Conference Tournaments and the Upset

Many of us cry when this guy does this thing

College basketball fans differ on their favorite part of the season: is it the NCAA Tournament itself, 64 (sorry, 68) teams going sudden death to determine the national title? What about BracketBusters, seeing the best mid-major teams go head-to-head, each game having a potentially major impact on their resume? While I do love the 3 weekends in March and April that result in college basketball's national championship, my favorite are the two weeks before then--the conference championships. In every conference except the Ivy Leagues, the majority of teams are given one final chance to make it to March Madness, the crowning achievement in many of the smaller leagues. To get the opportunity to play one of the big guys, to have their One Shining Moment, yada yada yada. 

What makes these tournaments more interesting is that they're the only way to get a reliable ranking for a team--while the AP and ESPN/Coaches' polls are nice, they're often biased in some way or another, by some means of horrible voters. However, over the course of a 16-or-18 game conference schedule, the good teams tend to rise to the top while the lesser teams sink to the bottom. For most of these conferences, the winner of that tournament is the only team to make it to the Dance, and therefore the conference tourney takes on more importance than any other week that season. What's incredibly interesting is how often the lower-ranked teams turn it on in those ultra-important tournaments. Hit the jump to find out:

Now that's what we're talkin about

I analyzed every conference tournament from 2006-2011: a total of 1,456 games pitting a lower-ranked team against a higher ranked team. Now, I didn't take AP or Coaches polls into consideration, letting the regular season decide solely which game was an upset and which game was not. Why did I do this? To eliminate luck. Could a higher seed have been a not-so-great team with a lucky schedule and a few injured opponents? Of course, just as a lower seed could have been a talented squad that lost a player for some time, only to get him back just in time for when it counted. So, in those 1,456 games, how often did the lower-ranked team win? 452 games--or 31% of the time.

I then broke it down into three categories: high-major (the "BCS" conferences), mid-major (C-USA, Horizon, Atlantic 10, Mountain West, Missouri Valley, CAA, WCC, and WAC), and low-major (everybody else). The BCS conferences had by far the highest percentage of upsets--out of the 389 tournament games since 2006, 36% of them, 139 games, had resulted in the lower seed beating the higher seed. The "mid-major" conferences had an upset in 127 out of 413 games, or 31%, while the "low-major" conferences had the lowest percentage at 28% (383 out of 654 games). Why is this? Well, the most obvious reason is where these tournaments are held. The major conferences are allowed to hold their conference tournaments in fairly neutral locations, like the Big East in New York City or the Big 10 in Indianapolis. The smaller conferences, unable to afford those sorts of arenas, usually host their tournaments on campus sites, eliminating that neutral-court bias.

So, what conference do you least want your team to be the favorite in? How about the Big East conference, whose monsterous tournament over the last 6 years has seen 42% of it's results going to the underdog. This is a tough number for many to swallow--their teams can have fantastic regular season, but have a 42% chance of losing a tournament game that could send them to the NCAAs? Luckily for the Big East members, a top seed in the conference tournament usually means an upcoming bid to March Madness anyways.  The conference where higher seeds are safest? You would think it would be one of the low-majors, but it's actually the Horizon League, which plays in a mix of neutral and non-neutral locations. The league of Butler, Valparaiso, and Cleveland State has only seen 18% of its conference games go for upsets in the last few years.

This year, 36% of all conference tournament games were upsets, the highest of any period in the study, an interesting note in a year where the "bubble" seemed to be the softest it had been in years. The safest year in recent memory was in 2007, when just over a quarter (56 out of 219) of the games were upsets.

So what does it all mean? When it comes to conference tournament week, if you haven't picked at least 1/3 of the higher ranked teams to lose...you're overestimating the power of the regular season. That bid to March Madness can cause some crazy things to happen in those few games.