THE POWER RANK

The Top 25 College Football Teams of 2013 by Recruiting Rankings

By Leave a Comment

Nick_Saban_StatueRecruiting rankings do matter.

Each year, Rivals assigns a rating or points value to each school that describes the talent of the players who signed a letter of intent. For Sports Illustrated, we developed a model that takes the Rivals ratings and predicts future team performance. To compare the rankings from this model with the preseason AP poll, we looked at which rankings better predicted the final AP poll.

The Rivals model did as good or better than the preseason AP poll on 46 out of 100 teams over the last 4 years. This is remarkable given the limited information the recruiting model has compared with the writers that vote in the AP poll.

To get the full story on SI.com, click here.

Before we count down the top 25 teams for 2013, we note the following about this regression model.

  • The regression model has learned from the past by relating recruiting ratings to the team ratings from The Power Rank algorithm. For example, recruiting data from 2009 to 2012 were fit to The Power Rank’s results for the 2012 season. Our team ratings from the regular season have picked the winner in 62.8% of bowl games over the last 11 years, a better percentage than the Vegas line (62.2%). Part of the accuracy of the recruiting model depends on our team ratings.
  • We used the final AP poll as the measuring stick for the accuracy of the Rivals model and the preseason AP poll. This poll has problems, as it strongly considers wins but almost ignores margin of victory. However, it was the most relevant measurement of team strength for a general college football audience.
  • The 100 teams we looked at were the top 25 teams in the preseason AP poll over the last 4 years. This puts the recruiting model in a better light, since this set of teams didn’t include teams highly ranked by the recruiting model but outside the AP top 25. For example, the recruiting model had Auburn 7th heading into the 2012 season. The preseason AP poll had Auburn 28th, which more accurately predicted their disasterous season.

With the passing of National Signing Day 2013, we now have the Rivals ratings to predict the 2013 college football season. The predictions are based the past 8 years of team performance.

25. Stanford. The Cardinal only had 12 scholarships to offer incoming freshmen. Since the number of recruits directly affects the Rivals team rating, Stanford only had the 61st ranked class this year, a far cry from their 5th ranked class a season ago.

24. Miami (FL). Miami coach Al Golden can’t walk off campus without tripping over a highly touted high school player. However, impending NCAA sanctions made recruiting difficult this year, as they had the 44th best class, much worse than their 9th ranking a year ago.

23. Virginia Tech. Despite two down years, coach Frank Beamer still recruited the 22nd ranked class. The Hokies need QB Logan Thomas’s passing to improve or groom a better passer.

22. California. The Bears continue to recruit well despite the coaching change from Jeff Tedford to Sonny Dykes. However, this ranking is probably too high. The Cal offense will be learning a new spread system under Dykes, while the defense lost coordinator Clancy Pendergast to USC.

21. Washington. The Huskies have continually improved their recruiting rank over the last 4 year, rising from 28th in 2010 to 18th in 2013. Hopefully, some of the linemen recruiting during this time will give QB Keith Price better protection next season.

20. Nebraska. The Cornhuskers pulled in the 17th best class, by far the 3rd best class in the Big Ten. Unfortunately, Michigan and Ohio State were way ahead, and Nebraska has to travel to Ann Arbor this season.

19. South Carolina. With all the media chatter about the talent in Columbus and the draw dropping hits from Jadeveon Clowney, you might think South Carolina is a top 10 recruiting team. In reality, their 16th ranking in 2013 was their best over the last 4 seasons.

18. Tennessee. The Rivals model had the Vols 13th in the nation last season, which was way too high. With new coach Butch Jones taking over, this year’s 18th ranking is also probably too high.

17. Oregon. Over the last 4 years, the Rivals model has consistently underrated the Ducks. The preseason AP poll was more accurate each year. Will this continue after head coach Chip Kelly handed the program over to Mark Helfrich?

16. UCLA. In talent rich Southern California, the Bruins always recruit well. They finally lived up to that talent last season under first year head coach Jim Mora. Their 11th ranked class this year should continue this trajectory.

