Writers of Pro Football Prospectus 2008

Most Recent FO Features

KerriganRya11.jpg

» Week 2 DVOA Ratings

Stomping the Jags leaves Washington No. 2 behind only Denver. But what can we really learn from one big win early in the season, before we are applying opponent adjustments?

25 Sep 2006

Confessions of a Football Junkie: Assume at Your Own Risk

by Russell Levine

A common observation coming out of Notre Dame's miraculous 40-37 win over Michigan State Saturday night was that the Irish had just pocketed a BCS at-large bid. Notre Dame stands at 3-1, and has seven straight games against unranked opponents before playing USC to end the season. A 10-win campaign should get the Irish into the BCS, no matter what happens against USC.

But relying on assumptions can be a dangerous game in college football this season. The weekend's action was devoid of upsets, but that doesn't mean everything went as planned. Just ask the Georgia Bulldogs. Undefeated and ranked no. 9 in the AP poll, Georgia was a four-touchdown favorite at home against a rebuilding Colorado team. Colorado had lost its home opener to Montana State, a middling Division I-AA program. The Buffaloes had just 23 points in three games, and were facing a Georgia defense coming off back-to-back shutouts.

But what looked like a laugher turned into a nail-biter, as Georgia trailed 13-0 after three quarters before rallying for a 14-13 win on a touchdown in the final minute.

So while it's presumed that no. 1 Ohio State will remain undefeated until it faces Michigan if the Buckeyes can beat Iowa on the road this week, one would be wise not to overlook an October road date at Michigan State, for example.

No such assumptions are made about the nation's no. 2 team, Auburn, which still must face both Florida and Georgia at home before a season-ending road date at Alabama. The same can be said of fifth-ranked Florida, which has trips to LSU and Florida State as well as the annual neutral-site game against Georgia still to come in the SEC's death march of a schedule.

USC is ranked third, and is a popular choice to play in the BCS championship game precisely because it's believed that only the aforementioned Notre Dame game November 25 stands between the Trojans and a 12–0 record. Oregon and Cal, the latter now fully recovered from its week one meltdown at Tennessee, would beg to differ. USC will host the Ducks and Bears on back-to-back weekends before playing Notre Dame.

Predicting an undefeated season for fourth-ranked West Virginia may be the safest bet, thanks to the lack of depth in the Big East, but the perceived weakness of its schedule may be precisely what keeps the Mountaineers from moving up the polls. West Virginia had a ho-hum 27–20 win over East Carolina on Saturday, the type of relatively unimpressive result that the Mountaineers must avoid if they are to win favor with the pollsters and the computers that will determine their BCS chances.

So which of the top teams has the best chance to avoid the land mines on its schedule and end up playing for the national championship in Glendale, Ariz., on January 8?

The answer is Ohio State. The Buckeyes' biggest concern entering the season -- a defense with only two returning starters -- has become a strength in just four games. Ohio State returned a pair of interceptions for touchdowns in the final three minutes to turn a tight game with Penn State into a laugher, and its defense has already dominated then-no. 2 Texas in a road win two weeks ago. On offense, quarterback Troy Smith has emerged as the leading Heisman contender of the early season; his whirling, 37-yard touchdown pass against the Nittany Lions will surely be shown on all the season-ending highlight reels. And the Buckeyes have found another go-to receiver in Anthony Gonzalez to complement their game-breaker, Ted Ginn Jr.

Ohio State's strength on both sides of the ball will help it survive during the inevitable game in which one unit or the other struggles. The Big Ten conference is strong enough -- with Michigan currently enjoying a top-10 ranking in a thus-far resurgent season, and Iowa a solidly ranked team -- to prevent the Buckeyes from having to face any questions about schedule strength, especially when the Texas game is factored in. That win over the Longhorns will probably stand up as the best non-conference result by any of the top teams at season's end, giving the Buckeyes another leg up on the competition for BCS spots.

Beyond Ohio State, it looks as if the preseason theme of a wide-open championship race is coming to fruition. The SEC is the nation's strongest conference, but its top-heavy depth could easily saddle all its best teams with a loss. Georgia, for example, has a championship-caliber defense but a very limited offense. That offense will keep the Bulldogs from contending for the national title, but the defense is good enough to upset Auburn or Florida.

Either West Virginia or Louisville is a fairly safe choice to emerge undefeated from the Big East, but such a result could create a new kind of controversy for the BCS. In past years, the BCS has dealt with issues arising from multiple one-loss teams and multiple undefeated teams. But suppose West Virginia knocks off Louisville and goes undefeated, yet ends up third in the BCS rankings behind one-loss Auburn?

The BCS title-game participants have always had at least the second-best won-lost record from all BCS-conference teams. While nobody would argue that the Big East is competitive with the SEC top to bottom, keeping an undefeated team home in favor of a team with one loss would certainly kick the BCS grumbling into overdrive.

But it's far too early to worry about such scenarios. As Colorado nearly proved at Georgia on Saturday, assuming too much about the college football season before the month of September is out is a fool's game.

The only thing we know for sure this season is what we don't know -- namely, the top two teams. That is unlike a year ago, when USC and Texas went wire-to-wire atop the polls before meeting in the championship game. While Ohio State has led the polls all season, the no. 2 spot has already been occupied by three teams: Texas, Notre Dame, and now Auburn. Odds are, a few more schools will rotate through that spot before bowl bids are handed out in early December.

John L. Smith Trophy

It will come as little surprise to anyone who watched Michigan State's second-half debacle against Notre Dame Saturday night that Spartans coach John L. Smith gets to keep his namesake trophy this week.

There was no single strategic decision by Smith that cost the Spartans a game they led by 17 points at the start of the fourth quarter. Instead, it was a top-to-bottom, wheels-come-off performance that included:

  • three turnovers in the final seven minutes;
  • three holding penalties (one was declined) and two wasted timeouts in a single series of downs;
  • a failure to give bruising tailback Jehuu Caulcrick the ball once in the final 14 minutes. Caulcrick's final five carries went for 18, 30, 6, 3, and 9 yards;
  • a kick returner, Demond Williams, who ran a kick back from his 6-yard-line to his 11 before inexplicably taking a knee with no one around him.

