The first was based on the team's non-conference schedule, and it used the team's opponents' typical RPI rankings. In the second category, bonus and penalty points were awarded for victories against the top 50 RPI teams and losses to teams rated lower than 150 on the RPI. The third method simply ranked the teams from highest to lowest number of points they could score. This was called the "margin of victory" method because it counted the number of more important victories (those that determined the winner of the game) over less significant ones (those that determined only the final result).
In short, the goal is to try to win as many games as you can while still keeping your rating high enough so that it doesn't drop too far when you lose some games here and there. If your rating drops low enough, you might not be invited to any more tournaments.
Every year, several new methods are tried by different organizations who compile ratings for the NCAA Tournament. Two of the most common methods are the "strength of record" method and the "elimination tournament seed system" method.
The strength of record method looks at how successful a team is during the season versus what their record shows. For example, if a team wins all of their games by an average margin of victory of 5 points or more, they would receive a "5 star" ranking.
The RPI is computed by summing three components. Part I (a quarter of the formula): The team's winning percentage The NCAA implemented a bonus/penalty system for the 2005 season, in which each home victory or road loss is compounded by 0.6 in the winning percentage computation. The difference between a home defeat and an away win is increased by 1.4. For example, if a team wins 60% of its games at home and loses 40%, its part I score would be.60+.40=1.00. Its total rating index score would be 1.00+0.84=1.84.
Part II (three-fourths of the formula): A team's strength of schedule Rank the other teams on your schedule from #1 (best) to #160 (worst). If two teams have the same record but one has played more difficult schedules, then give the advantage to the team that has played more ranked opponents. If they are still equal, then split the difference and add 1/8 of a point to each team. So, if a team has a ranking of 30 and another team has a ranking of 31, they would be equal with a total rating index of 3.125. However, since the higher-ranked team has played one more game against teams ranked lower than itself, we will add 1/8 of a point to that team's rank and divide by 2 to determine its final rating index score: 30+1.5/8=3.125.
The RPI, or Ratings Percentage Index, is a statistical calculation used in college basketball to help determine how teams are picked for the NCAA tournament and where they are ranked once there. The RPI takes into account all the games played by each team, with more weight given to winning teams. Teams are awarded points based on how well they do against other teams in their conference and nationally.
There are two main types of games used in calculating the RPI: non-conference games and conference games. In non-conference games, which make up about one-third of all games played, teams play so-called "out-of-conference" games. These are games that don't count toward their record but do affect the RPI. For example, if a team were to lose every game outside the conference, their RPI would still be affected because it would be difficult to win enough games to qualify for the postseason.
In conference games, which make up about two-thirds of all games played, teams compete within their leagues. Here, too, losses will affect their RPI, but only games against other league opponents will contribute to it. For example, if a team lost every game this season, excluding those against rival school Duke, their RPI would remain unchanged at 1,000 since these are games against non-conference opponents.
The rating percentage index, or RPI, is a metric used to assess sports teams based on their wins and losses as well as their strength of schedule. The strength of schedule is comprised of both the opponents' winning percentage and the winning percentage of those opponents' opponents (SOS). The RPI is used by college basketball coaches to determine who they will play in the NCAA Tournament.
Each team is required to submit a full RPI listing that includes all of their games played against ranked opponents during the season. The RPI ranges from 0 to 1 with higher numbers representing more difficult schedules. A team's RPI can change after each game because rankings consider all previous data including forfeits. For example, if a team loses to lower-ranked team and then beats another lower-ranked team, its RPI would increase.
Teams are given different levels of seeds depending on their RPI. A team's seed depends on how they rank compared to other remaining teams in the tournament. If there is an even number of teams remaining in the tournament, like this year, the committee uses a "random selection" process to decide which teams will play in what round. If there is an odd number of teams remaining, like last year when Maryland lost to Notre Dame before Michigan State played Georgia, they would have been given a bye into the second round.
In conclusion, the RPI is used to determine who plays in the NCAA Tournament.