Writers of Pro Football Prospectus 2008

Most Recent FO Features


» 2017 Adjusted Games Lost

Two NFC teams were hit hardest by injuries last year. One already set the AGL record in 2016, while the other has a coach with the worst AGL since 2002. Also: the Rams' incredible bill of health in L.A., and Tampa Bay's questionable injury reporting.

29 Aug 2005

Confessions of a Football Junkie: The Lights Go On

by Russell Levine

Sure, USC is gunning for its third consecutive national championship and quarterback Matt Leinart is the favorite for a second Heisman Trophy. But can Texas finally beat Oklahoma? Can Charlie Weis build a winner at Notre Dame? Will the Big East survive the loss of its strongest programs? And what's with this new poll in the BCS system? The 2005-06 college football season could be one of the strangest and most interesting in memory.

Coaching Carousel

Of the 23 Division I-A teams that changed coaches this offseason, the spotlight will shine brightest on Urban Meyer at Florida and Charlie Weis at Notre Dame. Meyer, coming off a 13-0 season at Utah, takes over one of the best programs of the past two decades from the fired Ron Zook. Meyer won at recruiting-challenged Utah, and that won't be a problem at Florida. Zook left the cupboard well-stocked, so the expectation is that Meyer will win right away.

Weis brings his three Super Bowl rings to Notre Dame. A Notre Dame grad, Weis will have to do quite a sales job to return the program to the recruiting elite, and those rings should help. But the short-term prospects are more muted thanks to a ridiculously difficult schedule and a talent-challenged roster.

Other new coaches expected to win right away include Les Miles at LSU, who arrives from Oklahoma State and inherits a national championship-caliber roster from Nick Saban, now coaching the Miami Dolphins. If Miles stumbles at all, his stay in Baton Rouge won't be a comfortable one.

The SEC's other new coach is South Carolina's Steve Spurrier, who returns to the conference after a failed stint in the NFL. His first South Carolina team will probably struggle, but that won't stop every press outlet in the country from sending a representative to the Florida- South Carolina game on November 12.

Not all the coaches on the hot seat are new; Joe Paterno has thus far resisted calls for his resignation after some truly dismal seasons, but his Nittany Lions need to turn things around before the program falls completely off the college football map. A few more mediocre seasons (by Florida State standards) could have the Seminoles' Bobby Bowden facing similar questions.

As the BCS Turns

For a time, it appeared this off-season could be the most tumultuous in the history of the BCS after the Associated Press demanded its poll no longer be used as part of the BCS formula. But BCS officials decided the status quo was preferable to a complete makeover of its selection process, and merely set about searching for a poll to replace the AP. Enter the Harris Interactive College Football Poll, which consists of 114 voting members spread evenly across all 11 Division I-A conferences and regions of the country. It simply subs for the AP; the BCS formula -- the Harris and USA Today coaches' polls will each count for one-third, with an average of six computer ratings accounting for the other third -- remains unchanged. The Harris won't release its first rankings until September 25 -- four weeks into the season -which helps to eliminate preseason poll bias, but it does nothing to solve the problem of having more than two worthy teams for the title game.

Play It Again

The biggest change on the field in college football this fall is the wide acceptance of the use of instant replay. After a successful trial in the Big Ten last season, it has been adopted by nine of the 11 Division I-A conferences. The system will vary from conference to conference, with availability dependent on television schedules in some of the smaller conferences. It will be used to review calls involving the boundaries, possession, and certain detectable infractions, such as too many men on the field or timing errors. If a head coach challenges a play that stands, the team loses a timeout and can no longer challenge a play for that half.

Heisman Dark Horses

Three of last year's four Heisman Trophy finalists -- Matt Leinart and Reggie Bush of USC and Oklahoma's Adrian Peterson -- have returned to campus, making them automatic candidates to capture the award this season. Others expected to contend include a pair of quarterbacks from major programs, Texas's Vince Young and Florida's Chris Leak; and two Juniors, running back Gerald Riggs, Jr. of Tennesse and receiver Ted Ginn, Jr. of Ohio State. But some of the nation's most exciting players are hiding out in smaller conferences and less-celebrated schools. Here are five dark horses you're not hearing about who could end up on stage in December after breakout seasons:

Brian Brohm, QB, Louisville Omar Jacobs, QB, Bowling Green Laurence Maroney, RB, Minnesota Reggie McNeal, QB, Texas A&M DeAngelo Williams, RB, Memphis

Other 'Can't Miss' Games

Are our conference games of the year not enough to satisfy your college football appetite? Here are another dozen you'll want to catch this fall:

  • Boise St. at Georgia, September 3
  • Miami at Florida St., September 5
  • Texas at Ohio St., September 10
  • Bowling Green at Boise St., September 21
  • Tennessee at LSU, September 24
  • USC at Arizona St., October 1
  • Iowa at Purdue, October 8
  • Michigan at Iowa, October 22
  • Auburn at LSU, October 22
  • Florida vs. Georgia (at Jacksonville), Oct. 29
  • Texas at Texas A&M, November 25
  • Florida St. at Florida, November 26


ATLANTIC: Boston College, Clemson, Florida St., Maryland, North Carolina St., Wake Forest
COASTAL: Duke, Georgia Tech, Miami, North Carolina, Virginia, Virginia Tech

Player of the Year: Devin Hester, CB/KR/PR, Miami
Watch Out for: Boston College
Falling Fast: North Carolina St.
Game of the Year: Miami at Virginia Tech, November 5

Unable to split its teams along geographic lines because of the heavy concentration of powers in the south, the ACC debuts as a two-tiered 12-team super conference this season with an "Atlantic" division and a "Coastal" division. Welcome to the most confusing conference in college football.

