TrojanWire

USC: 30% Chance of BCS Title Berth

1–2 minutes

November 23, 2008

#5 Oklahoma killed #2 Texas Tech, 65-21. Oregon State’s last minute field goal topped Arizona, 19-17, taking the Beavers within one game of going to the Rose Bowl. And USC?!…USC did nothing, but still managed to move up to #5 in the BCS. Now how does it all add up for the Trojans’ title hopes? Thanks to Adam Rose at All Things Trojan and USC engineer Anthony Christodoulou it’s down to a (not-so-exact) science:

“with six games to go, there are 64 possible outcomes remaining and 31 that will find USC in Miami on January 8. That assumes USC can win out over Notre Dame (in the Coliseum) and UCLA (in the Rose Bowl).USC’s best shot is if Florida State upsets Florida next weekend, and then Florida beats Alabama the following weekend in the SEC Championship game.”

31/64 possible outcomes for a title shot? I’m no Einstein over here, but that looks more like a 48% chance. So why is the magic number at 30%?

The article continues:

If we give Oklahoma a 65% chance over Oklahoma State, Texas Tech a 75% chance over Baylor, Texas a 75% chance over Texas A&M, Mizzou a 33% chance of winning their conference championship, Florida a 75% chance over Florida State, Florida a 60% chance over Alabama, and USC wins out … crunchcrunchcrunch …The Trojans have a 30.1% chance of playing in the BCS National Championship Game.

And if the cards don’t fall right and USC doesn’t end up in the title game, and AP #17 Oregon State manages to beat #19 Oregon to end up in the Rose Bowl (remember OSU already lost to Penn State 45-14 back in September), hope USC gets the SEC loser. Give the Trojans an Alabama or Flordia match up, it would be as big as the title game itself. Not Pac-10 vs. SEC, but USC vs. the SEC. Let’s end this once and for all.

5 responses to “USC: 30% Chance of BCS Title Berth”

Math Guy

November 24, 2008 at 11:49 am

Here’s why you are wrong.

If each game had a 50/50 shot of going one way or the other, then your inference would be correct.

Although I am dubious of the count. Can it really be that 1/2 of the outcomes of the 64 games leads to USC in the BCS Bowl? What else does that assume? For instance, would a one loss Alabama team get knocked down lower than 6. I guess that it would, but I doubt that anyone can know for sure, since voters will no doubt think a close game would suggest keeping them high and the computers might disagree as well.

The other guy has made edicts of what he determines to be the probabilities of the game outcomes. How did he do this? I think that the most scientific way would be to imply odds from the spread, but it doesn’t look like anything scientific was done here. Instead, it’s making things up. I mean Auburn probably doesn’t have a 50/50 chance of beating ‘Bama, but neither can it be said to be 30% either. That’s simply numerical hocus-pocus.

Reply
Ted

November 24, 2008 at 1:25 pm

TT, UT and FLA have close to 90% chance of beating Baylor, A&M and Fl St. respectively. And I would put Mizzou’s chances of knocking off either OK or UT at no better than 20%. SC’s chances of getting to the NC game are far less than 30%.

Reply
Sam

November 24, 2008 at 9:18 pm

Either way you look at it, USC’s computer rankings are really low (currently at 8), and given that our next opponents are mediocre at best (UCLA and Notre Dame), it’s going to take much more than those few upsets to get us into the NC game, even if we do get into the top 2 in the AP and coaches’ polls.
Sigh. If only we had adjusted faster in that Oregon State game, we wouldn’t need to worry about all this.
Fight on!

Reply
Reader

November 26, 2008 at 11:01 am

Obviously, nobody here has read the original article or the two preceding articles breaking down all the assumptions and mathematics. If anybody had bothered to look at their work, it’d be clear that these guys have done their jobs right. I’m pretty sure their idea is just to get an estimate anyways, as any probability they come up for a sports matchup has to have some sort of estimation involved.

Reply
Reader

November 26, 2008 at 11:02 am

Obviously, nobody here has read the original article or the two preceding articles breaking down all the assumptions and mathematics. If anybody had bothered to look at their work, it’d be clear that these guys have done their jobs right. I’m pretty sure their idea is just to get an estimate anyways, as any probability they come up with for a sports matchup has to have some sort of estimation involved.

Reply