On the heels of the NCAA releasing its first iteration of the RPI, the Internet computing machine is indicating that we've all officially entered NCAA Tournament projection season. "burnspbesq" took a first crack at an iteration of the field and Terry Foy of Inside Lacrosse published his first tournament projection dispatch yesterday. These are all fine, and I recommend reading each as they take different approaches to defining the landscape, but they both lack what I crave the most: Violent anger directed at the cockamamie way that the NCAA decides to create its lacrosse tournament field, relying heavily on a metric -- the RPI -- that is the equivalent of a 1940's refrigerator running on steam power.
Why is the RPI the equivalent of a dump truck racing in the Daytona 500? Easy: It does an uneven job at actually assigning the appropriate value to things. It's biggest issues are twofold: (1) It doesn't account for context, eschewing margin of victory -- which is vitally important in determining the relative strength of teams -- for blinding simplicity and ignoring how teams actually performed in their games; and (2) When applied as a means to measure strength of schedule, it fails notably, assigning more value to losses against good teams (regardless of the margin of defeat) than to wins against poor teams (regardless of the margin of victory) (in function, the simple act of playing an opponent is often given greater value than whether a team wins or loses that game). How is it possible in 2013 that the NCAA uses a faulty and non-contextual method to populate and organize its lacrosse field? We have, as a society, unwound the genetic sequencing of living things and yet the RPI -- a rubber band finger gun in an age of drone warfare -- is still the tool to which the selection committee must utilize to make decisions. It's incredibly frustrating, and the reliance on the RPI -- in all its forms -- almost makes one's brain want to secede from its skull/prison as it is being continually abused due to a failure to understand why humans would continue to do something impressively underwhelming.
Pertinently, the NCAA charges the lacrosse selection committee to create a field -- with respect to at-large selections and seeding -- based on a series of non-ordered considerations, many of which are, in some part, related to the RPI (again, an outdated method for establishing delineations between teams, akin to making a spacecraft out of a kite and running really fast). To wit:
31.3.3 Criteria for Selection of Participants -- Division I and II. The following criteria shall be employed by a governing sports committee in selecting participants for NCAA championships competition, and a governing sports committee that wishes to use additional criteria must obtain Championships/Sports Management Cabinet approval before doing so:
- Won-lost record;
- Strength of schedule; and
- Eligibility and availability of student-athletes for NCAA championships.
The lacrosse selection committee has received approval to utilize a series of primary factors to assist in the review of a team's won-lost record and strength of schedule. This primary criteria is, in general, an exercise is reiterating the RPI because it is built substantially -- an in some cases, exclusively -- on the tomato-cans-and-string telephone system that is the RPI:
- Strength of schedule index based on the 10 highest-ranking opponents in RPI. Two games against the same opponent count as two contests.
- Results of the RPI
- Record against teams ranked 1-5, 6-10, 11-20, and 21+.
- Average RPI win (average RPI of all wins)
- Average RPI loss (average RPI of all losses)
- Head-to-Head Competition
- Results against common opponents.
- Significant wins (against teams ranked higher in the RPI)
- Significant losses (against teams ranked lower in the RPI)
- Location of contests.
So, the only things in the primary criteria that have nothing to do with the RPI -- a functional attempt to make combustible fuel out of common arts and crafts paste -- is head-to-head results, results against common opponents, and the location of contests. That's it; everything else is tied to the RPI (a shovel with no handle). Does the selection committee even understand that it is charged to populate and organize a field based heavily on a metric that doesn't do the job that many think it does? Even more importantly, do selection committee members even know how and what the RPI measures? (My gut says "Hell no," which is sad considering that committee members always point to a team's RPI ranking and strength of schedule ranking (which is based on the RPI) and use this as a means for arguing why a team is in or out of the NCAA Tournament field. Education is important here, and given the loud shouts regarding why the RPI is a mess when used in a vacuum and the NCAA's continued deployment of the metric in despite these calls for reform, I doubt that any committee members are getting a crash course in how the RPI operates, its shortfalls, and what the RPI actually measures.)
How can this situation get fixed? I have a two ideas:
- Outfit the committee with more tools that supplement the RPI, tools that attempt to actually provide context as to how teams have performed in their games and how their performances relate to the performances of other teams. I don't care what data is utilized -- use LaxPower's ratings, use efficiency information, use stuff that I don't even know exists. Just arm these individuals with useful tools that assists the committee in its charge. An excess of options isn't always a good thing, but giving people appropriate information that is cogent to an analysis is paramount to making sound decisions. Right now, the selection committee is woefully under-outfitted when it comes to operative information that is useful in 2013.
- Reconstitute the selection committee with more committed lacrosse people. Right now, the NCAA Tournament selection committee consists of VMI head coach Brian Anken, Loyola head coach Charley Toomey (only two coaches -- at most -- are permitted on the committee at any time), Heather Lyke, an associate athletic director at Ohio State, Ellen Crandall, as associate athletic director at Hartford, and Jim Siedliski, the Big East's Associate Commissioner for Olympic Sports. Wash out the majority of administrative people and get coaches -- current and/or former -- that are watching these teams and games on a day-in and day-out basis on the committee; harness their innate ability to identify competitiveness. Coaches understand what makes teams great, and it's their job -- current or former -- to fill their days with that knowledge. They have information that others don't have on teams, and their network can outfit them with all the information necessary to make sound decisions on tournament selections. Draw from throughout the country and from every conference so that the entirety of Division I is represented and considered. If you also provide this group with operative information greater than simply the RPI, subjectivity is mitigated and intelligent lacrosse people can make intelligent lacrosse decisions. I'm not saying that administrators can't make good decisions, but they aren't lacrosse people and the majority that they hold on the current committee (especially when one of the coaching members needs to leave the room due to his team being under consideration) seems backwards. Hell, if you wanted to put an administrator or two in a position of oversight (something that they're professionally competent at doing) over the coaching group to limit conflicts of interest, that's fine. But more lacrosse people making lacrosse decisions is a good thing.