Boston College's arrival this year completes the league's two-year expansion. Much about the conference is new, including the inaugural title game on December 3, and participants could easily turn out to be two of the three Big East refugees (B.C. in the Atlantic, and Miami and Virginia Tech in the Coastal). The traditional power from the pre-expansion days, Florida State, suffered a tumultuous off-season and enters 2005 with question marks at quarterback, both lines, and the defensive secondary. Untested signal-caller Xavier Lee will get a baptism-by-fire against Miami on Labor Day, but at least the game is at home. If the 'Noles falter, Boston College, with all-world defensive end Mathias Kiwanuka could capture the division.

The entire Atlantic division is loaded with sleeper potential, with Clemson, Maryland, and North Carolina State all looking to rebound from disappointing seasons.

In the Coastal, Miami also has an unproven QB in Kyle Wright, but the experienced defense should be among the best in the nation, and CB/WR/KR Devin Hester is one of the nation's most exciting players. The Hurricanes' shocking home loss to Virginia Tech cost them the conference crown last season, and the Hokies are again a threat. Marcus Vick (Michael's younger brother) will start at quarterback after serving a year-long suspension. He showed flashes of brilliance two years ago, but has a long way to go to fill his brother's shoes.

Miami at Virginia Tech on November 5 could well decide the ACC title game berth. After those two schools, there's a significant dropoff to Georgia Tech, Virginia, and North Carolina in the Coastal.


NORTH: Colorado, Iowa St., Kansas, Kansas St., Nebraska, Missouri
SOUTH: Baylor, Oklahoma, Oklahoma St., Texas, Texas A&M, Texas Tech

Player of the Year: Adrian Peterson, RB, Oklahoma
Watch Out for: Texas A&M
Falling Fast: Oklahoma St.
Game of the Year: Oklahoma vs. Texas, October 8, Dallas

Based on the preseason polls, a changing of the guard has occurred in the Big 12 conference during the off-season. When they last met on the field, Oklahoma was putting the finishing touches on a fifth straight win over archrival Texas, a streak that until last year had kept the Longhorns from reaching a BCS bowl. But Texas quarterback Vince Young pulled a Michael Vick imitation in a Rose Bowl win over Michigan while Oklahoma was humbled by USC in the Orange Bowl.

OU was hit hard by the NFL draft and graduation, so most experts are pointing to Texas (no. 2 in both preseason polls) as the Big 12's team to beat. It's a sign of respect for Oklahoma coach Bob Stoops and sophomore tailback Adrian Peterson that the Sooners are ranked as high as they are (no. 7 AP; no. 5 USA Today). Conference supremacy will be decided at the Cotton Bowl in Dallas on October 8. This is not, however, a one-game season for the Longhorns, who face a titanic early-season test at Ohio State on September 10. Wins in both should write Texas a ticket to the National Championship game.

If the big favorites stumble, Texas A &M and dynamic quarterback Reggie McNeal appear ready to mount a challenge out of College Station.

In the weak-sister North Division, Nebraska is hoping for a major turnaround after a 5-6 season under Bill Callahan, whose West Coast offense should operate more smoothly with former junior college quarterback Zac Taylor under center. Colorado, the defending North champ, managed to hang on to coach Gary Barnett after riding out a serious recruiting scandal, and quarterback Joel Klatt's arm could be enough to keep the Buffaloes in the Big 12 title hunt in the North.

Big East

Cincinnati, Connecticut, Louisville, Pittsburgh, Rutgers, South Florida, Syracuse, West Virginia

Player of the Year: Brian Brohm, QB, Louisville
Watch Out for: West Virginia
Falling Fast: Cincinnati
Game of the Year: Pittsburgh at Louisville, November 3

If the ACC is the most confusing conference in college football, the Big East is the most diminished. Having lost Miami, Virginia Tech, and Boston College to the ACC over two years, the Big East is a league in transition. Temple was booted for ineptitude, leaving just Pittsburgh, Rutgers, Syracuse, and West Virginia as the league's founding members -- not exactly a murderer's row. This season, they are joined by Cincinnati, Connecticut, Louisville, and South Florida -- schools that, with the exception of Louisville, do more to help the conference's standing in basketball than in football.

Louisville, which is stepping up from Conference USA, has to be considered the favorite to capture the Big East's BCS berth. The Cardinals were 11-1 last season behind quarterback Stefan LeFors, who's now suiting up for the Carolina Panthers. He is replaced by the much-heralded Brian Brohm, who proved his mettle by nearly engineering a second-half upset of Miami as a freshman last season. The Cardinals should be able to fend off Pitt, Syracuse, and West Virginia to capture the league, and could conceivably go undefeated. If that happens, cue the BCS controversy, because schedule strength is likely to keep the Cardinals below even one-loss teams from better leagues.

Pitt, under former Dolphins and Bears coach Dave Wannstedt, is the defending conference champ and still has quarterback Tyler Palko, but the young defense is shaky at best. West Virginia suffered some key losses to graduation, but should be able to run the ball and still has a stout defense. Syracuse's new coach -- former NFL and Texas defensive coordinator Greg Robinson -- should improve the Orange's weak defense, but there's too much uncertainty on their offense to challenge for the crown. Meanwhile, Rutgers is still Rutgers (stuck in its endless rebuilding cycle), and Cincinnati, Connecticut, and South Florida will all struggle to post winning records.

