# Clyde Berryman's College Football Rankings

QPRS (QUALITY POINT RATING SYSTEM) - QPRS is a mathematical formula rating system for evaluating the performance of American football teams which takes into account a team's Schedule Strength, Won-Loss Record, and its Points Fielded / Points Allowed.

I am currently only able to do a final end of season college football ranking. My "day job" often involves long or unpredictable hours, travel, etc...Perhaps in retirement, I'll have the time to be able to do a weekly standings of sorts.

My QPRS system is not overly complicated but it is labor-intensive since I'm a shameful computer "dinosaur" and I do all my calculations by hand with pencil, paper, and a calculator. I rate circa the top 65 Div 1A teams (all those with a winning or even record plus any that have particularly noteworthy "tough" schedules) to then determine my Top 40.

## Clyde's "Quality Point Rating System"

QPRS - A Mathematical Formula-Based Rating System Created for
the Purpose of Ranking NCAA American College Football Teams

******

The Quality Point Rating System {QPRS} evolved as a result of this writer's frustrations with the week-to-week opinion poll system of ranking America's top college football teams. The media polls, whether Associated Press {AP}, United Press International {UPI}, CNN/USA Today, or other, reflect the subjectivity of their voting audiences. I always believed that a more methodical approach to rating teams such as the methods used by parent organizations to rate chess players or tennis players would be more desirable

At the same time, I decided to keep the rating system simple enough that it can be performed by any individual who owns a calculator, pen, paper, and who has access to complete football score results. Time is another factor. I find that it usually takes me a full workday {eight hours} to comfortably rate the top 20 teams of any football season using QPRS.

QPRS underwent considerable testing and revisions before I settled on its current form. Some of its rating conclusions on certain football teams will be sure to challenge long-held popular conceptions. Top teams in the general public's mind do not always fare so well while lesser known teams from particularly competitive seasons emerge at the high end of the scale. {Note: In comparing teams from different eras, I do not for one moment pretend that a team from the early years could beat its modern-day counterpart. Football tactics in college have greatly evolved while players have become more powerful, faster, and more professional than their forebears. Rather, I compare how one team did within its season against how well a later team performed against its opponents in its season.}

Unlike the polls which often take on a narrow "who-beat-whom" focus and which seem to rely on increasingly short-term memory as the season progresres, I wanted to create a formula which would accurately measure a team's strength based on its overall season-long performance. I decided early-on that to accurately gauge a team's greatness in comparison to its competitors, it would be necessary to examine how it fared in three key categories:

1. Its overall Won-Loss-Tie record,
2. Its Schedule Strength, and
3. Its average margin of victory or defeat (i.e., Avergage Points Fielded {APF} for Offense, and Average Points Allowed {APA} for Defense)

In order to demonstrate how the Won-Loss-Tie record, Schedule Strength, and average margin of victory/defeat interact in QPRS, the reader's attention is drawn to the examples which appear below. Each involves three hypothetical teams labelled, for the sake of simplicity, Team A, Team B, and Team C. In each example, Team A is the strongest team, Team B is in the middle, and Team C is the weakest team.

EXAMPLE ONE: Differing Won-Loss-Tie Records: All factors for each of the teams remain constant except for different Won-Loss Tie records. Each team played a 12-game season.
TeamWon-Loss-TieSchedule StrengthAverage Pts Fielded Average Pts AllowedOverall Rating
Team A11-1-040.0020.0010.00 347.40
Team B10-2-040.0020.0010.00 314.00
Team C 9-3-040.0020.0010.00 280.60
EXAMPLE TWO: Differing Schedule Strengths: All factors for each of the teams remain constant except for different Schedule Strengths.
TeamWon-Loss-TieSchedule StrengthAverage Pts Fielded Average Pts AllowedOverall Rating
Team A10-2-050.0020.0010.00 347.40
Team B10-2-040.0020.0010.00 314.00
Team C10-2-030.0020.0010.00 280.60
EXAMPLE THREE: Differing Average Points Fielded on Offense: All factors for each of the teams remain constant except for different Average Points Fielded.
TeamWon-Loss-TieSchedule StrengthAverage Pts Fielded Average Pts AllowedOverall Rating
Team A10-2-040.0030.0010.00 340.00
Team B10-2-040.0020.0010.00 314.00
Team C10-2-040.0010.0010.00 270.60
EXAMPLE FOUR: Differing Average Points Allowed on Defense: All factors for each of the teams remain constant except for different Average Points Allowed.
TeamWon-Loss-TieSchedule StrengthAverage Pts Fielded Average Pts AllowedOverall Rating
Team A10-2-040.0020.00 5.00 344.00
Team B10-2-040.0020.0010.00 314.00
Team C10-2-040.0020.0015.00 291.15
By transposing the above results to a graph, the reader will note that a better Won-Loss-Tie record or a more difficult schedule causes a team's overall rating to climb higher at a smooth, progressive rate. A higher Average Points Fielded, however, gains ground quickly but then tends to flatten out at the top of the curve. Conversely, a low Average Points Allowed tends to rise at the top of the curve for those truly tough defensive teams which only allowed a few points per game. In effect, this is a mild uay of putting a cap on the credit a team receives for rolling up scores against already-beaten opponents while also recognizing those unyielding defenses which allow opponents few opportunities to score. {NOTE: Some mathematical formula rating systems tend to handle runaway scores rather brutally and indiscriminately by "collapsing" scores when an arbitrary point margin is achieved.}