15. Texas A&M. New coach Kevin Sumlin is getting tons of credit for recruiting the 10th best class this year. Of course, it helps that Johnny Manziel (a 3 star recruit) led the Aggies to their best season in more than a decade.

14. Mississippi. The Rebels and coach Hugh Freeze had a magical signing day, landing two top 10 recruits on the offensive and defensive line. Their 7th ranking is by far their best since 2002. However, three of their rivals in the SEC west are ranked higher in these rankings.

13. Clemson. With their dramatic, come from behind win against LSU in the Chick-fil-A Bowl, the Tigers will surely be an overrated top 10 team in the preseason AP poll. This ranking in the teens seems more appropriate.

12. Oklahoma. In the past 12 years, the Sooners have recruited a top 10 class 7 times. However, none of these top 10 classes have occurred during the last 3 seasons. Coach Bob Stoops needs more talent on the defensive line to start contending for national titles again.

11. Texas. Even Mack Brown can’t recruit after 3 subpar seasons. The Longhorns had the 23rd ranked class, their worst since 2002. Their usually excellent defense really needs help after last season.

10. Georgia. How would the Bulldogs recruiting class would have fared if the coaches had told QB Aaron Murray to spike the ball during the waning moments of the SEC championship game? A win over Alabama would have landed Georgia in the national title game against Notre Dame. Instead, they had the 12th best class, a fine rank but the second worst for coach Mark Richt over the last 12 years.

9. Michigan. After a disasterous 3 seasons under Rich Rodriguez, Brady Hoke has turned around this program both on the field and recruiting trail. The Wolverines notched the 5th ranked class after finishing 7th last year. These are their two best ranked classes over the last 8 years.

8. USC. While we kept hearing about the defections from the Trojan’s class, no one mentioned that coach Lane Kiffin still had five 5 star recruits coming to campus, more than any other school (yes, even Alabama). Lack of talent will not be a problem for USC.

7. Auburn. Yes, feel free to call BS on this one. The Tigers continue to recruit well. But unless the next Cam Newton and Nick Fairley show up on campus this fall, Auburn will not return to elite status this season under new coach Gus Malzahn.

6. Florida State. Call BS on this one at your own risk. Despite disappointing loses to NC State and Florida this year, expect the talent rich Seminoles to exceed expectations next season.

5. LSU. It’s a bit shocking that top recruit Robert Nkemdiche picked Ole Miss over LSU. There’s more certainly in winning games in Baton Rouge. Still, coach Les Miles landed the 6th best class in the nation.

4. Notre Dame. The Fighting Irish took full advantage of their undefeated regular season and landed the 3rd best class in the nation. While we should expect Notre Dame to be good next year, 4th is probably too high.

3. Ohio State. The Buckeyes will not surprise anyone this year. In coach Urban Meyer’s first season, Ohio State started the year 18th in the preseason AP poll. Now, with an undefeated season and a 2nd ranked recruiting class, anything but a national title will be a disappointment.

2. Florida. Last year, first year coach Will Muschamp landed the 3rd ranked class despite finishing 7-6. Somehow, a 11-2 season this year got the Gators the 4th ranked class (although they did top the rankings before National Signing Day). Talent is never a problem at Florida.

1. Alabama. Duh. What did you expect? After going 7-6 in his first year at Alabama, coach Nick Saban still recruited the top ranked class in 2008. This started a streak of top ranked classes for Alabama in 5 of the last 6 years. The rest of the SEC should resort to a voodoo consultant to bring bad turnover luck to Alabama next season.

Outlook

No one should take these rankings too seriously. With the short season and the youth of the players involved, college football is incredibly difficult to predict during the preseason. And these rankings have their problems. There’s an incredibly high likelihood that Boise State will be better than the 60th best team in the nation next year.

However, these rankings are still useful, and not only because they are six months ahead of the preseason AP poll. As with all predictive analytics, use these rankings has a guide to help navigate expectations for next season.