Big Ten

Illinois, Indiana, Iowa, Michigan, Michigan St., Minnesota, Northwestern, Ohio St., Penn St., Purdue, Wisconsin

Player of the Year: Drew Tate, QB, Iowa
Watch Out for: Penn St.
Falling Fast: Wisconsin
Game of the Year: Ohio St. at Michigan, November 19

With three teams (Michigan, Ohio State, and Iowa) entering the season in the preseason top 11 in both polls, the Big Ten, that bastion of three-yards-and-a-cloud-of-dust football, has been reborn. It is heavy with breathtaking talent at the skill positions, including sublime all-purpose men Steve Breaston at Michigan and Ohio State's Ted Ginn Jr., and innovative coaches, including the man many consider to be the nation's best, Kirk Ferentz of Iowa. Then there's Purdue, which operates an offensive attack described as "basketball on grass" by coach Joe Tiller and could bump the favorites and claim the conference's BCS berth.

Michigan showed plenty of offensive flash last year, and despite the departure of receiver Braylon Edwards, the Wolverines could be even better on that side of the ball. Quarterback Chad Henne and running back Mike Hart were true freshmen last season and should only improve. Breaston and Jason Avant should soften the loss of Edwards, but it's Michigan's defense, shredded in season-ending losses to Ohio State and Texas, that could keep the Wolverines from a third-straight trip to Pasadena. If it does, it will likely be because of an inability to contain mobile quarterbacks such as Ohio State's Troy Smith, Iowa's Drew Tate, and Michigan State's Drew Stanton.

Ohio State, led by Smith and Ginn on offense and the nation's best linebacking corps on defense, has a difficult schedule, including a September 10 date with Texas and a road date at Michigan in November. Iowa gets Michigan at home but must travel to Ohio State. If those three beat each other up, Purdue, with 20 returning starters and no Michigan or Ohio State on the schedule, could sneak in to win the conference. Otherwise, look out for Penn State, which could finally emerge from its recent run of mediocrity if the offense improves under quarterback Michael Robinson.


Arizona, Arizona St., California, Oregon, Oregon St., Stanford, UCLA, USC, Washington, Washington St.

Player of the Year: Matt Leinart, QB, USC
Watch Out for: Arizona
Falling Fast: Oregon St.
Game of the Year: USC at Cal, November 12

USC enters the season as one of the biggest favorites in college football history. Consider that the Trojans likely would have received a majority of first-place votes in the preseason polls even if Heisman-winning quarterback Matt Leinart had not opted to return for his senior year. USC has lost one game since October 2002, and coach Pete Carroll has just reeled in yet another top recruiting class.

So can the Trojans be beaten? Of course they can. In fact, with road trips to Cal, Arizona State, Oregon, and Notre Dame, it wouldn't be a shock. Add in the departures of offensive coordinator extraordinaire Norm Chow and defensive line coach Ed Orgeron, and USC might not be as dominant as it appears on paper. It is difficult, though, to picture anyone or anything slowing down the Trojan offense, with Leinart, do-everything back Reggie Bush, sledgehammer LenDale White, and a pair of NFL-caliber receivers in Dwayne Jarrett and Steve Smith.

If anyone can challenge USC in the Pac 10, it might be Arizona State, which also features an explosive offense behind quarterback Sam Keller and another outstanding receiver duo in Derek Hagan and Zach Miller. California, despite the loss of standout quarterback Aaron Rodgers and 2,000-yard back J.J. Arrington, could also contend. Joseph Ayoob, like Rodgers a junior college transfer, will start under center, but the real excitement is over Arrington's replacement, sophomore Marshawn Lynch, who averaged 8.8 yards per carry in limited duty last season.

Oregon looks to be rebuilding, but could still pull off the surprise win -- even against USC -- playing in front of one of the nation's most hostile crowds at Autzen Stadium. Other sleepers are Arizona, which should be much improved in its second year under defensive specialist Mike Stoops, and UCLA, where coach Karl Dorrell, under pressure to produce, reeled in highly touted quarterback Ben Olson.


EAST: Florida, Georgia, Kentucky, South Carolina, Tennessee, Vanderbilt
WEST: Alabama, Arkansas, Auburn, LSU, Mississippi, Mississippi St.

Player of the Year: Gerald Riggs Jr., RB, Tennessee
Watch Out for: Alabama
Falling Fast: South Carolina
Game of the Year: Tennessee at Florida, September 17

All the usual suspects -- Auburn, Florida, Georgia, LSU, and Tennessee -- are expected to contend, but this is a league in flux thanks to an exodus of stars to the NFL and an influx of new coaches.

No team lost more talent than defending conference champion Auburn. But the departures of Jason Campbell, Ronnie Brown, and Cadillac Williams doesn't remove the Tigers from the SEC race in 2005. Auburn's offense should still be explosive with Brandon Cox throwing to Courtney Taylor and Ben Obomanu. The schedule sets up nicely with five straight home games before season-defining road trips to LSU and Georgia.

Georgia also lost talent to the NFL, but the offense is still potent behind experienced quarterback D.J. Shockley and tailbacks Thomas Brown and Danny Ware. Like most SEC contenders, Georgia's fate will be determined by its round-robin games against the other conference powers.