Those are not the signs of a well-coached team, and JLS may be running out of chances to turn things around at MSU. The Spartans' typical MO in recent years has been to knock off a good team, often Notre Dame, before folding like a house of cards the moment they start getting any national media attention. Last year's upset of Notre Dame was followed by a home-field beat down at the hands of Northwestern. Smith may have to beat Michigan in two weeks and get eight or nine wins to keep his job.

BlogPoll Ballot

This season, I'll again be voting in the BlogPoll, hosted by MGoBlog. I'll post my ballot in Junkie each week. Feel free to comment -- my rankings may change based upon your suggestions.

Rankings that require some explanation include:

  • Finally adding Iowa, though I remain relatively unimpressed by the Hawkeyes. Obviously they have a chance to change my opinion against Ohio State, this week, even if they lose (raising teams after losses and dropping them after wins are part of the unique charm of the BlogPoll).
  • Dropping Notre Dame after the Michigan State win. Did you see that game? Or the Michigan one? Notre Dame is not a very good football team. The defense is awful, the offensive line gets pushed around, and Brady Quinn has spent disturbing portions of the year throwing the ball nowhere near his intended target. That may have been the worst 300-yard, five-TD game in history.
  • Georgia likewise drops after its near-loss to Colorado.
  • Look for things to shake up in the coming weeks. I'm currently resisting the urge to move Michigan up ahead of Auburn and West Virginia until I see how the Wolverines handle should-win games at Minnesota and vs. Michigan State.
Rank Team Delta
1 Ohio State --
2 Auburn --
3 Southern Cal --
4 Michigan --
5 West Virginia --
6 Florida --
7 Louisville 1
8 Louisiana State 1
9 Georgia 2
10 Texas --
11 Virginia Tech --
12 Oregon --
13 Oklahoma 1
14 Tennessee 1
15 Clemson 4
16 Nebraska --
17 TCU --
18 Cal 3
19 Notre Dame 6
20 Iowa 6
21 Boise State 3
22 Florida State --
23 Rutgers --
24 Missouri 2
25 Washington 1

Dropped Out: Boston College (#18), UCLA (#20), Arizona State (#25).

Posted by: Russell Levine on 25 Sep 2006

65 comments, Last at 29 Sep 2006, 12:29am by chris

Comments

1
by Tarrant (not verified) :: Mon, 09/25/2006 - 9:37pm

Here's another doomsday BCS scenario for you.

Auburn is #2 in the AP poll. But the AP poll isn't used in the BCS anymore and has no impact on whether a team gets a spot in the title game.

In the Coaches' poll, and the Harris Interactive poll, each of which make up 1/3 of the BCS tally, Auburn is #3, with USC at #2.

Could there be a scenario where both teams win out, Auburn remains #2 in the AP, but #3 in both the Coaches' and Harris polls, while USC takes #2 in those, but #3 in the AP?

Unlikely, given that I see potential landmines for both USC and Auburn, but interesting.

2
by Fourth (not verified) :: Mon, 09/25/2006 - 10:40pm

It's really too bad there's no way any SEC team will go undefeated. LSU already took a loss to Auburn, and Florida beat Tennessee. Georgia, Auburn, and Florida all play each other. Florida has to play LSU. Georgia has to play Tennesseee. As a bonus, the east champ has to play the west champ in December, so you'll probably see an Auburn-Florida or Auburn-Georgia rematch. Meanwhile, West Virginia plays...Louisville. I hate West Virginia.

3
by ChrisFromNJ (not verified) :: Mon, 09/25/2006 - 11:05pm

(Disclaimer: I don't particularly follow or care about college football. But I do feel the need to set the record straight, just this once.)

#2- hey, hey, easy there. Lousiville isn't the only nationally ranked team that West Virginia has yet to play...

4
by Kyle J (not verified) :: Mon, 09/25/2006 - 11:37pm

As an MSU fan who witnessed the Saturday night meltdown first hand, I applaud your decision to award the JLS award to JLS himself. One more item for your list of errors: At some point, couldn't Smith or one of his assistants have told Stanton to stop trying to run the ball and throw it away when there was nothing there? He killed any chance we had to come back.

Fire John L. Smith! (Secondary recommendation: Hire Mariucci.)

5
by Charles (not verified) :: Mon, 09/25/2006 - 11:48pm

Ok, how about this scenario for the national title:

Undefeated #1 Ohio State vs undefeated #2...

Rutgers?

Five touchdown spread on the title game.

6
by Russell Levine :: Mon, 09/25/2006 - 11:50pm

Kyle, as a Michigan grad, even I feel your pain. To me, MSU is the biggest tease in college football. There is no reason that program can't be just as competitive as any in the Big Ten. They have excellent facilities (far superior to Michigan's) and the same access to talent that the Wolverines do. Yet time and again, just when you thnk they're going to matter, they don't.

I ge the feeling that this will be the last year for JLS, and I also believe the administration will try to make a big splash and land a big-name coach like the guy you mentioned.

7
by jed (not verified) :: Tue, 09/26/2006 - 12:05am

Usc could still lose to Oregon, Cal or Notre Dame ( or 2 of the 3). West virgina even if undefeated shouldn't play for the national title as its schedule is too weak and the big east should lose its confernces automatic berth. If Oklahoma loses to TX it probably won't get the big 12 south division title much less the conference title and hopefully won't be in a BCS game as a at large selection.

8
by Dave (not verified) :: Tue, 09/26/2006 - 12:59am

#7, the Big East is 9-7 vs. teams from the other BCS conferences this year, which gives them a better winning percentage by that measure than any conference except the Pac 10 (6-4). Any way you slice it that's a much better showing than the inept Big 12 (3-8) or the ACC (4-5) or SEC (4-5). The Big Ten at least has a winning record there (7-6).

Look, L'ville and WVU are legit top-ten teams (even if WVU has taken a page from the Kansas State school of non-conference scheduling). Rutgers is a legit top 25 team. And SU, UConn, USF, and Pitt aren't that bad. Heck, even Cinci stayed in the game vs. OSU and VT well past halftime.