Moreover, this is only the beginning of our preseason college football predictions. To keep up to date, sign up for our free email newsletter below.

Thanks for reading.

Check Out The Power Rank on Paul Finebaum

By Leave a Comment

The Power Rank's article on the BCS title game made the cover down South.I had the honor of being on Paul Finebaum’s radio show yesterday. Paul talks college football all year long from Birmingham, Alabama, and the New Yorker recently titled their profile of him “King of the South”. In prepping for the show, I was blown away with the engagement he has with his callers. He lets them say almost anything, making people like Tammy and Legend the stars of the show. He even traveled to Iowa to visit Robert from Waterloo, a frequent caller with cerebral palsy.

On his show yesterday, we talked about my Sports Illustrated article on the BCS title game between Alabama and Notre Dame. The article is on the cover of the magazine in the south, although it couldn’t beat out Michael Phelps nationwide. For me, these were the three highlights of being on the show:

  • Not getting bashed by callers when I said that Alabama was 2nd in my rankings. (Oregon has the top spot.)
  • Talking about Stanford’s Rose Bowl win and how the team rankings predicted the 6 point win.
  • Saying “computer code” in an interview on college football.

And if a team is predicted to win by 1 point, they have a 52.4% chance of winning, not the 53% I mention on the show.

To listen to the show, click here.

Did Wisconsin’s offense improve over the season? A Rose Bowl preview

By Leave a Comment

The Power Rank uses strength of schedule adjustments to evaluate Wisconsin's running game.

A few weeks ago, I was in New York City to meet with Mallory Rubin and Ben Glicksman, the college football editors at SI.com. I had an idea for a Rose Bowl story. Earlier this season, I wrote a preview of the Stanford Oregon game based on statistics for drives when quarterbacks Marcus Mariota and Kevin Hogan had played. (Oregon’s Mariota would often be pulled when the Ducks built a huge lead; Stanford coach Shaw only had the sense to start Hogan late in the season.) I thought a similar analysis might be useful for the Rose Bowl. Hogan has now been Stanford’s quarterback for 5 games, and Wisconsin has shuffled through 3 quarterbacks this season.

When I sent them the numbers early last week, they liked the analysis but thought it overlapped too much with Stewart Mandel’s excellent article on Hogan. They suggested I base the preview on Wisconsin. It was not the feedback I was looking for on Christmas day. But it did prompt me to look at something I had been interested in all season: what happened to Wisconsin’s offensive line? The traditional strength of the Badgers had played so poorly the first two games of the season that their position coach was fired after two games on the job.

The analysis of Wisconsin’s rush offense became the center of my analysis of their offense. To read the full article, click here.

The Shocking Truth About The Colley Matrix BCS Computer Poll

By 30 Comments

Ed Feng exposes flawed Colley Matrix computer poll used by BCSCollege football fans like us hate the BCS.

Unless you work for ESPN and the BCS contributes to your paycheck, the idea of allowing only two teams to play for the national championship is criminal.

And if you’re reading this numbers based blog, you probably know about the problems with the computer polls used in the BCS rankings.

First, only one of the six ranking systems gives enough details so that others can reproduce the results. The other five black boxes are shrouded in mystery.

Second, the BCS forbids the computers from using margin of victory in their calculations. It does not matter that a 33 point loss says something much different about a team than a 1 point loss. In the name of sportsmanship, the BCS will not give teams the incentive to run up the score.

Last, you may have even heard that Richard Billingsley, the man behind one computer poll, is not a mathematician. As he admitted to the authors of Death to the BCS, “I don’t even have a degree. I have a high school education. I never had calculus. I don’t even remember much about algebra.”

But it gets worse.

Why Strength of Schedule and Margin of Victory Matter

I wasn’t looking for a flaw in a BCS computer poll.

I was thinking about strength of schedule and margin of victory. In college football debates, most people agree that a ranking system should account for these factors. The intuition is obvious. Northern Illinois does not play the quality schedule that Alabama does. Oregon’s 24 point win over a solid Oregon State team says something much different about the Ducks than a 1 point win. However, no one has provided any quantitative evidence to support accounting for schedule strength and victory margin in rankings.