Florida is loaded with talent recruited by ex-coach Ron Zook, and new coach Urban Meyer is eager to use it in his innovative spread-option attack. If quarterback Chris Leak can master the complex system, the Gators could be national-title contenders. LSU is in a similar situation, with a team still stocked with Nick Saban recruits who will now play under former Oklahoma State coach Les Miles. It may take a year or two before the league's other new coach, South Carolina's Steve Spurrier, can make an impact in the conference standings; his Gamecocks are rebuilding.

Underappreciated Phillip Fulmer of Tennessee may finally get some respect if his Volunteers meet expectations, which are sky-high thanks to an explosive offense led by the quarterback tandem of Erik Ainge and Rick Clausen, as well as a defense that returns eight starters. It would be a shock if the title game participants came from outside the big five. If anyone could sneak in, it's Alabama in the West, so long as Brodie Croyle returns from injury as expected.


Conference-USA, MAC, Mountain West, Sun Belt, WAC, Independents

Player of the Year: Omar Jacobs, QB, Bowling Green (MAC) and DeAngelo Williams, RB, Memphis (C-USA)
Watch Out for: Miami (Ohio)
Falling Fast: Temple
Game of the Year: Boise St. at Fresno St., November 10

Will there be a Utah in 2005? Last season, the Utes became the first team from the non-power conferences to capture a BCS berth, going 12-0 and beating Big East champ Pittsburgh by four touchdowns in the Fiesta Bowl. The Utes will again be strong, but a repeat performance is unlikely with QB Alex Smith now in the NFL and coach Urban Meyer in Gainesville. Other teams from non- BCS leagues that could challenge for BCS berths are WAC powers Fresno State and Boise State. Boise has won 25 straight games on the "Smurf Turf," its distinctive blue home field, but will face its toughest tests on the road this season at Georgia and at Fresno.

Another possibility is MAC stalwart Bowling Green, which is led by Heisman-caliber quarterback Omar Jacobs. The Falcons test themselves early with road games at Wisconsin and Boise. Of the independents, Notre Dame will be better under first-year coach Charlie Weis, the former offensive coordinator of the New England Patriots. But the Irish schedule, which includes dates against three of the preseason top four (USC, Tennessee, and Michigan), plus games against Purdue and Pittsburgh as well as two West Coast trips to Stanford and Washington, will be too much to overcome. Weis knows he needs to do two things to return Notre Dame to championship contention: recruit, and lighten up the schedule.

This article was originally published in Thursday's edition of the New York Sun.

Posted by: Russell Levine on 29 Aug 2005

35 comments, Last at 31 Aug 2005, 7:18pm by Tarrant


by B (not verified) :: Mon, 08/29/2005 - 4:14pm

"Marcus Vick (Michael’s younger brother) will start at quarterback after serving a year-long suspension. He showed flashes of brilliance two years ago, but has a long way to go to fill his brother’s shoes."
Is it really that hard to throw the ball at your receivers' feet? Or maybe he just needs a a nickname to match his older brother's. I suggest Ryan Mexico.

by Russell (not verified) :: Mon, 08/29/2005 - 5:11pm

B: I was referring moer to Ron Mexico's college career, when he pretty much carried that team to the national title game as a redshirt freshman.

by Tarrant (not verified) :: Mon, 08/29/2005 - 5:43pm

I don't think this "Harris poll" starting on the 25th will matter at all. What are most of the voters going to use as a guideline? Does anyone really think they're going to start at zero and rank teams from scratch 4 weeks into the season? Of course not. They're going to look at the AP and/or Coaches' polls and maybe make a slight adjustment or two.

As for USC - I think the Oregon game is the real landmine. USC hasn't had to play Oregon, home or away, in years - like since 2001 or something. Autzen is a horrific place to play, and I think if the loss is coming, that's where it will be. Maybe Cal, but I think they're replacing so many people that Oregon will be the tougher away game.

Louisville will be the new Miami of the Big East - almost sleepwalking its way to the title and BCS berth.

Despite my belief that Texas is the better team, I think Oklahoma will again beat Texas this year - unless Texas beats Ohio State. Sometimes you just get the feeling that one team has it out for another and there's a mental barrier that is just very hard to overcome. If there's any year that Texas can break that barrier, it's this one, though, if they beat Ohio State.


by Pittsburgh Phil (not verified) :: Mon, 08/29/2005 - 6:10pm

Despite mounting popular opinion, I really don't think the Big East is going to be a cakewalk for Louisville.

They had a great season in 2004, and they played Miami close, but they had a weak schedule. Big East teams will play better D than the likes of Houston and Tulane.

I still think they're the favorite to win the Big East, but I certainly wouldn't hand them that BCS berth just yet.

by Peyton Manning (not verified) :: Mon, 08/29/2005 - 6:14pm

>Sometimes you just get the feeling that one team has it out for another and there’s a mental barrier that is just very hard to overcome.

You get that feeling too?

by Dennis (not verified) :: Mon, 08/29/2005 - 6:15pm

Am I still the only who thinks the BCS needs to get rid of the "human factor" rather than keep adding more of it because they don't like what the facts say? Get rid of the polls and just use the computer rankings and you eliminate the biases. But you can't do that, because when people disagree with what the computers (ie the facts) say, then obivously the computers are wrong, because people are always right.

by noahpoah (not verified) :: Mon, 08/29/2005 - 6:22pm

I suggest Ryan Mexico.

