The undefeated teams just won't die. Now all seven crowd the top of the latest BCS rankings following USC's loss to Oregon.
The big change isn't Texas moving to #2, as that's unimportant really. But Cincinnati jumping to #5—on the strength of beating Syracuse? Very noteworthy. The Bearcats and Broncos were neck and neck last week and Cincinnati made only slightly better gains in the human polls, while TCU gained more than both of those teams combined, pointwise. But the computers pushed Cincinnati ahead: on three computers they gained 8 spots, for a net gain of 7 positions, while Boise State gained only 1 and TCU lost 4 net. Now it's Cincinnati and TCU who are neck and neck. I hadn't expected Cincy to pass anyone on the week that they played Syracuse; it just goes to show how volatile some of the computer polls are.
Let's take a look at why it might have happened, using the Success power rating as a guide. The Success ratings are a BCS-style ranking that doesn't use score margin.
# Team rec Rating[LW] Change BCS Computers
1. Iowa 9-0 15.13 [ 1] -0.41 2
2. Florida 8-0 13.40 [ 3] +2.05 1
3. Boise St 8-0 12.42 [ 4] +2.71 8
4. Alabama 8-0 11.66 [ 2] -1.48 3
5. Texas 8-0 10.56 [ 6] +2.55 3
6. Cincinnati 8-0 9.76 [ 7] +2.37 5
7. TCU 8-0 9.46 [ 5] +1.35 6
Iowa lost points, not because of a close game, and not because Indiana is a weak opponent, but because four of their seven previous FBS opponents lost. In particular, Michigan lost to lowly Illinois, dropping their stock as a victory. The same problem hurt Alabama but even worse: their best wins over Virginia Tech, Mississippi, and South Carolina were all diminished, and they didn't add any points since they were idle.
One of the big gainers was Texas, of course, as they beat ranked Oklahoma State. But Boise State gained even more, and not from beating San Jose State; almost all of their gain came secondhand, from Oregon's win over USC. Florida gained by beating Georgia but also by Tennessee beating Mississippi. In all, five of their six previous I-A foes won, and two of them played each other so they were essentially perfect. That, in addition to Iowa's foes stumbling, is why Florida passed Iowa in almost all the computer rankings.
Cincinnati's rise, like most of the others, came not from beating lowly Syracuse but from how their previous foes re-evaluated. All of their previous I-A foes—even Louisville and previously winless Miami of Ohio—won last weekend, with South Florida beating West Virginia. This is what gave the Bearcats their push to move up. And TCU, though they gained points in the Success rankings due to winning and their opponents going 3-2, didn't gain enough to stay ahead of Texas and Cincinnati. As the Success rankings show, TCU was ahead of both those teams last week and fell behind them this week, just like in the BCS computer composite.
The undead undefeated refuse to die on Halloween
Let's look again at each team's remaining schedule and project each game, giving it a percentage-to-win value. Then we'll compute the team's overall % chance to finish the season undefeated. We'll do this by using the X-conference 726-team power rating to give every game a team has played a rating; we then cross-compare all of these ratings with the game ratings of their opponent and see how many of the cross-comparisons are victories. Basically it's a way to look at all the aspects a team has shown so far in the season, from their worst to the best performances, since on a given day any version of a team could "show up." The method is explained in more depth here.
For each team a record is projected: First, simply by counting wins and losses based on whether a team is favored in that game, and second, by adding the percentages to obtain a cumulative projection, or rather the record a team is most likely to finish at. Odds of going undefeated are computed by multiplying the odds of winning each game.
Florida remaining schedule & forecast |
||||||||
| 11/7 | Sat | vs. | *Vanderbilt | 99.9% | W | |||
| 11/14 | Sat | @ | *South Carolina | 95.8% | W | |||
| 11/21 | Sat | vs. | Florida International | 99.9% | W | |||
| 11/28 | Sat | vs. | Florida State | 90.6% | W | |||
| 12/5 | Sat | N | Alabama? | 53.1% | W | |||
Favored: 13-0 Projected: 12-1 |
||||||||
Odds of going 12-0: 86.8%, 13-0: 46.1% (up 10.4%) |
||||||||
Florida Gators (8-0; BCS rank: #1)
Florida's solid play in beating Georgia helped their odds beyond the relief of getting another game out of the way.