In QPRS, the Won-Loss-Tie record of a school and its Schedule Strength are inextricably linked. With regard to Schedule Strength, I look at Div 1A as being roughly divided between "normal schedule" {most major conferences, leading independents} and "weak schedule" {Big Sky, Big West, Mid-American, and to a large extent, Western Athletic}. Up until very recently {1990's}, a number of East coast teams {many now in the Atlantic Coast or Big East conferences} would from year-to-year fluctuate between "normal" and "weak" schedule category based on their opponents each year. Some independents cross the line from "normal" to "weak" from year-to-year such as So. Miss., Memphis St., Army, Navy, Ea. Carolina, Louisana Tech, etc. Some of these, such as Louisville, Cincinnati, Tulsa, and others appear to be making a determined effort to stay in the "normal" category of late. I treat Div 1AA opponents, when they occur in Div 1A play, as a third and weaker category choosing the category in which a team belongs can be somewhat subjective in cases where a close call is involved. This was particularly true ten or more years ago when the greater number of eastern independents created unevenness and some tough calls. As recently as 1982, for example, Penn St. played a number of close-call "weak schedule" opponents. However, since these "weak schedule" opponents were by-and-large winners, the net result is that Penn St. still comes out credited with a fairly tough overall Schedule Strength {53.74} for 1982.

For the sake of reader interest, the previous examples using Teams A through C relate to fairly successful teams whose performances would usually put them in the upper half of most end-of-season polls or formula-based rankings. While the Won-Loss-Tie record is self-explanatory, the following is a rough guide of how to interpret QPRS ratings in the areas of Schedule Strength, Average Points Fielded, Average Points Allowed, and the respective Offense, Defense, and Overall Power Ratings:

```Schedule Strength:

60s - Very Tough Schedule
50s - Tough Schedule
40s - Average Schedule
30s - Weak Schedule
20s - Very Weak Schedule

Average Points Fielded:                    Average Points Allowed:

40+ - Very High-Scoring Offense            - 5 - Very Tough Defense
30  - High-Scoring Offense                 -10 - Tough Defense
20  - Average Scoring Offense              -15 - Average Defense
1O  - Below Average Scoring Offense        -20 - Below Average Defense
less than 1O - Poor Scoring Offense        -25 - Poor Defense

POWER RATINGS:

Offense:      Defense:      Overall:

201 - 250     181 - 225     400 - 500  =  Outstanding
151 - 200     136 - 180     300 - 399  =  Good
1O1 - 150      91 - 135     200 - 299  =  Average
51 - 1OO      46 -  90     1OO - 199  =  Below Average
1 -  50       1 -  45       0 -  99  =  Poor
```
In sum, I hope the above.explanations are useful toward understanding the QPRS NCAA College Football rating system and that the reader will find QPRS rating results of past and present college football teams of interest

(c) June 1994 by Clyde P. Berryman

## Commentary on 2007 Season

2007 - the 'upset' season - witnessed fierce competition for the national college football title right down to the wire. One of the most exciting seasons in recent memory from a spectator's standpoint. From a purely ratings perspective, the fact that there were no stand-out teams with tough schedules who were able to dominate means that 2007 National Champion Louisiana State (12-2) has one of the lowest national champion ratings since the Washington Huskies back in 1984. In fact, 2006 Louisiana St. (11-2), which only ranked fourth in the QPRS standings last year (behind Florida, Ohio State, and Southern Cal), had an eerily close overall rating (380.123) and schedule strength (53.49) to this year's champion team. Just for historical context, I'm also attaching my QPRS American College National Champions list which goes back to 1940 and shows that increasingly, a team usually needed an overall rating in the 400's to become the champion. None of this year's teams broke into the QPRS Top 100 American College Football teams since 1940 list either.