9
by Kevin11 (not verified) :: Tue, 09/26/2006 - 2:02am

Russell, thanks for articulating a point that I've been making for many years- teams that pundits claim are "ready to run the table" rarely do. Just last year everyone thought Vandderbilt would finally go bowling after a 4 - 0 start with three wins vs name opponents, then BANG! They stepped on a landmine and lost to Middle Tennessee State.

Even now there are people that assume that TCU and / or Boise State are mortal locks to go undefeated and play in BCS bowls, but there's a lot of football left to be played.

10
by Travis (not verified) :: Tue, 09/26/2006 - 2:33am

Re: 7, 8

And the Big East hasn't lost any games to non-BCS teams, which is more than you could say for the Pac-10 (3 losses), the Big-12 (7 losses, including 1 to a I-AA team)), the ACC (4 losses, including 1 to a I-AA team), SEC (1 loss), or Big Ten (3 losses, including 2 to I-AA teams). It's not a great conference, but there's no Duke or Illinois or Stanford there, either.

11
by Fourth (not verified) :: Tue, 09/26/2006 - 3:02am

Re: 3

Ah, yes, 23rd ranked Rutgers at home. Pitt is decent as well. Nothing compares to what the very top-heavy SEC is going through this year though...happens to about 1 major conference every other year, where 3 or 4 or even 5 teams are good enough to be top 15 or 20 and none can make it to number 1 or 2. This is usually limited to the SEC and big 10, though everyone thought the ACC would be like this as well (oops, miami is awful under Coker).

12
by Erasmus (not verified) :: Tue, 09/26/2006 - 10:36am

The ACC is worse then the Big East. No more Big East slamming for the time being.

Alabama should be 4-0 going into the Florida game.

13
by Tarrant (not verified) :: Tue, 09/26/2006 - 10:41am

Conference USA suspended a replay official for making a mistake last weekend. After a coaches' challenge, the official ruled that a Houston player's forward progress was "clearly stopped" well before the ball was knocked out of his hands.

No one is disputing that his call was correct.

However, Conference USA rules state that forward progress can only be reviewed with respect to whether or not a player got a first down, and cannot be reviewed for any other incident (such as a fumble). So in reality, he should have upheld the call on the field despite indisputable visual evidence that the call on the field was wrong.

This is one of those "Why can't they review everything?" kind of things that just makes you shrug your shoulders about replay.

14
by Dennis (not verified) :: Tue, 09/26/2006 - 10:41am

Re #1: This is the fundamental problem with the BCS - it only "works" when there are exactly 2 undefeated teams. If there are 3 or more unbeatens, or 1 unbeaten and a bunch of one-loss teams, or no undefeated teams, then it's a mess.

15
by Pat (not verified) :: Tue, 09/26/2006 - 10:49am

I can't agree there: the main problem is when you have more than 2 undefeated teams. If you've got, say, 3 one-loss teams, there's no real way for them to complain if they don't get picked for the championship game, because they could've done something - they could've won the game they lost.

You can't say that for multiple undefeated teams, which is why it's not fair. Having some teams need to win more than others to get into the championship game is just saying that some teams have easier schedules than others, and that's just college football.

16
by Erasmus (not verified) :: Tue, 09/26/2006 - 11:17am

It's the same problem if there was no BCS either.

17
by Pat (not verified) :: Tue, 09/26/2006 - 11:22am

Oh, and brief thoughts on the weekend:

1) Don't give the Ohio State defense credit for those interceptions that sealed the Penn State game. Those were all Morelli, especially the second one. As soon as that ball left his hand, I said "oh look, another interception." Just a horribly thrown ball.

2) I can't believe how terrified Tressel became of the passing game versus Penn State after that first interception. It wasn't until late in the 3rd quarter that Smith even attempted another deep pass, and it wasn't until the 4th that he completed one. And most of the other passes were short passes to the sideline that couldn't easily be intercepted.

3) I really wish the field had been in better condition at the end of the game - on the interception returned for a touchdown, the announcers said "oh, that camera angle showed he was a few inches inbounds the entire way" except that he was closest to out-of-bounds on the portion of the run that his feet were obscured by a Penn State player. On the other camera angle, it really looked close, but the field was so beat up you couldn't tell. Regardless, though, that wasn't a touchdown - he spiked the ball at the one yard line (which he admitted afterwards, by saying "I'm just glad they didn't review that spike at the end"). Wouldn't that have been either a fumble or an illegal forward pass?

4) I don't know what was up with Brady Quinn for the first half of the game versus Michigan State. He just looked awful. He telegraphed his reads by about a mile. On one play, a receiver went in motion, right to left, and shifted upfield right at the snap. Quinn stared right at him after the snap all the way through the throw, and the entire Michigan State secondary converged around him. By the second half, though, he looked better, so maybe it was nerves.

5) The linebackers for Notre Dame were just awful for the first half, constantly out of position and getting burned by play fakes.

18
by John (not verified) :: Tue, 09/26/2006 - 11:27am

True, but at least then we wouldn't have the BCS committee (or whatever they call themselves) trying to convince us that it is, and somehow their messed up system can produce the one true champion without fail. I think they should either go back to the old system or have a playoff. What we have now tries to please everyone and in the end pleases no one.

19
by Dennis (not verified) :: Tue, 09/26/2006 - 11:28am

Re 15: If there are three 1-loss team, the team that is left out of the "championship" game has just as much right to be there as the other two. So what if they could've won the game they lost, so could the other two teams.

Re 16: Right, that's why the BCS is ridiculous. The only good thing it does is guarantee that if there are exactly 2 unbeatens, they'll play each other, rather than go to different bowl games because of their conference tie-ins.

20
by BC Pat (not verified) :: Tue, 09/26/2006 - 11:36am

wake forest should get a #25 vote if they stay undefeated. I know they have a historic level of suckage but you now you love giving the little guy some love every year.

21
by Derek (not verified) :: Tue, 09/26/2006 - 11:43am

#10,
Of course the distinction between BCS and non-BCS teams doesn't tell us a whole lot. Did they play Illinois or Ohio State? TCU or Eastern Michigan? Some 1-AA teams are equal to or better than bad Division 1 teams. Would anyone be suprised to see Temple lose to University of Northern Iowa, Montana, or Appalachian State? Yet the Owls were a BCS team only a few years ago.