The Gators' odds of beating South Carolina and Florida State increased slightly, and they broke out of the 50/50 tie they were locked in with Alabama, making them the official favorite to play for the national title.
Their odds of going 12-0 took a major leap, from 71.5% last week to 86.8% this week, and their chance to go 13-0 jumped from 35.7% to 46.1%.
Texas remaining schedule & forecast |
||||||||
| 11/7 | Sat | vs. | Central Florida | 99.9% | W | |||
| 11/14 | Sat | @ | *Baylor | 99.9% | W | |||
| 11/21 | Sat | vs. | *Kansas | 95.3% | W | |||
| 11/26 | Thu | @ | *Texas A&M | 79.7% | W | |||
| 12/5 | Sat | vs. | Nebraska? | 81.3% | W | |||
Favored: 13-0 Projected: 13-0 |
||||||||
Odds of going 12-0: 76.0%: 13-0: 61.8% up 16.4% |
||||||||
Texas Longhorns (8-0; BCS rank: #2)
Texas made an impressive leap forward in dispatching Oklahoma State, one that put them on top of the power ratings and made them a likely national champion.
The rest of their opponents had mixed results: Texas A&M became a more difficult foe, while Nebraska remained the projected Big 12 Championship foe but lost strength in terms of upset likelihood.
With just two "live" foes left in the way of going 12-0, the Longhorns are 76% likely to achieve that, up from 65.7% before the Oklahoma State game. More significantly, their odds of going 13-0 are back above 50% at 61.8%.
Alabama remaining schedule & forecast |
||||||||
| 11/7 | Sat | vs. | *Louisiana State | 89.0% | W | |||
| 11/14 | Sat | @ | *Mississippi State | 84.7% | W | |||
| 11/21 | Sat | vs. | Tennessee-Chattanooga | 99.9% | W | |||
| 11/28 | Sat | @ | *Auburn | 79.2% | W | |||
| 12/5 | Sat | N | Florida? | 46.9% | L | |||
Favored: 12-1 Projected: 12-1 |
||||||||
Odds of going 12-0: 59.7%, 13-0 28.0% down 6.0% |
||||||||
Alabama Crimson Tide (8-0; BCS rank: #3)
Not only did Alabama lose ground in the BCS last week, their odds of being national champs took a hit even though they were idle.
The Tide's odds of being LSU and Auburn remained basically static, but Mississippi State's chances to pull and upset jumped quite a bit after their win over Kentucky. With a tougher remaining schedule than Florida, they have a greater chance of losing before the SEC title game, and are a slight underdog in that game to the Gators.
The chance that they'll go 12-0 dropped from 68.1% to 59.7%, and their odds of going undefeated at 13-0 slumped from 34% to 28%.
Iowa remaining schedule & forecast |
||||||||
| 11/7 | Sat | vs. | *Northwestern | 96.3% | W | |||
| 11/14 | Sat | @ | *Ohio State | 40.7% | L | |||
| 11/21 | Sat | vs. | *Minnesota |
79.0% | W | |||
Favored: 11-1 Projected: 11-1 |
||||||||
Odds of going undefeated: 31.0% down 2.6% |
||||||||
Iowa Hawkeyes (9-0; BCS rank: #4)
Iowa was a trainwreck against Indiana until the fourth quarter, then all was well and they emerged as the only 9-0 team in the nation.Beating the Hoosiers didn't alleviate the concern about going into Ohio State, though, and even Minnesota gained some hope. But basically the scenario remains the same: Iowa loses to Ohio State but still goes to the Rose Bowl after Penn State beats the Buckeyes this weekend.
The Hawks' odds of going 12-0 fell slightly, from 33.6% to 31%. But with none of the other teams losing, Texas looking likely to win out, and falling to #2 in the computer polls, the Hawkeyes' chances of being the BCS title game at 12-0 took a bigger hit.
