The RPI is officially dead in NCAA men's basketball.
The NCAA announced on Wednesday a significant upgrade in the way its Division I men's basketball tournament will be seeded and selected. The new protocol is effective for the upcoming 2018-19 season.
The RPI, a long-used/official data point for the NCAA in how it rated out every team in the sport, will be put out to pasture. What's in? It's called the NCAA Evaluation Tool -- NET for short. Whereas the team sheets that the selection committee used were reliant upon the RPI not just for an overall ranking, but for strength of schedule and all other official markings therein, the NET's metrics will now replace the RPI on the NCAA's team sheets, which are staying.
This also means that the NCAA will not be opting to crowdsource a ranking -- to use a composite set of multiple different metrics -- as was heavily debated by the organization in the past couple of years.
According to the NCAA, the NET will rely on:
- Game results
- Strength of schedule
- Game location
- Scoring margin
- Net offensive and defensive efficiency,
- Quality of wins and losses
In addition, the NET will also tweak some data to "ensure fairness."
"To make sense of team performance data, late-season games (including from the NCAA tournament) were used as test sets to develop a ranking model leveraging machine learning techniques," the NCAA said. "The model, which used team performance data to predict the outcome of games in test sets, was optimized until it was as accurate as possible. The resulting model is the one that will be used as the NET going forward."
Important to note that the model is predictive in nature, which the RPI was not.
It's a progressive step forward. As to how effective it will be vs. other popular metrics (such as the KenPom and Sagarin Ratings) remains to be seen. This overhaul to how the selection committee assembles the field of 68 in men's basketball has been in the works for a couple of years; the NCAA men's basketball selection committee approved the new protocol on a conference call in late July after talking extensively about it at its summer meeting in Boston in mid-July.
The idea for the NET was formulated with the help of Google Cloud Professional Services, in addition to input from the National Association of Basketball Coaches (NABC) and insight from many of the leading analytics experts in college basketball, who met with the NCAA back in January 2017.
"The NET was built to create a ranking system that was as accurate as possible while also evaluating team performance fairly," the NCAA's release states.
Most importantly, when a game is played is not a factor. Whether it's the first game of the season or the last in a conference tournament setting, all win/loss results will be evaluated equally. End-of-season performance has no bearing in the NET'S calculus or in the team sheets.
As for scoring margin, it now matters -- to a certain point. The NET puts a cap on margin of victory/defeat at 10 points "to prevent rankings from encouraging unsportsmanlike play, such as needlessly running up the score in a game where the outcome was certain."
In the past two years, significant upgrades have been made to the seeding-and-selection model for the committee. Last season, the quadrant system was put into place, in an effort to bring more balance to team sheets and properly reward teams who not only won in road and neutral environments, but credit them for playing those games in abundance vs. other like teams vying for similar seed lines or inclusion into the tournament.
The quadrant system, which had some critics in its first year of use, will remain in place with the same lines of delineation in the NET and on the team sheets going forward. Here's how that breaks down:
- Quadrant 1: Home 1-30, Neutral 1-50, Away 1-75
- Quadrant 2: Home 31-75, Neutral 51-100, Away 76-135
- Quadrant 3: Home 76-160, Neutral 101-200, Away 135-240
- Quadrant 4: Home 161-353, Neutral 201-353, Away 241-353
The team sheets will continue to officially include, as reference, metrics such as KenPom, Sagarain, KPI and Strength of Record. But important to note that those metrics, while on the team sheets, do not factor into the NET.
For the RPI, it's a long-awaited death in college basketball circles. When the metric was introduced in 1981, it was considered important and progressive for establishing how good teams were vs. the rest of the field. But with the advancement of analytics over the ensuing four decades, the RPI has long since proven to be simplistic and manipulable -- if not outright misleading in some instances.
For now, the RPI will still be the default data point for other NCAA-sanctioned sports who rely on selection committees for their respective national tournaments. Women's basketball, for instance, will not see a change for the 2018-19 season.