The amazing upsets of this past year served to highlight just how off-target many of the pre-season polls or rankings could be. You obviously can't stop human nature and everyone enjoys to speculate and have their pre-season favorites. However, I have always been against polls or rating systems which actually make use of a pre-season ranking as a starting point by assigning teams a subjective pre-season ranking order. I think we all remember a few seasons ago how Auburn missed a chance at playing in the national title game only because they started off too low in the pre-season polls. An arbitrary starting pecking order based on pre-season 'gut feelings' only serves to pollute or skew the year-long validity of a poll or ranking which incorporates such subjective information. Serious rankings should be based only on results which take place on the football field once the season is underway.

One thing which QPRS also does is 'self-correct' with regard to the difficulty of a team's opponents as the year progresses. Just as an example, when Oregon St. beat California this year, it was facing a 5-0 opponent and it got a nice bounce in the polls for defeating a highly-ranked, unbeaten team. By the time 'lowly' Stanford beat them near season's end, California had lost five times and the game went by almost unnoticed. Does Stanford deserve less credit for beating California in 2007 than Oregon State? It was the same California team - pretty much the same players, same coaches. What happened is that Oregon State was simply the first team chronologically to learned how to exploit California's vulnerablities. These flaws were there when Tennessee, Arizona or Oregon played them beforehand but they simply didn't get the job done. In some polls or rankings, Oregon State would get credit for beating a highly-ranked 5-0 California (100%) while Stanford would only get credit for beating a forgotten 6-6 (50%) California which by then had dropped well out of the Top 25. So in QPRS, at year's end, both Oregon State and Stanford (and all the other teams before, after and in-between) get the same credit for having beaten a California team with a 7-6 season record. Obviously, margins of victory will differ and may affect the ratings but they all played against 2007 California.

## Commentary of 2016 Championship Game

```09-Jan-17 Alabama                     31 Clemson                     35 Tampa FL
```
My heart sank in the fourth quarter of the Alabama-Clemson game as I watched all that the Tide had done right in the beginning of the game start to come unglued. The last minute score by Clemson sealed it. How did Alabama get it so wrong when they were rightfully the odds-on favorite to win by at least 7 points? Overuse of RB Bo Scarbrough until he was finally injured was one reason. He did all you could expect of him with his first-half touchdowns but the Tide needed to vary the offensive game lest it become too predictablewhich it certainly became in the end. They thankfully threw to O.J. Howard and belatedly brought in Damien Harris to replace Scarbrough but they vastly under-utilized star performers like ArDarius Stewart, Calvin Ridley, and Gehrig Dieter among other explosive offensive weapons. QB Jalen Hurts should have been allowed to run far more frequently on QB-run designed plays in the second half. The question will always remain whether Offensive Coordinator Lane Kiffin should have been left in charge of the offense he was familiar with face-to-face and on the ground rather than bringing in Steve Sarkisian for his first game as Offensive Coordinator as he may have lacked the needed personal chemistry with individual players in a championship game situation. On defense, the Tide did reasonably well in the first half but the perennial problem remains that while its defensive corners were fast enough, they remained too short to deal with some of the tall Clemson wide receivers. Clemson QB DeShaun Watson was rightfully praised by one of the TV Commentators as being like an Offensive Coordinator and QB all rolled into one. His passing precision while under pressure was exceptional.

While Clemson won the championship game, Alabama remains the best team of 2016 in the QPRS College Football Ratings. Lest I be accused of bias in that Alabama is indeed my alma mater, I will point out that more than two thirds of the individual math-formula rating systems carried by Massey Ratings (http://www.masseyratings.com/cf/compare.htm) all agree Alabama was the top 2016 team. This is not surprising. Alabamas sole loss came in the championship game against Clemson (14-1) while Clemsons only loss was during the regular season to Pittsburgh (8-5). Alabama won ALL but ONE of its games (against Mississippi) by a margin of more than a 7-point TD/PAT. Clemson, however, won almost half of its games this year by only 7 Points or less! Both Alabama and Clemson had very similar schedule strengths and their relative Points Fielded on Offense were also quite close but where Alabama was in a league of its own was in Points Allowed its defense only gave up 195 points in 15 games! 2016 was definitely Coach Nick Sabans strongest Alabama team to date and it ranks as the 10th greatest college football team ever since 1920 in the QPRS Top 100 College Football Team rankings. 2016 Clemson ranks as the 16th greatest team since 1920 and is the highest ranked Clemson team of all time (the 1981 Clemson Tiger national champions, for example, are ranked 79th in the Top 100.)