Cincinnati remaining schedule & forecast |
||||||||
| 11/7 | Sat | vs. | *Connecticut | 98.4% | W | |||
| 11/13 | Fri | vs. | *West Virginia | 98.4% | W | |||
| 11/28 | Sat | vs. | Illinois | 93.8% | W | |||
| 12/5 | Sat | @ |
*Pittsburgh | 65.6% | W | |||
Favored: 12-0 Projected: 12-0 |
||||||||
Odds of going undefeated: 59.6% down 2.1% |
||||||||
Cincinnati Bearcats (8-0; BCS rank: #5)
Knocking off Syracuse didn't relieve any uncertainty about Cincinnati's quest to go 12-0, and Illinois' win over Michigan showed they're still a threat.
The home contest against the Illini looked like a sure thing a week ago: Illinois hadn't played a game as good as the Bearcats' worst game of the season. But by beating Michigan 38-13, Illinois showed they can beat a decent team, and if Cincinnati has an off day they're vulnerable even at home.
The odds of Illinois winning still aren't that great—just 6.2% of the game-comparisons fall their way—but it's enough to actually reduce Cincinnati's odds of going 12-0 from 61.7% last week to 59.6%. On the bright side, the Bearcats' jump to #5 in the BCS keeps their slim national title game hopes alive, when last week it was looking like they might only pass TCU on the final week at the earliest, if at all.
TCU remaining schedule & forecast |
||||||||
| 11/7 | Sat | @ | *San Diego State | 99.9% | W | |||
| 11/14 | Sat | vs. | *Utah | 96.9% | W | |||
| 11/21 | Sat | @ | *Wyoming | 99.9% | W | |||
| 11/28 | Sat | vs. | *New Mexico | 99.9% | W | |||
Favored: 12-0 Projected: 12-0 |
||||||||
Odds of going undefeated: 96.9% up 8.9% |
||||||||
TCU Horned Frogs (8-0; BCS rank: #6)
It's nearly a foregone conclusion: TCU will be undefeated this season at 12-0.
The game against San Diego State now scores as a lock, along with the road game at Wyoming and home game against New Mexico.
The only stumbling block left is Utah, and that game is at home and the Horned Frogs will be heavy favorites, more than would be suggested in a game between ranked foes.
So barring an embarrassing stumble, TCU will be 12-0. But what will it mean? Last week we showed how USC had little chance of avoiding a loss the rest of the way, and with a potential loss by Texas on the books, TCU had a real shot at being in the national title game.
But just one week later those dreams look shot. Oregon's dominance of USC gave the Ducks new credibility in the human polls, Cincinnati jumped TCU in the computer rankings, and Texas appears much more likely to avoid a stumble. So now TCU must hope to stay ahead of Boise State to get a guaranteed BCS bowl game invite. Depending on how much credit Boise State gets for beating Oregon, that's not guaranteed, either.
It doesn't help either that TCU's remaining schedule is so easy. For computers that average a team's strength of schedule in a formula, it's bad news. And for formulas that add points for victories, the Horned Frogs won't gain much. Pretty much their best hope is to keep blowing out opponents—Utah especially—and impress the humans who make up 2/3 of the BCS formula.
Boise State remaining schedule & forecast |
||||||||
| 11/6 | Fri | @ | *Louisiana Tech | 90.6% | W | |||
| 11/14 | Sat | vs. | *Idaho | 97.2% | W | |||
| 11/20 | Fri | @ | *Utah State | 87.5% | W | |||
| 11/27 | Fri | vs. | *Nevada | 89.1% | W | |||
| 12/05 | Sat | vs. | *New Mexico State | 99.9% | W | |||
Favored: 13-0 Projected: 13-0 |
||||||||
Odds of going undefeated: 68.9% down 5.7% |
||||||||
Boise State Broncos (8-0; BCS rank: #7)
Boise State knocked off a foe they were a lock to defeated, basically, while the odds of their future foes rised slightly.