Also, the Big East only has 8 teams which means less chances for weaker teams to lose games to non-BCS and 1-AA competition.

Despite all that, I agree that the Big East gets criticized too harshly. If Louisville and West Virginia keep winning, it will change. College football is about tradition so you have to prove yourself for a few years before anyone will notice.

22
by Derek (not verified) :: Tue, 09/26/2006 - 11:51am

#17,
I posted about the Ohio State spike at the 1-yard line during the game...funny to see that the player admitted his mistake. Of course neither the replay booth nor the announcers even noticed. You would think somebody in the production truck might have pointed out the most critical aspect of a play that big.

Hopefully Penn State gave my Hawkeyes enough of a blueprint to beat OSU in Kinnick this weekend. I agree with Russell that Iowa has yet to play a complete game. It would be nice if they put it together this weekend.

23
by Pat (not verified) :: Tue, 09/26/2006 - 11:52am

If there are three 1-loss team, the team that is left out of the “championship� game has just as much right to be there as the other two.

What? Why? The other two teams would've played a more difficult schedule. They earned their way there. How is this any different than the tiebreakers to get into the playoffs in the NFL?

This is like saying the NFL playoffs are unfair because last year, the Patriots got into the playoffs, and the Chiefs didn't, even though they both went 10-6. They don't have equal right - the Patriots won their division. The Chiefs didn't.

A system is fair so long as every team has some chance at the beginning of the season, based purely on their performance, to make the championship. Equal chance isn't important, because you can't ever guarantee that.

So what if they could’ve won the game they lost, so could the other two teams.

How the other two teams got in isn't important. It's why the last team got left out that's important. If an undefeated team gets left out, and the reason they get left out is their schedule, then they legitimately had no chance at the beginning of the season to make the championship game.

Even a coin flip, or drawing straws, at the end of a season to select the championship teams from the undefeateds would be better, because then at least every undefeated team has a chance.

24
by Ben (not verified) :: Tue, 09/26/2006 - 12:37pm

Bad officiating, huge comeback, no real legit team (even OSU has some major flaws), this is what football is all about. In fact, this is what football has been about since its inception waaay back when. Officiating hasn't gotten any worse than say a decade ago...it's just televised more allowing us (the fans) to see more of the mistakes missed. Because it hasn't gotten better (MSU-ND game) we think it's still getting worse. I may be wrong, but if you look at that huge fiasco, in the replays, it looked like some coach grabbed that players face mask (MSU #6?) and jerked it down. That's right before he went back "into" the fray. I may be wrong (I know, I already stated that, but I don't believe I am, so I have to cover myself) but it really did look like a blue-sleeved arm reached out and jerked down.

25
by Tarrant (not verified) :: Tue, 09/26/2006 - 1:04pm

Re: #23.

Pat, you know better than to say when there's three teams with one loss and one is left out, it's usually because of the schedule.

9 times out of 10 in college football it's going to be because the team left out lost later in the season than the other two, or (also true for multiple undefeated teams) started the year ranked lower than the other two and just couldn't catch up (e.g. Auburn 2004).

Right now, OSU is #1. USC is #2. Auburn is #3. WVU is #4. If all four teams win out, will those poll numbers change significantly? Probably not.

It seems that teams are best served by putting their hardest ("BCS qualifier") game as early in the season as possible. If they win, they get the poll bump they need to get an early edge in the BCS race. If they lose, they've got a "quality loss", they get credit for playing a "real opponent", and they'll be the first one-loss team in line if everyone else falters.

If when your team loses matters more than who you lose to, then it really isn't at all equivalent to tiebreakers for the NFL playoffs.

26
by Pat (not verified) :: Tue, 09/26/2006 - 1:32pm

9 times out of 10 in college football it’s going to be because the team left out lost later in the season than the other two

If that's true, weight the computer rankings higher. They'll do a better job on the schedule adjustment, and they're unbiased as to the timing.

It probably should be a 50/50 ranking, as the computers do a better job with the schedule adjustment, but humans do a better job with margin of victory adjustment.

That being said, though, I think you'd be surprised. One of the big advantages that the BCS ranking has over the real polls is the fact that teams don't actually have a "1, 2, 3" rating - if two teams are very close in terms of #1, they'll get very close rankings in the BCS poll portion.

27
by DMP (not verified) :: Tue, 09/26/2006 - 1:52pm

I don't even have the strength right now to complain about John L and MSU. I think he knows how to get plays to work, but he seems to have little common sense on when to run them. For instance, if it's a torrential downpour, and you are running the ball with great efficiency on a week defensive team, and you have a 17 point lead and running also helps you kill time, well, why not just keep running the ball? Why call a little tight-end screen on 2nd and 5 when you can just run for it? It's slippery and it's crowded -- not the best place for a touch pass.

And why is it that year in and year out MSU defenders over-pursue and take bad angles? And why is John L always having DBs lay back off WRs so much? Now, bad players can make even the best coach look bad - witness Bill Belichick and the last Patriots game. The thing that gets me about MSU under John L, though, is that the players don't just make mistakes, they make GIGANTIC mistakes that border on the comical. A defender who over-pursues and runs right past the ball carrier at 80mph, missing the tackle. A questionable play call that turns 2nd and 7 into 2nd and 17 to 3rd and 42. (I don't remember this sequence very well anymore. Withdrawal symptoms, I think.)

Drew Stanton has always had guts, I don't think there is any question about it. However he makes dubious decisions too many times. I don't see where people think he's a good pro prospect at all. (I have similar opinion of Brady Quinn, by the way.)

Finally, Jehuu Caulcrick has a butt the size of my TV, legs as thick pine tree trunks, the stockiness of a keg barrel wrapped with another keg barrel. YOU MUST HAVE HIM RUN THE BALL AGAINST A WEEK DEFENSIVE TEAM!

28
by Opiwan (not verified) :: Tue, 09/26/2006 - 2:35pm

Re: #17 sub 1)

I've taken to calling him "PA Lottery" Morelli the last couple days because of the Pick Sixes... I didn't know whether to throw myself or the TV out of the window after the first one, and the second one was just icing on the cake.

29
by Dennis (not verified) :: Tue, 09/26/2006 - 3:22pm

If that’s true, weight the computer rankings higher. They’ll do a better job on the schedule adjustment, and they’re unbiased as to the timing.

I've been saying that since the BCS started. But every time it doesn't work out how people think it should (which is almost every year), they adjust the formula to give more weight to the human polls and less to the computer rankings.

That being said, though, I think you’d be surprised. One of the big advantages that the BCS ranking has over the real polls is the fact that teams don’t actually have a “1, 2, 3″ rating - if two teams are very close in terms of #1, they’ll get very close rankings in the BCS poll portion.

That's totally untrue because 2/3 of the BCS formula is the polls, which have a very distinct "1, 2, 3" ranking. #25 is absolutely correct that when you lose is much more important than strength of schedule or any other factor.

Another very important factor is where you are ranked in the preseason polls. If a team that is unranked or low-ranked in the preseason finishes 11-1 and a team in the presesason top 5 finishes 11-1, odds are the top 5 team will be ranked higher in the final poll even if the unranked team played a tougher schedule.

30
by Tarrant (not verified) :: Tue, 09/26/2006 - 3:37pm

Note that Pat is correct to a point, actually. The BCS doesn't use the strict "1" "2" "3" of the polls anymore - they use the percentage of points a team has out of the total possible. A team with every possible #1 vote would be 1.000 in the new BCS. A team with every possible #2 vote would be .9600, etc. So if two teams were only a handful of points apart, the polls would put them at a straight #2 and #3, but the BCS would have them quite close.

In such a situation, the computers would cast the "deciding" vote, as it were.

But in cases where there's dozens of points separating two teams (like one team has most of the #2 votes, while the other has most of the #3 votes), the computer rankings would have to be disparate in order to matter, since the polls carry 2/3 weight.

31
by Pat (not verified) :: Tue, 09/26/2006 - 4:24pm

But every time it doesn’t work out how people think it should (which is almost every year), they adjust the formula to give more weight to the human polls and less to the computer rankings.

Not really. The changes that they've made were pretty substantial, and while it seems like they keep discrediting the computer rankings, they're really balancing them better.

Previously, the 'quality win+strength of schedule' etc. components were actually giving the computers two bites at the apple, so that's why that went away. It was artificially boosting the portion of the computer rankings.

Probably the main problem right now is that the Harris Poll isn't really a good independent source of information as well.

It'd be nice if they got rid of the wacko computer rankings (Billingsley) as well.

32
by kal (not verified) :: Tue, 09/26/2006 - 4:54pm

I still say that a college national championship isn't that important relative to being able to do historic bowl matchups, and if that results in a confusing national champion ranking, so be it. I'd much rather my team go to the Rose Bowl against a Big-10 team even if the matchup isn't as awesome for whatever reason. That's part of history. This? Makes every bowl game just a good game except for the championship.

That's kind of sad to me.

33
by mactbone (not verified) :: Tue, 09/26/2006 - 5:03pm

I have to say though, last year had some phenomenal games. I wish BCS weighted the computers more to limit poll inertia, but man, last year had some great games and that's what I want.

34
by NewsToTom (not verified) :: Tue, 09/26/2006 - 6:02pm

I've been meaning to do a longer piece on this, but there are two real problems with the BCS. (1) You have available after the regular season one football game between two teams. (2) With the limitation of (1), you have to design a system ex ante when the number of "deserving" teams ex post is not necessarily equal to two. Three times, the BCS system has worked perfectly and given us games we would not have gotten without the BCS: Miami-tOSU, USC-Oklahoma, and USC-Texas. Two of those games were absolute classics. FSU-VaTech would have occurred under the Bowl Alliance, but was obviously the right result. OU/LSU/USC is impossible to resolve equitably with fewer than 3 games. Ditto OU/USC/Auburn+Utah. Then you have OU+FSU/Miami/etc and Miami+Oregon/Colorado/Nebraska/etc, both awful messes. I'm not trying to say that the BCS is perfect, but I do think it's better than what came before it and the problems that it has are not easily to solve without changing the structure.

35
by Travis (not verified) :: Tue, 09/26/2006 - 6:22pm

I prefer a system in which all undefeated teams (e.g. Auburn, Boise State, and Utah in 2004) who meet minimal criteria (e.g., can't schedule 7 Sun Belt and 5 I-AA teams), plus any team with a valid BCS or poll #1 or #2 ranking (USC in 2003, but not Michigan) would be allowed to play their way into the BCS title game. Those play-in games, like in Major League Baseball, would only be scheduled if necessary.

I advocated this last year, and was mostly shot down, but this seems like the ideal thread to bring it back up.

36
by NewsToTom (not verified) :: Tue, 09/26/2006 - 6:43pm

But how would you play your way in? When is the game, and what's the effect on the loser? Say you have an undefeated Boise State team this year, tOSU is undefeated, and Michigan, Florida, USC, and Texas all have 1 loss and are ranked 2-5 respectively in the BCS with Boise State #6. Does Boise State get a play-in game against #2 Michigan on New Year's for the right to face Ohio State the next week? Or do you do it the week after the regular season ends, when finals are for most every school?

Alternative scenario: tOSU, USC, and Boise State are unbeaten. tOSU is #1, USC #2, Boise State #6 in the BCS rankings. Does Boise then play USC for the right to play tOSU for the BCS Championship?

37
by Travis (not verified) :: Tue, 09/26/2006 - 7:11pm

Well, obviously Michigan could not be #2 in any almost any scenario where Ohio State is undefeated thanks to the bubble nature of the polls, but yeah.

Seed the teams with a valid claim of being the national champion based on regular season performance (undefeated, or #1 or #2 in the polls) using the BCS or something like it, so an undefeated team like Boise State in a year where there's no other undefeated team doesn't wind up #1 or #2. Let the teams play on New Year's or some other mutually agreeable date either in a bowl game or at the higher seed. In a rare year like 2004 with 5 or more undefeateds, have the lowest seeds play each other Christmas week to see who advances to the next round.