Nah, too bland. How about Tony Honduras?

by zip (not verified) :: Mon, 08/29/2005 - 6:27pm


Coming up with a mathematical representation of which college football teams are the best is a difficult if not impossible process. Imagine if the two teams that met in the superbowl were determined by DVOA -- that's essentially what you're asking the BCS to do. And a DVOA-like stat would be even harder to evaluate college teams, since they have so few common opponents and there is a bigger disparity from the top to bottom.

Also, I love how everyone says "the computers" about the BCS, as if there's some group of PCs that have opinions on college football. It's just a formula, written by humans.

by Vinny (not verified) :: Mon, 08/29/2005 - 7:02pm

Russ, two points of contention. First, I don't think Greg Robinson can improve anything other than maybe the traffic situation outside the Carrier Dome. I have no idea how he got that job.

Along those lines, the only offense that Michael Robinson can succesfully run requires an X-Box.

by zlionsfan (not verified) :: Mon, 08/29/2005 - 7:14pm

But at least the formulas are consistent throughout the year. They don't change based on voters' whims. Mack Brown can complain all he wants and it still won't change the properties of mathematics.

That's still not a good enough solution for me. There aren't enough games in the season to decide the top two teams in the country, no matter how you do it, unless you have a p ... a p ... darn NCAA censors!

Autzen is a very tough place to play, which makes IU's victory there last year that much more impressive. I remember watching it in a bar somewhere and doing a double-take when they first flashed the score (it was 23-0 Indiana at halftime). Even as a Purdue grad, I thought it was impressive. Oregon cut the lead to 13, kicked off ... and 97 yards later, IU was up by 20.

It seems like the weather wasn't that good, but the recap on Yahoo doesn't mention it.

by Vinny (not verified) :: Mon, 08/29/2005 - 7:16pm

Remembering the highlights from that game, the weather looked fine to me.

by Tarrant (not verified) :: Mon, 08/29/2005 - 7:24pm

The computers can only get a good representation of what team is better with a significant sample size - 11 or 12 games is not usually enough.

One can say "Humans have biases!" however, football is, more than any other sport, a game of intangibles. The computers don't take into account injuries, or how many yards a team has gained, etc. Or that one team's close game happened in bad weather on the opponent's field while another team's was at home. All the computers take into account are wins and losses - which would be great if you were choosing 8 teams, but you aren't, you're choosing two. Even something as complex as DVOA, which does take into account some of these things, can't really get ahold of the intangibles, those things that make you say "This team is better."

The computers said Oklahoma was, far and away, the best team in the nation in 2003 - so much so that getting blown out in their last game didn't even displace them from #1 by a wide margin. They lost, and the game wasn't as close as the score indicated. The computers also said Oklahoma was a solid #1 last year - they got blown clear out of the water. They said Nebraska, not Oregon, was the team in 2001, etc.

Noting this, one can point out that the computers have been directly responsible for almost all the "BCS disasters" since the inception of the system - cases where one team that clearly didn't seem to deserve it made it into the title game, and was (more often than not) blown out.

In a case like this, only people can really look at the teams and really get a good idea of which teams are superior.

There's also another factor - since there's no playoff, the championship system is only credible if the general college football public believes in it. The general college football populace distusts the computers because the computers have blown it multiple times in the past few years - they BELIEVE in the polls, since the polls have "worked" (used loosely) for 100 years or so. That's why there's zero controversy if the polls override the computers, but lots of the computers override the polls. The BCS is shaky enough as it is - they have to operate in such a way that people maintain faith that their champion is "the champion" (which is why they couldn't just tweak the system after 2003's split title - they HAD to completely revamp it, even if they didn't want to, because the public believed it failed).

The problem with the polls is that the people responsible for doing them, don't look at the teams necessarily - reporters are too busy reporting, and coaches are too busy coaching.

But I'd say that, since the start of the BCS, the humans have gotten it right (although the teams the humans chose didn't always get in to the game) much more often than the computers.


by Adam (not verified) :: Mon, 08/29/2005 - 7:35pm

Where, exactly, is Temple falling FROM?

by Pat (not verified) :: Mon, 08/29/2005 - 7:41pm

Otherwise, look out for Penn State, which could finally emerge from its recent run of mediocrity if the offense improves under quarterback Michael Robinson.

Bleah. I'd feel much more comfortable if there was another QB. Robinson just never lived up to any potential at QB. Their previous QB, Mills, was the most inconsistent thing I've ever seen in my life. I'm convinced it was his nerves, because when he was on, he was on fire in 2002. If it hadn't been for a crap pass interference call, he might've even been able to pull off a comeback versus Ohio State that year.

But Robinson was always inserted as a QB during the Mills era to do one thing - scramble. Guy never could pass worth a damn.

If Penn State's offense does well this year, that says a heckuva lot more about Williams and King than it does Robinson, in my opinion.

by Dennis (not verified) :: Mon, 08/29/2005 - 10:32pm

Re #12: I disagree that the computers have blown it. I think they have done as good a job as you can do picking 2 teams that - as others have said - have a limited sample size and probably don't even have an common opponents. And how do you know that the computer rankings only factor in wins and losses? As far as I know, the formulas haven't been made public. So let's make them public, and if the majority doesn't like them, then let's make some that most people can agree. Then we'll have clear objective criteria up front and if people don't like the results, they'll have nothing to complain about.

