Saturday, November 22, 2008

Let's All Pile On Bob Kravitz And the BCS

Howdy. Two posts in two days--still too early to call it a streak, but a good trend nevertheless.

Zinglebert beat me to the punch with the Bob Kravitz college football playoff article, but I thought I'd add my two cents (Yes, I could have done this in the comments section, but this will probably be too long and I needed something to post. So there.).

I agree wholeheartedly with Zinglebert's notion that Kravitz has the right conclusion but idiotic reasoning. I actually agree with Kravitz on this issue: I'd rather not see a playoff in college football, either.

God, I feel dirty admitting that I agree with Kravitz. And it gets worse: I agree with his general idea that playoff systems can (and often do) devalue the regular season. Of course the regular season still matters--you have to play well in the regular season to first make the playoffs--but once the playoffs start, the regular season doesn't matter at all. Oh, sure, it matters for playoff seeding, but how much has seeding mattered in deciding the championship? Using the last year as an example, only the NBA had the best two teams in the regular season meet for the championship:

NHL Stanley Cup Finals: Detroit (#1 seed in West) over Pittsburgh (#2 seed in East)
NBA Finals: Boston (#1) over LA Lakers (#1)
Super Bowl (NFL): NY Giants (#5) over New England (#1)
World Series (MLB): Philadelphia (#2) over Tampa Bay (#2)

Looking at larger trends: in MLB, the overall best record in the league hasn't won the World Series since 1998, and only 4 times in the last 10 years has a #1 seed made it to the World Series (they are 2-2 once they get there). In the NFL, the last time the best team in the regular season won the Super Bowl was 2003 (New England). A #1-seed has made the Super Bowl every year since, but has lost.

The playoffs are the ultimate "small sample size" argument: any team can beat another team on a given day. Yes, the New York Giants beat the Patriots in the Super Bowl, so they were the better team that day. However, a reasonable person would not say that the Giants were a better team than the Patriots last season--in fact, the Patriots beat the Giants 5 weeks earlier. By the time the Patriots and Giants met for the rematch on February 3rd, the regular season meant squat.

Zinglebert makes a valid point that in the regular season, good teams don't always play the other good teams. And that sometimes makes it hard to determine who the "better" team is--unless both teams play identical schedules (and they never do), then a head-to-head matchup helps decide a championship.

Back to Kravitz. Zinglebert already dissected the article, but some parts are worth repeating:

Every week, every game involving a Top 25 team is, in its own way, a playoff game. It's not necessarily a one-and-done, but it serves essentially the same purpose.

One of the reasons college football's popularity has soared in recent years is because every regular-season game matters and matters deeply. When USC loses early in the season at Oregon State, it's tantamount to a first-round playoff loss. When Texas Tech knocks off Texas, it's a playoff game.

As we've noted a few times now about Kravitz, he has the uncanny ability to take the correct position but use entirely crap reasoning to get there. Every game involving a Top 25 team in college football is NOT "essentially a playoff game." It depends on when the loss is in the season--teams can recover from early losses, depending on the outcome of the rest of the season.

And I love how he uses Texas as an example of a team that was knocked "out of the playoffs" by an earlier loss to Texas Tech--thanks to the 65-21 beat down of Texas Tech by the Rodgers & Hammersteins tonight, Texas may very well end up #2 in the Bowl Championship Series (BCS) standings!

I like the fact that with one bad performance, Penn State's claim to a national title is all but dead.

So, isn't that the exact definition of a playoff: one bad performance, and you're out? Penn State's national title hopes went down not just because of their loss, but because a) they lost after USC lost and b) no one above them has lost since (until tonight). USC lost to Oregon State earlier this season, but USC still has a remote chance of playing for the national title.

It's difficult to quantify, but there is surely value in the continual "who's No. 1?'' debate in college football.

Sure, I suppose that's true in the sense that it keeps people interested. But that doesn't mean that the current system that's in place is automatically the best system.

The best thing about the BCS is that the regular season matters--Kravitz and I agree on that point. However, the BCS has one giant flaw: people. Specifically, the people who vote in the polls.
The BCS standings are made up of three components: a media poll, a coaches/former coaches poll, and the average of six computers that analyze opponents, schedule strength, etc. In other words, two-thirds of the BCS is subjective.

Why are the polls such a problem? The main reason is because the polls punish teams who lose late in the season more than the teams who lose early on. An example: Florida and USC lost the same week in late September: Florida to Mississippi 31-30; USC to Oregon State 27-21. But Penn State lost to Iowa 24-23 on November 8. Florida, USC, and Penn State each has one loss, but Penn State is #8 in the BCS, USC is #6, and Florida is #4. Why? Mainly because Penn State's loss came in November, and they won't be able to go back up in the polls unless a miracle happens and 5 or 6 teams in front of them lose.

In fact, Florida is 4th in the BCS because of being ranked 3rd in both polls. The computer rankings have Florida 5th. Texas, who has all sorts of quality wins and whose only loss came in the final seconds against Texas Tech, is somehow behind Florida in both the Harris and USA Today polls. Again, Texas lost later than Florida, so Florida has the poll advantage.