Yes, USC would likely mop the floor with Boise State, but what about 2004 Auburn? Or 2004 Utah, with the #1 pick in the NFL draft at QB, and a coach now at Florida?

38
by Pat (not verified) :: Tue, 09/26/2006 - 7:33pm

so an undefeated team like Boise State in a year where there’s no other undefeated team doesn’t wind up #1 or #2.

You could eliminate teams like Boise State using a minimum schedule strength qualifier. Do it over a 5-year period and there shouldn't be any issue there.

That is, basically say "in order for an undefeated team to get into the playoff, the schedule over the past 5 years must contain X teams that have been ranked in the BCS top 25 at the end of the year."

That's really the only way to avoid a team scheduling a cupcake schedule to get into the BCS via an undefeated loophole. It might also force teams from conferences on a down cycle to schedule harder.

I'm actually amazed that a "minimum schedule strength" hasn't been introduced yet. As much as people want to say "going undefeated and not playing for the National Championship is crap", Boise State's schedule was really, really easy.

39
by NewsToTom (not verified) :: Tue, 09/26/2006 - 8:03pm

Re #38
I see what you're trying to do, but I'm not certain it will work very well. First, I'm pretty sure it doesn't solve the K-State problem of cupcake scheduling. Enough Big 12, Big Ten, SEC, and likely Pac-10 and ACC teams will be ranked that you won't have to play good non-conference opponents. In fact, I'd bet X would be low enough so that you didn't have to. Second, given that scheduling is often done several years in advance and that programs fluctuate over time, there can be no guarantees of quality. Colorado is a prime example of this. I would submit that a better rule is to schedule an average of one BCS-conference team as part of your non-conference schedule each season, but I would probably declare Notre Dame to be a BCS team under this rule. For independents, it's 9 of 12 games per year against BCS teams on average. For non-BCS teams, maybe 1.5 BCS conference foes per season.

Overall, I think the non-BCS team playing for the national championship issue is overrated. In the past 15 years, the only undefeated non-BCS I-A teams I can think of are Utah and Tulane in 1998. Compared to the other issues, I think this one is small potatoes. But, ceteris paribus, I would prefer a system where we have an opportunity to see how just how good a Utah in 2004 really is.

40
by Travis (not verified) :: Tue, 09/26/2006 - 9:09pm

In the past 15 years, the only undefeated non-BCS I-A teams I can think of are Utah and Tulane in 1998.

Marshall, behind senior QB Chad Pennington, went 13-0 in 1999.

41
by Harris (not verified) :: Tue, 09/26/2006 - 10:16pm

Wow, if Notre Dame wins out maybe they can drop completely out of your poll.

42
by Pat (not verified) :: Tue, 09/26/2006 - 11:04pm

Enough Big 12, Big Ten, SEC, and likely Pac-10 and ACC teams will be ranked that you won’t have to play good non-conference opponents.

So? Why is this a problem?

BCS conference teams should perfectly be allowed to schedule weaker non-conference opponents than non-BCS conference teams. Their conference schedules are nasty enough.

Really, if you look at the computer rankings, the major conference schedules are so ridiculously harder than non-major conference schedules that they don't need to make them harder.

43
by Pat (not verified) :: Tue, 09/26/2006 - 11:07pm

In the past 15 years, the only undefeated non-BCS I-A teams I can think of are Utah and Tulane in 1998.

Boise State, 2004.

There were 5 undefeated teams that year. But Boise State's was actually worse than Utah's, and that took effort.

That's actually the situation I'd be trying to avoid, mainly because a 4-team playoff would handle every other situation that's occurred historically.

44
by Travis (not verified) :: Tue, 09/26/2006 - 11:49pm

Boise State, 2004.

Toledo also went undefeated in 1995, although they did have one tie.

There were 5 undefeated teams that year. But Boise State’s was actually worse than Utah’s, and that took effort.

Utah tradionally schedules 3 mid-level BCS schools (in 2004, UNC, A&M, and Arizona, though in previous years Cal, Oregon, Indiana, and Michigan) plus Utah State, so they would mostly likely pass any strength-of-schedule barrier.

That’s actually the situation I’d be trying to avoid, mainly because a 4-team playoff would handle every other situation that’s occurred historically.

Yes, but a fixed 4-team playoff means that teams that have no valid claim to the title (e.g., Ohio State, 2005, who lost to #2 Texas and #3 Penn State) have another chance. Let Utah and Boise State play a play-in game on December 23rd in the rare 5-team case.

45
by NewsToTom (not verified) :: Wed, 09/27/2006 - 12:09am

Jeez, I can't believe I'd forgotten that many undefeated non-BCS teams. The only one that really concerns me is Utah in 2004, the only team other than 1995 Nebraska to win all their games by at least 14 points. The other problem to keep in mind is 1983, when a BYU team that IIRC didn't play anybody in the Top 25 is the only undefeated team.

Like Travis (#44), I am concerned about a team clearly not one of the best 2-4 in the country (variable, depending on year) winning the "national championship." E.g., in 2002 a system that produced a winner other than tOSU or Miami produces a wrong result. In a 4, 8, or more team playoff, this is an inevitable result. I've come to believe that (i) the BCS championship game is almost certain to include 2 of the 3 best teams in the country, (ii) this is better than what existed before the BCS, and (iii) the BCS's biggest avoidable problems come from the games other than the national championship game. I don't know that it's economically feasible to move away from games other than the national championship game away from the BCS, but that would be my preference.

46
by NewsToTom (not verified) :: Wed, 09/27/2006 - 12:16am

Actual game action alert!
UCF failed on a 4th down with 2:35 to go in the game. Southern Miss got the ball at the UCF 35. UCF is out of timeouts. Normally, you have three plays and a 40 second play clock to run off. Even with the clock starting on change of possession, you'll have to at least punt. But USM Coach Jeff Bower, unlike most coaches, has actually thought about late game clock control. He starts by taking a delay of game on the change of possession. Bang, clock down to 2:10. He then has the QB drop seven yards and wait before kneeling down. And then again. Lou Holtz in the booth comments about how this is a stupid strategy because it's killing his QB rushing average. But Bower knows better. In fact, he takes another delay of game penalty before third down, taking the clock down to 37 seconds. A quick kneel-down, and the game is over. Sure, his team lost almost 30 yards between the 3 kneel-downs and the 2 penalties, but he removed any doubt whatsoever about which team was going to win the game (barring, of course, the Herm Edwards Jets scenario). In the midst of so much clock idiocy, it's refreshing to see a coach who's actually thought about things.