Across the board these results reduce Boise's odds of going 13-0 from nearly 3/4 last week to just over 2/3 this week.
With TCU almost certain to go undefeated, Boise State needs to hope for losses from other teams that might push them out of BCS bowl contention. An Ohio State win over Penn State, followed by an Iowa win over the Buckeyes, would keep the Big Ten a 1-BCS-team conference. A Texas loss in the Big 12 championship game would be a disaster, so Boise fans should root for Texas to go 13-0. And Notre Dame has to lose another game somewhere; Stanford and Pitt are good possibilities. Another USC loss helps a lot, too. If enough of these things happen, Boise State will join TCU as a second non-BCS conference representative.
There is the chance that voters will prop up Boise State, given their win over Oregon, and put them above the Horned Frogs; and as TCU's strength of schedule slides, the Broncos might be first in line. But Boise State's schedule strength doesn't do anything but slide, too. If TCU has a few scare games, or just barely gets past Utah, maybe the Broncos would pass an undefeated TCU team.
How Many Teams Will Go Undefeated?
It's looking like we'll have four undefeated teams out of the current seven. Five teams (including the aggregated Florida/Alabama) have a greater than 50% chance; therefore Iowa and one other team should suffer a loss. The odds of six teams going undefeated took a jump, from 4.3% to 5.6%, and the odds of all seven teams suffering a loss fell from 0.13% to 0.03%.
Who will the undefeated teams be?
- TCU: 96.9%
- Florida/Alabama: 74.1%
- Boise State 68.9%
- Texas: 61.8%
- Cincinnati: 59.6%
- Iowa: 31.0%
If this happens, we'll see the Florida/Alabama winner against Texas. If the Longhorns lose, then it gets interesting of course between Cincinnati, TCU, and Boise State.
LSU is a potential wild card only if they beat Alabama. Oregon is rising in the BCS and is currently #8, but their loss to Boise State makes them a very unlikely candidate as a 1-loss team even if a number of the undefeated teams lose.
With Florida favored over Alabama and Texas expected to win out, here are this week's four most likely BCS title game matchups:
Texas 51.6% over Florida
Texas 60.9% over Alabama
Florida 67.2% over Cincinnati
Alabama 64.1% over Cincinnati
Texas had been expected to lose last week only by cumulative odds; there was no game in which they weren't the favorite. By beating Oklahoma State and improving their cumulative odds, Texas now looks likely to make the national championship game and is a slight favorite over Florida or Alabama. If Texas loses and Cincinnati slips in, they'll be a fairly heavy underdog to either SEC team.
Big wins by Oregon, Texas demonstrate how USC, Oklahoma State were overrated
Here's the conventional wisdom after Saturday's big games: Texas "proved" themselves by whipping a tough Oklahoma State team 41-14, and Oregon shocked the world by destroying #4 USC 47-20.
Here's my take: Oklahoma State and USC were both exposed by blowout losses. Oklahoma State as a team who'd been having success despite not being of top 25 quality, and USC because they were staggeringly overrated at #4 by whatever measure you use.
USC: The Trojans were overrated
They struggled to beat marginal top 25 teams Ohio State and Notre Dame. Yet they were placed ahead of several undefeated teams.That USC was #4 in the AP poll shows ridiculous favoritism to the Trojans based on what they've done in past years, not for this year's play. They had a loss to Washington, a tough team but one with a losing record, and commentators decried the BCS system that dared to have them lower.
The pro-USC contingent went as far as to prop of Matt Barkley, a decent quarterback but certainly nothing special, for the Heisman Trophy. True, he's doing fantastic for a true freshman, and the Heisman race is wide open. But for commentators after the Oregon State game try to sell him as a true candidate—when he was 15 of 25 for 202 yards, 2 TDs and 2 interceptions—was laughable. Do these guys even look at the stats? He wasn't even in the top 50 in pass efficiency.
