After a two year hiatus, DJ Ross Power Performance Rankings are back. For the uninitiated, the PPR is a rankings system based on fundamentals. One can always guess, speculate and tweak formulas based on statistical probabilities in order to come up with nothing more than a popularity contest. I have instead compiled a list of fundamentals that indicate what has been done on the field, not on what some boardroom categorically decides should determine who is ranked and who isn't. Instead of assuming superiority because Team A belongs to conference XYZ, I let the results of competition decide.
...so...given all the various rankings systems around why is mine any better? Well here are some points to consider, and who knows maybe they will convince you enough to make sure that the PPR is part of your weekly watch list for collegiate football.
Every season each of the 119 teams in Div IA start at 0. That is right. No pre-season rankings boost to unfairly position teams before the whistle has even blown and the game clock started. Every year several ranked teams end up falling out as they prove that they are not up to expected ranking levels. We have seen ranked teams lose up to 3 games in the early season only to hang on by their teeth to a 25th or 24th spot. We have also seen teams that are undefeated by mid season only to be getting votes while several two loss teams are ahead of them in the rankings. Now there may be a motivation to that, but since it isn't based on fundamentals, we can only speculate. The PPR starts each team at the same level at seasons beginning turning it into a footrace.
SOS is only counted into the formula once. Unlike many statistical rankings systems who double dip through giving bonus allotments for victories over “ranked” teams, and using poll data that has already been influenced by so called superior conference rankings, the PPR uses SOS solely as a defining factor in separating teams with apparently identical points rankings. For example if Team A has a earned 100 points and Team B has earned 100 points, than by comparing their SOS we can discern which team has a better performance ranking.
While statistical rankings base many factors on probabilities, the PPR only deals with what is. Various fundamentals parameters are used to determine the power or strength of each team. Parameters such as: Wins, losses, Overtime games, Road Victories, Wins over Div IAA teams, Losses by Div IAA teams, Conference Victories, Shutouts,Winning Streaks, Losing Streaks,etc. Factors such as these are real, not speculative. There is no guesswork involved when Team A travels to Team B and beats them 28-7. There is no guesswork when Team C has won 5 straight and has lost the last 3 games. There is no guesswork when team D has been shutout 3 times over the season.
The PPR eliminates two inherent problems with the current rankings system. One it eliminates the need or interest for preseason polls. Since it is based on fundamentals it literally shows just how strong or weak teams are based on what they have accomplished. Two it eliminates this mysterious need for such a late release for other official polls. Waiting until October for a poll to have worked itself out and thus being ready to show the public doesn't really lend confidence in said poll. This just shows that the PPR is not beholden to any specific interest, but illustrates just how well all teams performed during the season.
The PPR eliminates the ambiguity between good and bad teams. It is frustrating for many fans who look at polls and either never see their team listed since the poll only focuses on the top 25, or find their team listed as 45th and have no real understanding as to what the difference between 45th and 44th or 50th for that matter. The PPR provides a clear point of reference for all teams based on performance. If the team is having a great year it really shows, and if the team is performing badly, it also really shows.
Given the changes in the BCS starting in 2006, rankings will be even more important on a conference level when determining automatic bids to the all important BCS bowl games. The PPR provides an easy and clear picture of conference performance strength. This more than anything should be considered when determining which conference champions should be granted a bid.
In conclusion the PPR provides poll and rankings watchers everything that they are missing: A clear cut picture of just how good or bad their favorite teams are compared to the rest of Div IA football.
--Douglas J Ross
Explaining a few of the fundamentals in the PPR'd...
While wins and losses are obviously vital to any rankings system, there are some other parameters that might not be so obvious or at least understood:
STREAKS-Winning and Losing streaks are exciting and devastating depending on which side of the streak your team is on. Each week fewer and fewer teams are undefeated as competition heats up. While winning a game is valuable in showing how powerful the team is, a constant string of wins is exponentially more valuable. Each time a team wins, the likelihood of them continuing to win is lessened. Some of the greatest winning streaks throughout history boggle the mind. It is extremely rare for a team to go through an entire season undefeated. The 2004 season was an enigma where 3 teams ended the season without a loss.