47
by Travis (not verified) :: Wed, 09/27/2006 - 12:30am

The other problem to keep in mind is 1983, when a BYU team that IIRC didn’t play anybody in the Top 25 is the only undefeated team.

It was BYU in 1984 that won the national title without playing anyone in the Top 25, including Michigan in the December 21 Holiday Bowl (the WAC was locked in; BYU went 7 straight years). BYU did finish #1 in the pre-bowl poll, but it was such a strange year that I'm not sure who the BCS #2 team would be.

48
by NewsToTom (not verified) :: Wed, 09/27/2006 - 2:07am

Re #47
Grr, I'm dumb today. 1983 was the great Nebraska team that lost a close game to a good team in their home stadium(!) in the bowl game when they decided to go for an outright win instead of a tie after blowing people out all year, then dropped to #5.

49
by Dennis (not verified) :: Wed, 09/27/2006 - 9:40am

Re #45 Like Travis (#44), I am concerned about a team clearly not one of the best 2-4 in the country (variable, depending on year) winning the “national championship.� E.g., in 2002 a system that produced a winner other than tOSU or Miami produces a wrong result.

So then the NFL should get rid of the playoffs, because the Steelers were clearly not one of the best 2-4 teams last year. Since Indy or Seattle didn't win, it was obbviously the wrong result.

50
by Travis (not verified) :: Wed, 09/27/2006 - 10:21am

So then the NFL should get rid of the playoffs, because the Steelers were clearly not one of the best 2-4 teams last year. Since Indy or Seattle didn’t win, it was obbviously the wrong result.

No, because the NFL structure is far different from college. One of the unique things about the college regular season interesting is that every single game has an impact on a team's championship hopes. A fixed 4 (or far worse, 16) team playoff system would devalue that regular season, much in the way the NCAA basketball season has little meaning.

51
by Dennis (not verified) :: Wed, 09/27/2006 - 10:39am

Except every game doesn't have an impact because as has happened several times, teams can win all of their games and still not have a chance to win the championship.

52
by Pat (not verified) :: Wed, 09/27/2006 - 10:41am

Yes, but a fixed 4-team playoff means that teams that have no valid claim to the title (e.g., Ohio State, 2005, who lost to #2 Texas and #3 Penn State)

True, but even a two team playoff has that possibility. Had USC lost to Notre Dame, and probably another team as well, and Penn State ended up losing another game, Ohio State could've ended up #2. Well, it's at least possible.

53
by Travis (not verified) :: Wed, 09/27/2006 - 10:42am

Except every game doesn’t have an impact because as has happened several times, teams can win all of their games and still not have a chance to win the championship.

Which is why I'm advocating an expandable playoff system, with the number of entrants dependent on the number of unbeatens beyond the #1 and #2 teams.

54
by Travis (not verified) :: Wed, 09/27/2006 - 10:47am

True, but even a two team playoff has that possibility. Had USC lost to Notre Dame, and probably another team as well, and Penn State ended up losing another game, Ohio State could’ve ended up #2. Well, it’s at least possible.

In that case, Notre Dame or Oregon would have been the #2 team, but I see your point. I can't see a scenario, however, where a team would just be declared the national champion; I'm willing to give the #2 team, but no one else, another bite at the apple.

55
by HLF (not verified) :: Wed, 09/27/2006 - 3:19pm

Sorry, but though you guys mean well you're missing the forest for the trees.

You have a situation where you're deciding who's "national champion" based on the formula figure skating uses -- some wierd combination of beauty (in the eye of the beholder), uniforms, musical accompaniment, degree of difficult (schedule), timing, and behind the scenes influence peddling. At no time do you decide this purely based on a level playing field, or on the field activities. It's as on the level as an USSR election, and you're complicit in the corruption when you, meaning well, acknowlege that these ice dancers are champions of NCAA div. 1 football.

May I ask why this splendid system isn't used in men's div. 1-AA NCAA football, since some above object to the NFL comparisons? Sure, you say, the Steelers don't count, but why do the arguably 6th or 8th place regular season teams in 1-AA get a chance a true Championship?

How about Div 2, or Div 3? These atheletes get no scholorships (unless their academic scholorships), yet somehow they manage to decide a champion on the field, giving all reasonable contenders a chance.

Saying that only an undefeated team (or a select one loss team in pretty uniforms who's in with the judges) deserves even a chance to win a championship is on it's face absurd. The best college football team in America frequently has one loss, often even two.

In every serious sport in America, we play the season to determine a number of qualifiers for a post season tournament. That number of qualifiers is made large enough that all reasonable entrants are included, erroring on the side of inclusion. Then, that tournament determines the championship, not how your nails look or where some fat ignorant sportswriter went to school, nor who that coach's secratary's assist who doesn't even watch the games fills out on some ballot. There is no possible justification for the current "system" other than institutional corruption, in kind if not severity with organized crime.

Let's call it a gold medal instead of a national championship, at least, if we're going to award it with all the integrity of the Olympics (or the WWE).

Hopeless Lions Fan (and college football fan),
Seattle

56
by Pat (not verified) :: Wed, 09/27/2006 - 3:36pm

yet somehow they manage to decide a champion on the field

There's still a selection committee for the championship tournaments in all major college sports - you can't have a 119-team tournament.

Division IAA does it via a selection committee for the 8 non-conference champions in the tourney. How is that strictly better than using something like a BCS ranking? As far as I can tell, the BCS ranking is slightly better - it's more open.

I think a 16-team tourney is overkill. There's never been a year when more than 4 teams could legitimately claim to be the best in the country. No offense to Boise State fans, but their schedule was crap. Even Division IAA has a minimum-schedule strength qualifier (must have more than 7 Division I wins).