This phenomenon happens with USC every year. They win a few big games, are declared the best team in the country (perhaps best college football team of all time), then they lose a game. After a few weeks of "what's wrong with the Trojans?" articles, they blow out a few more opponents and again are called the best team despite having a loss on their record. The BCS system is blamed again for the fact that the Trojans aren't number one on the list.
What's different this year is that USC wasn't close to being the best team, and they ran into a better team on Saturday. By any measure, Oregon should have been ranked ahead of USC going into the game. The Ducks' only loss was to an undefeated team, not 3-5 Washington, whom they destroyed the previous week. They beat Cal just as bad as USC did, and held a win over ranked Utah. So in an AP-style poll, the Ducks deserved to be in front. What about true strength? In a basic score-margin-based power rating, the Ducks were slightly ahead, ranking 8th to USC's 11th. And as the BCS computer rankings demonstrate, Oregon had accomplished more in their season than USC so far, averaging 6th to USC's 9th. It was only the biased human observers who had USC ahead.
The USC mystique is pervasive; I even picked the Trojans to win the game despite it being at Oregon. USC has consistently done well against ranked opponents and I figured they'd win again. But seeing Oregon win didn't shock me, though the 27 -point margin was unexpected.
Oregon: Soon to be overrated themselves at Boise State's expense
Now the commentators are falling over themselves about Oregon. After all, if they beat the great Trojans, they must be incredible! Talking heads are putting the Ducks ahead of Boise State, who beat them fair and square. The excuses for not counting this game include: it was the first game of the season, it was in Boise, that was before Oregon "hit their stride." How about the real excuse: we don't think of Boise State as a real team, but USC is a real team. Tell me, if they'd lost the first game to USC under the same circumstances, who would disregard that game? No one.
One commentator (Craig James?) said he was moving the Ducks ahead of Boise, "not to take anything away from Boise State." Well, you are taking something away from Boise State. It's ridiculous and inexcusable to move a 7-1 team ahead of the 8-0 team that beat them, and it should never be done. Since we don't have a playoff, every game of the season has to be counted. Boise State 19, Oregon 8 was their playoff, and the Ducks lost. I guess losing a 1,000 yard rusher was a boon for the Ducks?
Oklahoma State: Exposed
The Oklahoma State Cowboys should have been a great team this year. And if you just look at wins and losses, they were doing pretty well. Fairly high in the Success rankings (#16) and the BCS computer rankings (average #17), they'd lost only to 7-1 Houston. But dig a little deeper and they weren't doing that well at all, ranking just 40th in Strength. They weren't winning convincingly, and it showed up in power ratings. And with Dez Bryant out for the year, the test they were able to give Texas was minimal, even though this game was supposed to tell us a lot about the Longhorns.
Now, after winning 41-17, instead of questioning whether Oklahoma State is even a top 25 quality team, everyone is talking about how Texas is finally playing good football.
Texas: They were already playing fantastic football
I'm not sure why Texas was getting so little respect all year. Most of it probably had to do with Colt McCoy's occasional interception, and not having as spectacular a year as last season. Some of it had to do with Texas' easy schedule. But the complaints peaked after they beat Oklahoma 16-13.
Think of how ludicrous that is: Texas wasn't a good team because they only beat Oklahoma by 3 points. Of course, the Sooners could easily have come into that game undefeated. While Oklahoma had two losses, they were still one of the best teams in the country, at least the equal of USC. And 16 points was the most they'd allowed except for Miami's 21. The game was played on a neutral field. If anything, Texas was a bit sloppy in that game, and they'd played sloppy football before. But the end result was always impressive.
Texas has consistently been in the top three in the Strength power rating and even before the Oklahoma State game they were #1 in both versions (I-A only, and the 726-team all-division version). It's strange that with Florida and Alabama struggling it was Texas who caught the most flak. The last few weeks (road wins over Missouri and Oklahoma State) should help nudge perception of Texas back into reality: they are—at least—the equal of Florida and Alabama at this point.
November 01, 2009 at 02:19 PM in analysis, commentary | Permalink | Comments (3)
Digg This | Save to del.icio.us