The opposite side of the coin is also true. Even the worst teams have a difficult time losing every game during a season, which is why it is uncommon for that to happen as it is for teams to remain undefeated. Even with an average of two teams going undefeated or winless each season, that still only represents 1.7% of all Div IAA schools. So the PPR considers both winning and losing streaks, and the more a team wins or loses, the more points or negative points they receive. This parameter is measured exponentially which is why a team with ten wins or losses in a row would have a performance ranking much higher or lower than a team that can win/lose two or three games before going the other way.
W/L DIV IAA TEAMS- Big time football is big time, and if you are currently one of the 119 Div IA teams, than you should play that way. Div IAA is AA for a reason. They are smaller programs with less resources and visibility (with some marked exceptions like the Ivy league). Sure one could argue that there are currently teams in IA that don't belong in IA, but hey I don't make the rules. If they are currently accepted as being Div IA, so be it. So what does it mean to beat a Div IAA team? Not as much as beating a Div IA team therefore the value of a win vs Div IAA is prorated. What does a loss at the hands of a Div IAA team mean? It means a whole lot more than being beaten by a Div IA team. If a IAA team can come into your house and beat you, than you teams legitimacy in the Div IA ranks is in question, therefore negative points awarded for a loss to a Div IAA team are enhanced.
SHUTOUTS- Regardless of how powerful a team is, shutting out the opponent is a difficult task, and only occurs sporadically throughout the season. Even in matches where strong Div IA teams play Div IAA teams, a shutout is not that common. In week one of 2005 only two shutouts at the top level were recorded, and no Div IAA team were among the two teams shutout. Again this is a difficult task and therefore worth something. Getting shutout is also worth something on the negative side.
CONFERENCE VICTORIES- Much is made of which conference is better or best, and makes for great tailgate fodder. The truth of the matter is, it is conference victories that are most important to a team because they translate into championships. Sure one could argue that Conference A is much better than Conference B, and therefore their wins are as powerful or important because they don't play as tough of competition in their league. Many also quip with the classic argument that if their favorite team were in league A they would never win any games because it is so much tougher in that league. At the risk of offending some readers, that is bunk plain and simple.
The truth of the matter is that most teams where the resources are applied to developing the football program are able to make the necessary adjustments to compete in a new league, even at a new level of competition. The transition may be rough or it may not be. It depends on so many factors that it is impossible to quantify. The reality is that it is harder to play and beat the team you meet every year than it is to beat the team you meet once every five years. Familiarity breeds difficulty. Coaches begin to read each other, understand the small things about their opponents that make it that much harder to surprise or dominate. We see this type of thing happen every year when a 2 win team knocks of their conference leader in week 8 or so because they know that team so much better than their opponents. A conference victory is much more difficult therefore conference victories as well as annual rivalry games between inter-conference teams such as ND vs USC or UC vs CSU are given a value in the PPR.
OVERTIME- Another factor that doesn't often occur. Athletes have been out on the field giving everything they have to win. This is why overtime is so important. It requires the teams to dig down that much deeper in order to bring home the victory. The PPR attributes a value to this as well.
ROAD VICTORIES- With todays college game more dependent on big revenues than ever, the home game has become an economic Mecca. The NCAA is tinkering with adding a 12th game to schedules in order to provide that all to important revenue stream of ticket and concession sales for schools looking to pay the bills. Without the advantage of a home field, winning becomes more difficult. Thus winning on the road is more impressive than winning at home considering what it takes to take 80 players from point a to point b on a season schedule while keeping them fit, focused and ready to fight. Not an easy task in the best of times. Also in college football the scheduling is such that you don't get a home and away in the same season. The PPR definitely attributes a value to this.
No comments:
Post a Comment