Bowl games at neutral sites provide a simple quantitative test for ranking systems: how often does the higher ranked team win each game? For a system that incorporates neither schedule strength and victory margin, rank teams by winning percentage. For a system that accounts for strength of schedule but not margin of victory, rank teams with the Colley Matrix of the BCS. Last, my rankings account for both.

I spent some time digging into the details of the Colley Matrix.

Colley Matrix does not consider the results of each game

It was easy to get mesmerized by the beautiful mathematics behind Colley’s method. His paper discusses Laplace’s die problem, a symmetric positive definite matrix and solving a linear system of equations. I spent the weekend telling my wife that if college football games had only winners and losers, this would be a dandy little ranking algorithm.

Then thinking back on the equations, it hit me.

The method does not care who a team loses to in ranking them. It considers the win loss record of each team and the number of games played between each pair of teams. However, the specifics of who won each game are not an input to Colley’s method.

Omitting specific game results in a ranking system is like disabling the guiding system on a missile. The technology will do its job, but it will not be that accurate.

You can check this yourself by reading the descriptions of equation 18 and 19 in Colley’s paper. It’s possible to solve for the rankings (equation 17) knowing only each team’s record and how many games each pair of teams played.

As a mathematician, I find this omission appalling. To see why, take Alabama in 2012 as an example. The Crimson Tide lost to Texas A&M, a respectable loss to another top 10 team. But the Colley Matrix does not account for this. Suppose Alabama beat Texas A&M but lost to a bad Florida Atlantic team. Since a top team almost never loses to a bad team, this bad loss should lower Alabama’s rank. It doesn’t.

You can check this with your own example. Wesley Colley has set up a page on which you can add and remove games and recalculate the rankings.

When I first discovered this omission in 2012, Stewart Mandel of Sports Illustrated suggested looking into whether this flawed computer poll was helping Kent State. The Golden Flashes were 11-1 heading in the MAC championship game and ranked 17th in the BCS. If they moved up to 16th or better, they would earn a BCS bowl bid.

Sure enough, the Colley Matrix had Kent State ranked 15th, the highest rank in any computer poll. It did not consider that their lone loss came at Kentucky, a 2-10 team that won zero SEC games that year. This flawed computer poll played a small role in placing Kent State 17th overall in the BCS rankings. I wrote about this on SI.com.

“Sam Feng’s article is a perfect example of anti-science”

In response to my article, I got this tweet the next day. Amidst a flurry of four letter words, a blogger blasted the mathematics behind my analysis of Colley’s method.

I disapprove of Feng because I don’t know what the f*$% he was doing, and I don’t think he knows what the f*$% he was doing either.

I guess that can happen when your writing jumps from a small blog to SI.com. At least he could get my name right.

I promptly replied to his post, and a conversation ensued about the details of the mathematics. In the end, the blogger verified my main conclusion that Colley omits the results of each game. The post started with a rant about “anti-science”. It progressed with a dense mathematical discussion in the comments. It ended with this in the last comment.

I’d rather have Ed’s rankings making the decisions than Colley’s or a roomful of NCAA bureaucrats.

The blogger still had a problem with the example I used in the article, a criticism with some merit. The example is equivalent to the Alabama scenario above. In exchanging the loss to Texas A&M for a loss to Florida Atlantic, the records of these two opponents change. Since the Colley Matrix does consider each team’s record, the rankings do change. However, Alabama’s rank does not change. This makes no sense when a team loses to a cupcake.

To be precise, one can show the rankings remain exactly the same under certain changes of wins and losses. For an example in 2012,

  • Stanford beat Oregon
  • Oregon beat Washington
  • Washington beat Stanford

Suppose we change the result in each game.

  • Oregon beat Stanford
  • Washington beat Oregon
  • Stanford beat Washington

Since all teams have the same record, the rankings stay exactly the same. Oregon would remain 7th despite losing to an average Washington team. It just doesn’t make sense.