As an aside, I still haven't heard a good reason why I-AA and every other division can have a playoff but not I-A (and losing the bowl money isn't a good reason IMO). My suggestion is a 16 team playoff with the winners of the 10? conferences and 6 at-large teams. Play the first 3 rounds on campus with the championship at a predetermined sites. Non-qualifiers and teams that lose in the first round (maybe even the second) could still get bowl bids. But it makes sense so it will never happen.

by pm (not verified) :: Mon, 08/29/2005 - 10:49pm

Ted Ginn Jr. is a sophmore not a junior. I don't know why ohio State doesn't try Ted Ginn at CB more. With his speed and size he could be a charles woodson clone and would greatly help his heisman chances.

by Brian (not verified) :: Tue, 08/30/2005 - 12:53am

"I suggest Ryan Mexico.

Nah, too bland. How about Tony Honduras? "

The official name of the BlogPoll is "Ron New Mexico".


by Tarrant (not verified) :: Tue, 08/30/2005 - 10:04am

Re: #15

We know that the computer rankings take into account primarily wins and losses because the BCS tells us so - it puts limits on what the computers are allowed to do, and this is one of them. They rank based on wins, losses, and strength of schedule (which can be calculated only based on wins and losses, and cannot take score, margin of victory/loss, home/away, etc. into account).

Now that doesn't mean that every computer calculates SoS the same way - in fact, the way a given computer calculates SoS, and gives "bonuses" or "penalties" for playing high/low-SoS opponents, or being in a high/low-SoS conference, is where most of the differences between the various computer rankings come from.

The computers are NOT allowed to take into account scoring of any form, yards, total offense, etc.

Sagarin, for example, thinks this is stupid, and during the season, releases two rankings each week - one using "BCS numbers", and one using his "real formula". The two are often quite different, and his "real numbers" are usually a bit more accurate than the "BCS numbers".

Note that "originally", BCS computers WERE allowed to take scoring into account, but teams were leaving starters in all game and wildly running up the score to get extra BCS points, so "margin of victory' was capped at 21. That led to teams getting rid of all potentially "tough" nonconference matchups, so they could make sure to always win by 21. Thus any computer rank due to scoring was forbidden entirely by the BCS.

As to the computers "doing a good a job as any". I disagree - if they did, then there wouldn't be the problems we have now. The fact is that 3 or so out of the 7 BCS years, the computers "vetoed" a team that the polls would have selected for the title game, and in each case, the team pushed out won its game convincingly, and the team the computers put in was blown out of the water. That's circumstantial evidence, sure, but it's the only kind we have here, without a playoff. The only reason nothing was changed until after 2003 was because the OTHER times it happened, the #1 team always won, so no big deal (after all, it's always assumed that if the polls' #1 and #2 played, the #1 would win anyway). In 2003, it was the #1 team left out, a split title resulted, and the system had to be changed so the #1 team (at the minimum) could never be left out again.

The current system essentially says that if both polls agree 100% on the two title-game participants, it would be difficult if not impossible for the computers to override their selection. If the polls DISAGREE, then the computers break the tie.

Remember that since there's no playoff (and you are correct, there's no good reason to not have a playoff), the system of selecting the mythical national champion only works if the general football public believes it is valid. The general football public DOES NOT BELIEVE THE COMPUTERS ARE AS VALID AS THE POLLS because the computers have never 'vetoed' a matchup and ended up with a BETTER one, only the reverse - and thus the computers had to be pushed aside - their 'veto' had to be removed, relegated to a tiebreaker of sorts, or the BCS would have simply collapsed (and they knew it, hence the huge changes made for 2004).


by Carl (not verified) :: Tue, 08/30/2005 - 11:36am

I would like to make a point for Markov chain models, but fear the backlash. Anyway, Russell, I've been sent a copy of the Notre Dame press guide for the upcoming season. If you want it, let me know and I'll send it out to you.

by mistamaxwell (not verified) :: Tue, 08/30/2005 - 11:50am

The general public doesn't believe the computer polls are valid because they can't understand them. What I think they should do, in lieu of the obvious playoff, is make most BCS teams get there via a 'conference-championship style game'. For instance, there's always a team like Boise or Utah that's trying to crack the BCS, why not let them playoff against the Big East champ to get there? It's only fair since the Big 12, ACC, SEC, etc make their schools playoff in a championship game that provides no benefit to the teams playing the game other than money being made for the schools. This way the Big East can keep it's BCS standing, the mid-majors can get their BCS standing, the schools from the major conferences aren't punished for playing a conference championship, and we get to see more competitive football. Of course, a playoff is much simpler, better and there really is NO reason that there isn't one all ready.

by Vinny (not verified) :: Tue, 08/30/2005 - 11:59am

Re: 16

OSU tried Ginn at DB early last year but felt he really couldn't learn the defense fast enough. So, along with having him return kicks, they put him at WR and designed a few simple plays for him. It was the quickest way to get him on the field. And it paid huge dividends.

Seems like Senator Tressell finally started to depart from his conservative ways on offense half way through the season and that's when OSU really hit its stride. Should be interesting to see if he picks up where he left off or if he gets conservative again.

by Russell (not verified) :: Tue, 08/30/2005 - 12:32pm

Carl, you know what, I need something to elevate my computer monitor a couple inches. I think that ND guide might come in handy :)

by Domer (not verified) :: Tue, 08/30/2005 - 12:36pm


We'll talk after Sept 10.

Good day, sir.

by Tarrant (not verified) :: Tue, 08/30/2005 - 1:16pm

While it's true the public doesn't understand them, the public can look at them and say "They were wrong this many times." They can say "In 2001, the polls got it right and the computers were wrong." It doesn't matter if the result had the polls gotten their way would have been similar (say, still a blowout) - the computers chose a team that everyone who watched the teams play knew was inferior, because football is more than just a game of raw numbers - there are lots of times where a computer might say "Team A is better than B" but anyone that watches those two teams play can say "No, B is definitely better."