I just can't see how selecting 16 teams from a pack of 122 by a behind-closed-doors committee is so much better that it makes an open algorithm based on solid mathematics as well as two nation-wide polls look like a USSR election.

The only thing wrong with the BCS is that there are frequently more than 2 teams that deserve consideration. There are never 16 teams.

57
by zlionsfan (not verified) :: Wed, 09/27/2006 - 4:31pm

A 16-team tournament ensures that all conferences are included, which I believe would be an NCAA requirement, and ensures that all legitimate contenders would be included by providing a sufficient number of at-large spots.

The BCS ranking is based on what the BCS conferences believe will put their most deserving team in the title game, not what is most likely to rank the best teams 1 and 2, which is why it's been changed so often. A selection committee would have the opportunity to address that issue in a specific season without changing the entire formula - for example, the silliness with the margin-of-victory component, telling people who've spent many, many years tweaking their formulas that they must be changed because the BCS people say so.

Even if the BCS people had legitimate reasons for changing the formula, the problem is that they'd be doing so after the fact, which doesn't help the team(s) that were excluded that season. Including 16 teams ensures that a scenario that excludes a deserving team will be very, very rare. (For the cynical, it also provides 8 more games than an 8-team tournament, meaning lots more money for the NCAA and its member institutions.)

The BCS formula could be one of the tools they would use during the selection and seeding process, much as the RPI is used for basketball. Hopefully it would weight the polls less heavily - as HLF points out, it's highly unlikely that most people who vote are capable of accurately ranking the top 10 teams in the country, never mind the top 25.

7 Division I wins isn't a SOS qualifier any more than 5/6 I-A wins + count only one I-AA win is. It's simply a rule designed to encourage teams to schedule games with other teams at their level (read: teams in weaker conferences). I don't see the BCS adding a SOS qualifier any time soon, though. The whole point of the additional game was to pacify the non-BCS conferences. It would be pretty clear at whom the qualifier would be aimed, and there's not much those teams can do about it. Boise State already has two top-50 Sagarin teams on their NC schedule (Utah and BYU). Who else is going to agree to play them?

Ben (24): Yes, it was Matt Trannon who was presumably grabbed by a Notre Dame person on the sidelines.

I like it when Stanton runs (except at the end of games when they need to put the ball in a HB's hands). It makes my fantasy team do better.

58
by Pat (not verified) :: Wed, 09/27/2006 - 5:43pm

A 16-team tournament ensures that all conferences are included, which I believe would be an NCAA requirement

Honestly, if that's true, then the proper solution is simple - Division IA needs to be split in two. The talent disparity between BCS conference teams and non-BCS conference teams is just ludicrous.

The BCS ranking is based on

One third of the BCS ranking is based on math. Math that's been around for longer than the BCS, or you, or me. Laplace figured it out back when he wanted to rank players in a chess tournaments. Hundreds of years ago.

for example, the silliness with the margin-of-victory component, telling people who’ve spent many, many years tweaking their formulas

Including simple margin of victory in a ranking designed to decide the best team is a seriously flawed idea, and most people who were involved in the process agreed with this. Scoring margin is a biasable metric independent of wins and losses. It's not a smart thing to do.

And, incidentally, those changes weren't arbitrary. They were done in concert with a lot of statisticians, and most of the changes were by the recommendation of those statisticians.

The biggest problem with the BCS right now is the fact that they're including some seriously wacko rankings (Billingsley) and some overly simple versions (Sagarin's ELO-CHESS is very dated) and neglecting much more stable and accurate versions that still would be W/L only.

59
by HLF (not verified) :: Wed, 09/27/2006 - 6:05pm

Pat,

It astounds me that you find that the biggest problem with the BCS right now is which particular rankings are included. I have to believe I read that wrong.

HLF

60
by Pat (not verified) :: Wed, 09/27/2006 - 6:41pm

That's the biggest problem with the BCS rankings. Not the BCS itself. The BCS itself should go to a four-team playoff with a minimum schedule strength requirement, in my opinion.

61
by rams fan (not verified) :: Wed, 09/27/2006 - 7:23pm

"A system is fair so long as every team has some chance at the beginning of the season, based purely on their performance, to make the championship. Equal chance isn’t important, because you can’t ever guarantee that.
"
which is why even if wva beats everyone they play they have a good chance of not getting in to the national championship?

62
by Pat (not verified) :: Wed, 09/27/2006 - 10:17pm

I never said the current system is fair. It's not.

63
by Brian McClafferty (not verified) :: Thu, 09/28/2006 - 2:44pm

After 4 games, the Sagarin Ratings ranked ND's schedule the most difficult in the nation. Make anything of that you want. I do believe, however, that the Irish are 3-1 after those four games.

64
by NewsToTom (not verified) :: Thu, 09/28/2006 - 4:46pm

Re #63
It's actually very simple. Notre Dame is the only good team that's not a member of a conference. Teams that are members of conference are most willing to play a good team in a non-conference game before the beginning of the conference season, reserving weeks during conference season mostly for healing up and correcting mistakes against patsies like Buffalo. Notre Dame has played what are likely 4 of its 5 most difficult opponents. No other team ranked in the Top 25 can make that claim.

65
by chris (not verified) :: Fri, 09/29/2006 - 12:29am

Other than the fact it doesn't give you a champion (by any measure), the biggest problem with the BCS is that the people running it can't leave it alone.

It would have been interesting to see how things turned out if any of the versions of the BCS formulas were allowed to "just be" for more than a year or two.

Unfortunately, backbone tends to be lacking amongst those running the system. So the system has to change with each public squawk.

Of course, a 16-team playoff including automatic bids for smaller conferences makes the most sense. But there are some people who are never going to agree with it, particularly the crowd with the "it doesn't count if the stadium holds less than 90,000" mindset.

(It's rather odd that the way to solve depth disparities is to cut off smaller conferences from access to the very things that help to close the talent gap in terms of recruiting and building a fan base -- the opportunity for exposure against the top teams without totally screwing yourself over, and the chance to pick up extra money for wins.)