However, for the sake of simplicity, I went with the example in which one team traded a loss for a win. At the end of the day, Colley’s method disregards massive amounts of useful information.

Northern Illinois busts the BCS.

Before the MAC championship game in 2012, Kent State threatened to bust the BCS with their ranking of 17th. However, their opponent, Northern Illinois, wasn’t too far behind at 21st.

After winning the championship game, Northern Illinois jumped to 15th in the final rankings to earn a BCS bowl game against Florida State. In the computers, the Huskies made massive jumps in the polls of Richard Billingsley and Peter Wolfe.

Billingsley does not describe this ranking method, so no one knows why he bumped Northern Illinois from 19th to 12th.

However, Peter Wolfe describes his method and even offers a few references for his Bradley-Terry model. This nice academic article by Keener describes the model in some detail, making it possible to reproduce the results. After carefully reading this paper, I didn’t find any problems with the ranking method. It does matter that Kent State lost to Kentucky. The math seems to favor teams that play an extra game, which most likely helped Northern Illinois jump from 23rd to 12th after their win over Kent State.

A Playoff on the Horizon

I didn’t think anything could make me hate the BCS more. I was wrong.

At least the current computer polls will be banished when a four team playoff arrives in 2014. A selection committee similar to the group that determines the field for the NCAA men’s basketball tournament will pick the four teams. I only hope their debates will be aided by better algorithms for ranking teams.

Thanks for reading.

How To Use Simple Division to Evaluate Notre Dame at USC

By Leave a Comment

The BCS calculus for Notre Dame is simple. Beat USC on Saturday and play in the national championship game. Can the Fighting Irish do it?

Most of the headlines focus on Notre Dame’s defense. Led by linebacker Manti T’eo, the Fighting Irish have allowed 10.1 points per game, best in the nation. With the injury to USC quarterback Matt Barkley, Notre Dame’s defense will most likely have a good day.

But what about Notre Dame’s offense?

Notre Dame’s Offense

The raw statistics suggest Notre Dame has a poor offense. For example, the Fighting Irish pass for 211 yards per game, a number that includes negative yards from sacks. Their 74th ranking places them below average nationally.

However, total yards per game is a misleading statistic.

Yards per game depends on how many times a team throws the ball. Notre Dame doesn’t throw the ball as often as the spread offense of West Virginia. Using simple division, we divide total pass yards per game by the number of attempts to obtain a better metric of pass offense. Notre Dame averages 6.83 yards per pass attempt, 51st best in the nation.

Simple division turns a below average pass offense into an above average one.

Dividing total yards by the number of plays lets us account for the pace of the football game. It is analogous to the tempo free statistics that Dean Oliver introduced into basketball. Looking at points per possession shed a new light on offenses that didn’t fast break and looked for the best possible shot late in the shot clock. Yards per attempt is the first step in incorporating these ideas into football.

Match Up With USC’s Defense

At The Power Rank, we take yards per pass attempt and adjust it for strength of schedule. Since Notre Dame has played top pass defenses such as Stanford, Michigan State and Oklahoma, our algorithm bumps up their pass offense from 51st to 28th. Their rating of 7.04 gives a predicted yards per attempt against an average pass defense.

Below, we show how our adjusted numbers for Notre Dame’s offense match up with USC’s defense. A better defense has a blue dot further to the right. Then the unit with the dot further to the right is predicted to have the advantage in the match up. By overall yards per play, Notre Dame’s offense enjoys a slight advantage over USC’s disappointing defense.

The Power Rank shows how teams match up with data visualization.

These visuals appear on our interactive team pages, the heart of our premium college football product. For more information, click here.

Outlook

Our team rankings predict a very even game (USC by 0.3 points). However, these numbers reflect a USC team with quarterback Matt Barkley. The Vegas line has Notre Dame as a 6 point favorite.

Is the injury to Barkley worth 6 points? We’ll find out Saturday.

Thanks for reading.