They haven't generally said that about the polls since the inception of the BCS.

I think it would be better for the computers if the BCS rankings were put off by another few weeks. One of the things people always make fun of, computer-ranking-wise, is the anomalies that crop up early in the season due to not enough games having been played - where you get a middling team that played 3 nonconference nobodies ranked #6 by one (but rarely all) of the computers because that particular formula treated something they did differently than the rest. Or where a team is wildly underranked. By 2/3 of the way through the season, those things have usually worked themselves out.

But when a team that is #3 in, say, 5 computers is #11 in the 6th, people will look at that and say "The computers make no sense." You and I know that the computer is simply spitting out a result based on the formula put into it, but stuff like that leads to further distrust of the computers vs. the polls. Additionally, when you have results like the last two years - Oklahoma a solid #1 in all the computers, but lower-ranked than their opponents in the polls, yet losing both years, and BADLY last year, tends to reinforce people's opinions that the computers are flat-out wrong and shouldn't get their way.


by Carl (not verified) :: Tue, 08/30/2005 - 5:04pm


To which I say, "It's not my fault the general public doesn't understand the Perron-Frobenius Theorem for Irreducible Matrices or the sublime beauty of the Markov Chain."

The important thing to remember is that what "the computers" tell us isn't always demonstrated correctly to the public. If you're using a dynamic programming model, you should be discussing probabilities. It's not as if Oklahoma is always going to beat USC, but the model "predicts" that it's more likely than not, given the statistical information we have in the dynamic model, that they will, say, six out of 11 times.

You're estimating win probabilities, not guaranteeing future success.

Personally, I'm a big fan of using logistic regression to plot the transition probabilities of a Markov Chain.

Vive Callaghan, Porter and Mucha!
Add in a touch of homefield advantage, a pinch of "margin of victory," some SoS and start cooking!

You see, I don't like binaries. Never have. Too reductive. Besides, I like to fire up the SPSS!

This method of predicting winners has been more successful than the writers' and coaches' polls and even the Vegas line than any other methods for college and NFL football and NCAA and NBA basketball.

Plus, I like it!

Go team!

by Carl (not verified) :: Tue, 08/30/2005 - 5:11pm

Another thing. Injuries. You could control for injuries in the regression by simply determining which injuries, at which positions, mean what to a team's success or failure.

Heck, we do this. So do several NFL teams.

So when State U loses its QB for the FATNHAPPY.COM BOWL, and their opponent State Tech goes without their TE, you can compute for (predicted) effect.

by Tarrant (not verified) :: Tue, 08/30/2005 - 5:26pm

Of course, that is the exciting thing about football - the single-game nature of it, that makes many of the statistical numbers less relevent. In the NBA or MLB, with 7-game playoff series, the better team can lose game 1 but will usually come back to win the series.

Statistics in 2002 would probably have said that the Rams would have beaten the Patriots 6-7 times out of 10. But the game was only played once, and the Patriots won the Super Bowl. Similarly, OU was favored by oddsmakers, columnists, and computers in both 2003 and 2004, but lost both times. If played 10 times, they might win the majority (although last year I doubt it).

Some of the BCS computer rankings go beyond a simple rank (even if they only turn into the BCS a 1-25 number), and some of them are very interesting and can indeed give more of a probability number.

We don't have a lot of data points, unfortunately - we have three or so instances where the polls and computers disagreed on who the top two teams were, the computers "won" the disagreement, and the computer-chosen team got blown out. That does not an overwhelming argument make - but when your entire championship model relies on choosing two teams from a pool of 117 and then arbitrarily naming one (or two) mythical national champions, rather than having a playoff, the public has to believe that those champions are actually champions, and that the team(s) left out were not more deserving than the participants - and the computers are not going to ever effectively convince them.

Perception, in such a case, is reality. We both know that, Carl.


by Carl (not verified) :: Tue, 08/30/2005 - 5:27pm

And one last thing, Russell. You forgot Navy.

I can see this team going 11-0. Their toughest test, I think, is the opening day tilt against Maryland.

Since the Terps weren't exactly stellar overachievers last year, I think the Midshipmen might sneak out a win at home.

They follow that up with a Who's Who of recently dysfunctional Division I squads -- Stanford, Duke, Kent St., Rice, Rutgers, Tulane (oy, the alma mater), Notre Dame, Temple and, of course, Army.

More brainpower than you would find in the Ivy League (sorry, FOs), but not exactly the top tier of the Big 10.

I know they lost a lot on defense, but it doesn't seem completely improbable that Navy (9-2 last year) could beat all these scrubs. Even with only six returning lettermen!

The problem is that Navy is a proven winner with a system against which elite teams don't want to play. The Midshipmen can't help it if no one wants to take the field against them.

Go Navy!

by zlionsfan (not verified) :: Tue, 08/30/2005 - 6:09pm

Let's face it - there is no good way to choose the "best" two teams from a group of 110+ that play outside their conferences 3-4 times per season (in most cases).

I prefer the computer rankings because I know (well, I think I know) that the systems are being applied equally to all teams being considered, whereas to my knowledge, few of the coaches and writers who vote in their polls are able to rate all teams based on knowledge.

