A Substitute for War

Basketball philosophy

March Madness as a Playoff System

with 10 comments


Image via Wikipedia

I’ve previously analyzed the playoff systems of the 4 major professional sports leagues in the US, looking at fairness, which I’ve defined as follows:

Ideal fairness means that we get rid of the unevenness of the regular season schedule without adding too much randomness.  If you’ve got a variety of divisions or conferences that hardly play against each other, the idea that you can have a single champion without a playoff tournament of some sort is absurd – but of course playoffs in some sense always mean throwing out a larger sample size for a smaller one, which never entirely good.

We’re in March Madness season so it’s worth considering college basketball’s playoff system, arguably the most successful in terms of financial gains relative to regular season. This happens to be a particularly good season to consider this because all of the favorites are gone. Every team left has at least 8 losses, which either indicates a stunning amount of parity, or a ridiculous amount of luck.

The 4 Pro Leagues

With my previous analysis, I went back to 1995, and tallied how dominant the top regular seasons were in the regular season and how they did in the playoffs. There were a couple charts in the post, but I’ll include one here that gets at the crux:

At the professional level, football and basketball are significantly more able to be dominated by superior teams than baseball and hockey, and this carries on in to the playoffs. It’s safe to say that there is just far more luck in baseball and hockey, and this means that a sample size-reducing playoffs has a danger of decreasing the merit of the eventual champion.

Noteworthy also is that while football appears even more able to be dominated by superior teams than basketball based on the regular season, come playoff time that reverses. I argue this has everything to do with the fact that football plays single elimination playoffs, thus reducing sample size even further than other playoff systems, and reducing merit. Even so, the NFL’s champion merit does quite well compared to leagues other than the NBA, so they seem to have a pretty good balance between fairness and the excitement that gets maximized when every game is a must win for every team involved.

A Season of Madness

So here we go, doing the same thing for March Madness the same time span using the Final AP polls on Sports-Reference, and our new table:






What do we get? That the top college basketball teams dominate their peers significantly more than the analogous teams in any professional sports league, but they are less likely to win the playoff title than even the kings of randomness in baseball.

Now, obviously my analysis is quick and dirty. There are things that can be done to make a more precise and thorough analysis, but I can’t anyone finding this conclusion all that odd. The event is called ‘Madness’ and forces a team to play 6 win-or-go-home game, of course that’s going to up the randomness.

In terms of performing the initial goal of creating a fair playoff system, March Madness fails horrendously.

March Madness: The junk food of American sports playoff systems

And yet people love it. Even more interestingly, every year college football gets bashed to no end because they don’t have a playoff system, but can there be any doubt that the best college football team is more likely to be crowned champion than the best college basketball team? I’m all for a college football playoff system, but clearly the notion that March Madness gives us a true champion and college football does not is irrational. But again, people love it, why?

Well, one aspect of it of course is that March Madness is embraced by a lot of people who admittedly know nothing about the sport. We could expect for such people to really not have any sense that they are watching inferior teams get crowned with superior accomplishments. However, the hard core watchers of college basketball don’t seem to be bothered by the randomness of it either. Does that mean we as human just don’t really care about whether the competitions we watch actually say anything about who the better competitor is?

I don’t think it does. Why would people get so up in arms for the lack of playoff system in college football if they didn’t care who was truly better? Why would we be so dissatisfied with tie matches if we didn’t think that winning the match meant something?

I think instead what it means is that people really do care about merit, and they just are weak at judging whether something really measures merit. If you could put an event together that superficially looks good – no obvious biases for or against a certain team, no glaring errors by referees – people are fooled into think the results are meaningful, even if more analytical scrutiny will quickly reveal otherwise.

One other takeaway from this for me, is just how silly it is to think that reasoned arguments would have an effect on decision makers even if the decision makers were completely sold on the argument. Last fall, I wrote a piece on why I think it’s a bad idea to expand the MLB playoffs, because I know it will only make the playoffs less likely to crown the best team as champion. However, if people aren’t complaining about March Madness, the most randomness-inducing event you’ll see in major American sports, adding another round of randomness to baseball is unlikely to make a dent on the public consciousness. Thus, it behooves Bud Selig to just do whatever will up baseball’s revenue.

Ah well. Not all is doom and gloom. I’ll be watching the Final Four with enthusiasm ignoring the analytical voice in the back of my head, just as I do all the time when it tells me to not do something fun. So here’s to March Madness, the junk food of American sports!

10 Responses

Subscribe to comments with RSS.

  1. Problem with this calculation: you didn’t adjust for the number of teams involved in the leagues. The top 2 in the NCAA is just the top 0.57%, and the next several percent are just about as good. Perhaps instead of the top 2, you should look at the top x%? Or something like that?


    March 30, 2011 at 2:18 pm

  2. This is a solid point to bring up.

    I would argue that that the focal point here behind what fraction of the “league” that should be looked at should be ideally neither a blanket top 2 nor a top X% but based on how much uncertainty we have in who really has been the top teams during the season.

    In college football, it’s perfectly reasonable to assert that this means we at least need the impressive champs from each conference, but in basketball, I don’t know if we even need that because there are plenty more games in the regular season.

    Here’s what I’ll concede though, an argument that using the top 4 teams instead of the top 2 is more reasonable is understandable. And if we went by that, the probability jumps to 64.7%, which would yield a significantly different conclusion, when compared to the numbers of the other league.

    At the same time though, if I let the NBA go by their top 4 teams, we know that result would jump a good amount too (we’ve only had one non-top 5 team win in the NBA in the time frame I’m looking at). So let us not doubt that the randomness of March Madness is inducing far more randomness than the NBA playoffs.

    Even if the March Madness championship still can be argued to have more meaning than something like the MLB playoffs, I think it’s still pretty clear that the having of the tournament does not get us significantly closer to understanding who truly was the better team.

    Matt Johnson

    March 30, 2011 at 2:38 pm

  3. I think the high win percentage of the 2 top teams in the NFL may be because of the small amount of games. Compare this to taking the highest batting avg of the top two players in a 30 player population after 16 or 82 games. The outliers will regress as the amount of games increase.


    March 30, 2011 at 2:49 pm

    • This is an excellent point. I’ve thought about that, but I don’t see obvious signs that this is clearly the case.

      If you look at my playoff systems post, you’ll see a round by round look at results in the NFL playoffs. In their first playoff game against a top 8 team, they win more than 70% of the time, and have a gradual decay down to the low 60s by the Super Bowl.

      All that seems pretty much like I’d expect from a team whose 83% winning percentage against average teams was legit.

      Matt Johnson

      March 30, 2011 at 2:56 pm

  4. […] March Madness as a Playoff System (asubstituteforwar.com) […]

  5. […] matchup circumstance and just sheer variance. Instantly replaying the NCAA Tournament — a highly volatile event — would likely produce a different winner. My guess is, the average sports fan would be […]

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 )

Google photo

You are commenting using your Google 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 )

Connecting to %s

%d bloggers like this: