BCS Rankings: A Closer Look at Strength of Schedule, Part II

The Florida Gators are One of Several Teams Whose Strength of Schedule Provides a Big BCS Advantage

In part one, we looked at the undefeated and one-loss BCS conference football teams and compared the win percentages of their opponents so far to estimate strength of schedule. However, we came up with some results that just didn’t look right. I mean, Texas Tech at no. 1, Oklahoma at no. 5 — ahead of both Alabama and Oregon. What’s wrong here?

If you looked carefully, you may have detected the problem already — we are giving Texas Tech and Oklahoma credit for opponents they lost to! Why should the Red Raiders be able to count Oklahoma’s five wins if they weren’t able to beat the Sooners head-to-head? Why should the Sooners take credit for Kansas State‘s 7-0 record?

So we need another table, this time only showing winning percentage of the opponents each teams actually beat.

That’s better; at least an unbeaten team (Florida) is back on top, though the pesky Red Raiders are still hanging out in second place. Alabama and Oregon move in front of Oklahoma in this list as well, which is intuitive. Louisville remains at the bottom — the Cardinals haven’t lost, but they haven’t really beaten anybody either. Ohio State and Kansas State both move in front of Cincinnati, since the Bearcats are not allowed to count Toledo‘s seven wins.

There is just one last issue with this chart: it addresses quality of wins, but it doesn’t really address how bad the losses are. So let’s do one more thing — let’s add the losses for any opponents that a team lost to (call it “my enemies losses”). That way a team still gets no credit for the wins of a team that beats them, but they are saddled with the losses if they lose to a team that is not undefeated themselves. With that in mind, here’s the final table:

There you have it. The two teams who have proven themselves against the strongest competition so far? Florida and Notre Dame. Does that sound reasonable to you? It makes sense to me, and what is more, I wouldn’t be surprised at all if those teams play for the BCS championship when all is said and done, though there is still a lot of football to be played.

What about the bottom? Are you surprised to see Georgia there? The Bulldogs have really only played one good team so far — South Carolina — and they lost to them. Georgia gets another crack at the big stage when they play the Florida Gators this weekend… but something tells me that will only eliminate the ‘Dogs from this list when they receive their second loss.

USC placed below Clemson in strength of schedule? Well, the Trojans lost to their toughest opponent so far, in Stanford. Their second toughest opponent? It’s a tie between Syracuse and Washington. No offense to our future conference mates, but when the Syracuse Orange are the strongest team you’ve actually beaten, that isn’t a strong resume (this year).

This list just wouldn’t be believable without LSU ahead of Cincinnati! This last tweak fixes that, too.

Does this mean these teams will finish in the order shown? Not at all. This is only showing the road traveled so far. For example, Clemson and Florida State both have played some very week non-conference teams, but both have a chance to move up in these charts because their remaining opponents are stronger than those they’ve already beaten (with the exception of when they played each other).

But if Alabama and Oregon have not played the toughest schedules, why are they ranked so high? And why is Cincinnat still ranked above FSU and USC? Those are questions for another article…

Read more from Hokie Mark over at ACCFootballRx, where he gives his prescription for fixing what ails the ACC on the gridiron.

About these ads

6 thoughts on “BCS Rankings: A Closer Look at Strength of Schedule, Part II

  1. The one note I will add — and this is something that I’ve actually been trying to create a metric for myself — is that the opponent win totals can still be a bit deceptive, too. In Rutgers’ case, they get a boost from Howard’s five wins, even though they’re a FCS squad. You almost need a way to measure the quality of your opponent’s opponents to really capture it. Washington and Syracuse, as you mentioned, are each 3-4. But take a look at the teams each have faced. UW’s met with Oregon, LSU, Stanford and SC (and went 1-3, unsurprisingly). SU has played against SC, Northwestern, Minnesota and Rutgers (and went 0-4 against those opponents, but only went down by an average of a touchdown per game). Point is, just like not all 6-1 records are alike, not all 3-4 opponents are either.

    • That’s exactly right, John. In order to measure the true “power” of a team, you need to consider things like the opponents of your opponents, margin of victory, etc. The math can get very hairy, very quickly. Perhaps we’ll take a layman’s view of power ratings later.

      Of course, in the case of the BCS poll, there is another insidious factor – preseason rankings. It’s “the elephant in the room” that the guys on ESPN SportsCenter don’t want to talk about… but hopefully WE can address that in a future article, too!

      • Definitely a key factor. And in many ways, we need it to go away. Though after 75 years of polling, I sincerely doubt it’s ever removed from the game — in preseason or otherwise.

        What I am curious about is how the playoff selection committee will present their
        “teams in contention” week by week, once that system starts in 2014. They’re not using BCS rankings (though I wish they would use some computer metric), so curious to see what type of SOS-related formula they put together to come up with an RPI-type number for teams. And how weighted will aspects of it be? We’ve obviously got time for al that, but fun to think out a couple years, in any case.

    • Thanks for the link. I guess what I find disconcerting there is the de-emphasis of strength of schedule. Several teams with terrible SOS numbers ranked pretty high. In an ideal world, metrics would account for these issues.

  2. All of this (and more) point to why it is so difficult to rank teams against each other. There are so many possible factors to consider. A game between an FBS and FCS team is more or less like an MLB team playing against a AAA team … you’ll occasionally have upset wins by the FCS teams, but not too often. A 5-2 FCS team vs. a 5-2 FBS team are not nearly the same thing.

    I definitely appreciate the insight on the opposing winning percentages though. There’s no “true” way to analyze which team is the “best” team. It’s ultimately an interpretation of what factors are more important than others.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s