However, I still believe that the best solution is to have a 16-team playoff for I-A - all conference champs and the next best teams, from whatever system you like. I think it's pretty unlikely that the second-most-qualified team in the country would not be included in the five at-large spots that would be available.

Yes, I want to include all conference champions. No team plays as well as a team that doesn't get the chance to play.

by a-dam (not verified) :: Tue, 08/30/2005 - 7:40pm

looks like tmq no longer has the only all-haiku football preview. it seems espn's ivan maisel has taken a little uncreative license...

by Tarrant (not verified) :: Tue, 08/30/2005 - 7:55pm

Well, TMQ has only said that he's the only all-haiku NFL predictions. Maisel's are college :)


by ElJefe (not verified) :: Wed, 08/31/2005 - 3:24pm

A little late to the party, but ...

What seems to be missing in any explanation of the BCS is that, really, it should almost never "work". The only situation that won't generate controversy is if 2 (and only 2) teams from the major conferences finish the season undefeated (Utah and Louisville need not apply). In that scenario any of the frills of the BCS are unnecessary since the humans will inevitably rank those two teams #1 and #2. The BCS "works" in this case because it removes the bowl tie-ins that may have prevented those two teams from meeting. But anything else and the BCS (and most likely the computers) are going to get criticized.

Another thing that goes unsaid is that although the computer rankings are superficially unbiased, most are created by a statistician with an axe to grind. I seem to remember that of the computer rankings used initially by the BCS, one (Sagarin?) was created to prove Washington was the real national champion in 1991 (instead of Miami), and another was designed to show Michigan to be the supreme team in 1997. There's a lot of variables and weightings available to the statistician to indulge his own beliefs.

The primary reason the human pollsters haven't "screwed up" is that before the BCS they really didn't have the chance. Without the BCS system, Auburn, Oklahoma and USC would have all played in separate bowl games last year, and in the previous year OU, LSU and USC would also have played in separate bowl games. The other game cited as a computer screw-up (Miami vs. Nebraska) would have still taken place, just in the Orange Bowl instead of the Rose.

It just happens that there have been a couple of lousy title games in last few years, and the human pollsters formula of (rank by losses), (if same number of losses, rank by earlier date of defeat), (if same date, rank in order of pre-season ranking, based on results with almost entirely different set of players and opponents) would have worked better. What is missing from the computer rankings is the human bias of "what have you done for me lately".

Without a playoff, college football is doomed to controversy more years than not. And I think the folks who sell the rights to the games like it that way.

by Sophandros (not verified) :: Wed, 08/31/2005 - 3:25pm


The Alma Mater beat Navy 42-10 last season, and looked stronger before Katrina disrupted life.

BTW, can anyone tell me how the computers determine their preseason and week one rankings?

by ElJefe (not verified) :: Wed, 08/31/2005 - 3:58pm

Re: Division I-A Playoff

I think no one is willing to articulate why they oppose a playoff system, but I figure I can openly speculate. :)

I think coaches oppose a playoff system due to job security. There are ~50 postseason spots available under the current bowl system, and a lot of alums can be made happy with a trip to Shreveport after a 7-4 season. But f a playoff system were in place a lot of fairly successful coaches would never sniff the postseason and a lot of heads would roll.

If you wanted a 16-team playoff, you would have to reserve some of those spots for the minor conferences in order to get them to participate (assuming the NCAA is actually a democracy). That would likely lead to a dramatic redistribution of the revenues in college football. If the MAC gets a spot, they should get 1/16th of the money. If they don't get a guaranteed spot, they won't vote for a playoff. That thought alone would throw a wet blanket on all the administrators at football power schools.

I think a playoff would be worth even more money than the bowl system, and that might scare administrators. Look at the number of arrests/expulsions/otherwise anti-social behavior that occurs at major college football programs. Jack up the dollars and make the difference between 8-3 and 10-1 much much larger, there is going to be a lot of pressure to be even more lenient in admissions and enforcement to get/keep good players on the field. This hasn't happened in college basketball because you only need to let 3-4 players a year into the college/university that otherwise wouldn't have been accepted. There are enough assistant coaches and student hangers-on that you can fairly effectively supervise those players. Sitting on 8 basketball players is a lot easier than sitting on 80 football players.

The players themselves probably aren't in favor of it. Generally, what distinguishes a I-AA (II, III) player from a I-A player is their future profession. Playing 4 extra games and risking injury isn't going to dramatically effect your future career as an actuary. But if you aspire to the NFL .... Willis MacGahee is the exception.

Notice none of this has anything to do with the burdens of being a "student"-athlete. That line of reasoning is pure BS.

by Tarrant (not verified) :: Wed, 08/31/2005 - 7:18pm

I don't know if it was Sagarin's, but I definitely recall one of the original BCS computer rankings, run by I think it was the Seattle Times, for years vastly overranked the Washington schools. I think their excuse was that their formula took into account the weather or something (and of course, it's almost always raining in Seattle).

The New York Times computer had a tendency to give the benefit of the doubt to Notre Dame, etc.

Since none of the BCS computer ranking people (save one) have ever been willing to make their formulas public, you're correct that it isn't out of the realm of possibility that the formulas themselves have internal biases that we don't know.

Furthermore, such biases don't even have to be things like "IF (team == "Washington") {rank++;}". It can simply do things like calculate SoS in a manner that benefits cetain areas of the country, or give greater weight to some different things - like in the above case, weather (or so they said - note that they aren't part of the BCS anymore and never did make the formula public).