A further problem is that the system is biased towards the traditional football powers and the teams that are expected to do well going into the season. The preseason polls almost always feature the big names: Florida, Alabama, Texas, Oklahoma, USC, Michigan, Ohio State, etc. Other teams who aren't the "name brand" variety have to start lower down in the polls and work their way up. The poses two problems: the first is that a loss for these schools early on effectively ends any BCS hopes, the second is that even if one of these teams goes undefeated, it can only reach the championship game if every other team loses.

Take Utah as an example. The Utes are #7 in the Harris Poll and #8 in the USA Today poll. However, Utah's computer average is #4--meaning that 6 computer analyses think Utah is the 4th best team in the land--but they won't get a sniff at the championship game because of the polls. And, since 12-0 Utah has already finished its season, it will most likely go down because it doesn't have another game, and other teams have potentially two games remaining due to conference championships. Had Utah been given any respect at the beginning of the season, perhaps it would be #4 right now, and in a good position to sneak into the championship game if Texas and/or Florida falter. As it stands, Utah is behind six teams, so no chance.

Am I saying Utah deserves a shot at the national title? Maybe. The Utes are 12-0, after all. They played Michigan, who sucks this year, but it's not like Utah knew that when it made the schedule. But then I'm reminded that the 6 independent computers have Utah #4 (heading into today's games), so I'd say no to the national title chance. If Utah ends at #1 or #2 in the computer rankings, then I'd say yes.

Why do I put so much faith in the computer rankings? For one, the computer rankings are objective. There are no "style points" in the computer analyses; for BCS purposes, margin of victory is not factored in to the computer ratings (so Florida's 70-19 win over The Citadel today may impress pollsters, but the computers will ignore the final score). The other reason I like the computer rankings is because each computer factors in the entire season, so any loss has equal weight no matter when it occurs. Strength of schedule, opponents' strength of schedule, road wins, etc. are all factors in the rankings. In other words, the computer will try to equate Utah's 12-0 record with Florida's 11-1 record despite the teams having no common opponents. I believe this to be a more reliable method than some guy who writes for the Chokoloskee Courier-Sentinel-Examiner in Florida (note: paper may not actually exist) who is lucky to see Utah play once during the season making an "objective" analysis of who's the better team based on what he's seen on SportsCenter. Would it really be surprising who he'd pick?

The coaches' poll is no better. Coaches are busy during the season, so they don't have time to watch all of the teams they need to watch. And coaches are also known for their biases, too--most notably, voting from teams from the coach's own conference.

It's obvious that the polls are the biggest problem with the system, yet it seems no sportswriter is able to recognize that--or, if he or she does recognize it, isn't willing to say so. Could it be because the sportswriters like to feel important? I'd say that's a pretty safe bet. If the polls were removed from the BCS equation, the sportswriters would lose their influence on the outcome. If we've learned anything about sports writers and reporters over the years, it's that they've always managed to overestimate their importance.

Think about it: almost all of the talking heads who have been screaming for a playoff system to replace the BCS want the teams selected for the playoffs to be determined by the polls--the subjective part of the system. What these people forget is the BCS was instituted in part because fans were upset with the system in place before--namely, no system other than the polls!

The BCS is certainly not perfect. Perhaps a playoff system will come to major college football someday. Until then, the BCS should improve itself by dumping the polls and letting the computers do the work. Since computers don't have feelings, they won't show bias for a particular team. They also won't care about any criticism, either.

Labels: , , , ,


Blogger Zinglebert Bembledack said...

Good points, Slut. I did not get into the argument of the polls and how losses at the end of the season are somehow different that at the beginning of the season because of the "human" element polls, so I am happy you did.

I am conflicted as to whether there should be a playoff or not. On one hand, a playoff can give you a more definite champion, but just like other playoffs, but as Slut said, the "best" team does not always win the playoffs. What most people do not seem to comprehend is that baseball and basketball teams can play multiple games in a day or on back to back days. Football teams need that week in better to recover and nurse their wounds. Adding a playoff can suddenly force teams to play 15-16 games a season plus the added cost of travel for teams and fans.

Gone are the days when playoffs were a matchup of the two best teams between leagues. The World Series up until 1969 was the example of what playoffs were meant to be, a series between the best two teams of the rival leagues that did not play each other during the regular season. Now we have TV deals and greedy owners that try to get the most bang for their buck and squeeze the most they can out of the playoffs.

Sometimes is does make for a more enjoyable show, but it does not typically reward those teams that did their job during the regular season.

There is no right or wrong answer as far as a College Football Playoff goes, but the BCS system is flawed and will not fully work until the sportswriters and coaches polls are taken out or reduced in value. Until then, at least it gives us fodder for the website.

November 26, 2008 at 10:59 AM  

Post a Comment

Subscribe to Post Comments [Atom]

<< Home