Be aware that contests which depend on single-game results do not guarantee that the best team or competitor over an entire year will win the championship event. The Super Bowl winner has not always been the best season-long NFL team either for that matter. Playoffs exist to create buzz, sell tickets and additional advertising , but they can often skew who is best based on a thorough analysis of the results of an entire 12 to 16 game regular season. (The Chase in NASCAR auto racing was similarly devised to create artificial excitement with a reset at the end of the regular racing season but especially, to sell more tickets and create more televised advertising opportunity.)

## Commentary on 2017 Season

Alabama has won its fifth national championship title under Coach Nick Saban. Alabama during the first half of the season looked totally unstoppable. Then something happened in a few of the latter games where they had to work harder to get the win (notably Miss. St.) and then came the loss at Auburn. Some will say it was because the opponents got tougher but the previously mighty offense was slowing a bit. Was it the play calling? Was QB Jalen Hurts taking longer to make decisions after the snap, was he being too cautious or unable to find the gaps as quickly as before when forced to scramble? In the second half of the championship game against Georgia, Saban took the gamble of taking Hurts out and putting in Freshman QB Tua Tagovailoa who proved remarkably clear-headed and precise in his passing when thrown into such a high-stakes situation. Does this mean he automatically gets to be the starting quarterback at Alabama in 2018? Not so fast. There is actually a very strong argument for Alabama to have two starting quarterbacks given what very different strengths Hurts and Tagovailoa each bring to the table. They should be used interchangeably depending upon the defensive strengths of each opposing team they will face, the given situation as the game progresses, and to some extent, to keep the opponent guessing who they will face and what kind of play calling changes may come with each QB substitution. I am usually not in favor of a two-quarterback starting situation but these two particular QB's could make a very powerful argument for such an arrangement next season.

The SEC was no longer the all-dominant conference we have seen in recent years past. The Big 10 was probably the strongest overall in 2017 with Wisconsin, Ohio State, Penn State, Northwestern and Michigan State all posting very impressive records. There is also increasing parity between the 'Power 5' conferences (ACC, Big 12, Big 10, Pac 12 and SEC) and select schools from the next conferences, most notably the American Athletic Conference and Mountain West. Undefeated Central Florida is naturally disappointed not to have been given a chance to be in the College Football Playoff but, not to take away from its remarkable year, there is a difference between playing almost every weekend against 'Power 5' caliber opponents versus once or twice in a season. With regard to schedule strengths, schools like Maryland, Notre Dame, and Iowa faced some of the toughest schedules this past year.

I am still not in favor of a College Football Playoff as the method of selecting the college National Champion each year. I am not a big fan of ranking a team based only on 'who beat who' on any single day. The winner was certainly the better team during those sixty minutes of play but if the match-up were repeated 100 times, which team would win the majority of the time? That would be a much more representative result. The NFL team which wins the Super Bowl is not always the best NFL team that year - it is the one which avoided injuries to significant players and which has hit its stride as a cohesive unit toward the end of the season. A wild-card team that finally gets 'hot' at the end of the year can win the Super Bowl by beating a team with a 14-2 regular season record but which is now plagued with injuries and/or not playing as well as they did in early-mid season. In all the systems out there, from pre-season AP polls which create a flawed ranking based on no scientific evidence (but which can skew the standings for the rest of the season) to playoffs (or the NASCAR 'chase' if you like), the end of the year for no good reason becomes much more important than the beginning . Playoffs like the Super Bowl and the 4-team college playoff exist to get the public excited but mostly to make more money for TV networks, sponsors, etc. We are kidding ourselves if we think they are seriously designed to identify which was the best football team for that entire year. The only playoff system (or 'who beat who' method) which has any semblance of validity would be a Swiss round-robin system (such as is used in some chess tournaments) where from the outset, winners keep playing winners while losers are increasingly relegated to playing other losers and that is how over several rounds, a pyramid of top players develops. That would be too unwieldy in football given conference alignments, schedules determined years in advance, not to mention distances covered. The BCS era was probably the closest we came to identifying the best year-long teams. The average of the different computer systems that were used as a part of the BCS also eliminated human bias although no system had a perfect math algorithm for measuring football strength - that is why an average of several systems was wisely put into place.

Clyde Berryman / hardbraking at hotmail dot com