLA: First year of Cori Close's tenure ends with disappointing 14-16 mark. California sunshine is still sweet for Bruins, freshman G Nirra Fields scored 20 points in McDonald's All-Star Game and was named Ms. Basketball of California and Parade All-American. Touted freshman G Lauren Holiday's brother Jrue is starting PG for Philadelphia 76ers.
#32: Texas Tech: Lady Raiders started the season 14-0 and finished it 7-14, finishing next to last in the Big Twelve (next to departing Missouri). Tech wins its 800th game in women's basketball, picks up Minta Spears, who scored 33.7 ppg for Bullard High School.
#33: Penn State: Head coach Coquese Washington was asked by Michigan in the off-season but will stay at Penn State. No wonder Washington in demand - finished at #9 in final ESPN Coaches poll, NCAA Sweet Sixteen appearance, Big Ten regular season title. The Lady Lions are roaring again!
#34: Miami (FL): The Hurricanes cracked four digits in attendance, were a Top Five ranked team at one time last year and Shenise Johnson was the #5 pick in the WNBA Draft. As Katie Meyer says "We're relevant", but now Miami has to stay relevant - that will be tough. Recruiting still seems to be working with F Keyona Hayes and C Macy Keen leading the recruiting pack.
#35: DePaul: Head coach Doug Bruno earned an Olympic gold medal in the off-season, took team to NCAA for ninth consecutive year months earlier. Injuries forced Bruno to rely on his "Magnificent Seven"; six of them will be back for 2012-13.
#36: James Madison: Impressive year takes James Madison all the way to the WNIT Finals. Freshman guard Angela Mickens and future G Chania Ray (class of '14) should help the Dukes dominate the CAA for a long time.
#37: Oklahoma State: Cowgirls finished an emotional year - they lost coach Kurt Budke mid-season - by winning the WNIT title. Jim Littell to take OSU into the future, wing freshman Brittney Martin among top twenty recruits nationally according to ESPNU.
#38: Wisconsin-Green Bay: Matt Bollant - who led the Phoenix to a 148-19 record over five seasons - is off to Illinois, former coach Kevin Borseth (216-62 for UWGB) comes back. Four starters come back, Julie Wojta drafted at #18 in WNBA draft.
#39: California: First season under head coach Lindsey Gottlieb ends at 25-10, second round loss in NCAAs. Of eight players who started games last year, seven are back. F Kendall Waters joins the Golden Bears in 2013.
#40: Dayton: The Flyers won their first Atlantic 10 conference tournament and picked up their third straight NCAA appearance four months after losing 78-38 to UConn. Dayton had two players on the A-10 all-Rookie Team, including sophomore guard Andrea Hoover who was second in points per game on the team last season (10.0 ppg).