Will the tourney keep getting crazier?

The 2010, 2011 and 2012 tourneys have marked the maddest three-year stretch of the 28-year, 64-team era. They came hard on the heels of the sanest three-year period of the NCAA tournament. How do we know that 2007-09 was the chalkiest and 2010-12 was the craziest stretch of dances in the modern era? Because we devised a simple statistic to measure the madness. I call it the “Madometer.”

The Madometer works by measuring the seed-position differences between actual winners and perfect high-seed success or failure in all six rounds of the dance. If the higher seed advanced in all 63 games of the tourney (perfect sanity), the cumulative seed value of the winners would be 203. If the lower seed always advanced (sheer madness), their cumulative seed value would be 868. The difference between the two—665—is the predictability range.

Let’s take a closer look at the 2012 dance. If you added up the seed positions of all the teams that advanced through the tourney, the number would come to 317, certainly closer to perfect seed dominance (203 positions), but still 114 positions toward madness along the 665-point predictability range. That works out to a Madometer reading of 17.1%. To put that number in context, the average tournament over the 27 years of the 64-team era has deviated from by-the-seed results by 14.1%.

That makes 2012 the seventh most unpredictable tourney since the field expanded to 64 teams. The craziest dance occurred just last year, when tourney advancers deviated from perfect high-seed dominance by 19.8%. And the year before that, in 2010, saw another 17.1% mad tourney. Here’s how the last three years stack up against the other 25 dances on the Madometer:


Since 1985, the 64-team tourney has ranged from 4.1 percent “John-Woodenesque” high-seed predictability in 2007 to 2012’s 19.8 percent “Jimmy Chitwood-like” darkhorse craziness. (Chitwood’s the guy who famously sunk the last-second jumper that gave little-known Hickory High its upset win in Hoosiers. His name is actually Marias Valainis. But I digress).

Looking at the Madometer readings, it becomes abundantly clear how unique the 2007 dance was for its by-the-seed results. The tourney deviated from perfect seed supremacy by a scant 4.1 percent (27 of 665 possible low-to-high-seed differences). And just two years later, after the only dance where all four top seeds reached the Final Four, the 2009 tourney logged the second-most predictable Madometer reading.

Then, after three years of yawn-inducing dances, the tourney turned wildly unpredictable. The closest three-year period to the unpredictability we’ve seen most recently is 1999-2001. Over that time, the average Madometer reading was 17%; since 2010, the gauge has averaged 18%.

It would be one thing if the modern dance were reliably unpredictable; then we would know to take more chances in our bracket picking. But what has been going on since 2006 has been erratic—call it unpredictably predictable. We’re in the sort of period that drives tourney pool players nuts: from one year to the next, we don’t know whether to go chalky or go rogue. Put it this way, if you look at the deviation between one tourney and its predecessor, the three biggest swings between madness and sanity occurred over the last six years. From 2006 to 2007, the Madometer shifted 13.3 percentage points, then lurched 8.2 percent, before a slight downturn from 2008 to 2009, followed by another jolt of 7.9 percent, then a bump and dip of 2.7 percent. Overall, the deviation over the six-year period averaged 6.3 percent. Compare that to the consistent 2.8 percent deviation during the next craziest era of the dance, from 1997 to 2003.

If all of that just hurt your head, there’s a more basic way to measure the madness. You simply tally the number of upsets in each tourney. I define an upset as any game in which a longshot knocks off a favorite that’s seeded four or more seed positions above it. So a 10 beating a seven in round one isn’t an upset; but a five seed beating a top seed in the Elite Eight is.

By this definition, there were nine upsets in 2011. That’s slightly more than average, but it falls four shockers short of the record. The 2012 tourney saw 13 upsets, tied for the highest total with the 1985, 1986, 1990 and 2002 dances. Here are the nine shockers from last year:

  • 15 Norfolk State beat 2 Missouri in round 1
  • 15 Lehigh beat 2 Duke in round 1
  • 13 Ohio beat 4 Michigan in round 1
  • 12 VCU beat 5 Wichita State in round 1
  • 12 South Florida beat 5 Temple in round 1
  • 11 Colorado beat 6 UNLV in round 1
  • 11 North Carolina State beat 6 San Diego State in round 1
  • 11 North Carolina State beat 3 Georgetown in round 2
  • 7 Florida beat 3 Marquette in the Sweet Sixteen

While there weren’t as many upsets in 2012 as 2011, two of them were arguably more unlikely. Considering that 15 seeds were 4-104 against two seeds before last year, having both Duke and Missouri go down in the same dance was extraordinary. And it did serious damage to most people’s brackets.

Since 1985, the average tournament has seen 8.6 upsets. The three years before 2010 were remarkably upset free, with just three Cinderellas in 2007, eight in 2008 and five in 2009. Take a look at the rest of the upset totals throughout the 28 years of the modern era:


The 2009 tourney marked the sixth dance in seven years of below-average upsets. So 2010-12 marked a big trend reversal. Will we see more of the same in 2013? Before the 2010 dance, I stumbled upon a more concrete way to assess whether a dance would be unpredictable or not. I compared the Pythag efficiency values of the top 13 seeds in the tourney field to the Pythag averages of the top 13 seeds over the previous seven tourneys. (For those who don’t know about Ken Pomeroy’s possession-based Pythag data, check out www.kenpom.com.) What I found was that the higher performing seeds over the last three years were weaker on average than previous elite seeds. Take a look at this curve comparing the quality of seeds in the 2012 dance to the average seed quality since 2004.


At every seed position except eight, 12 and 13, last year’s teams were weaker than previous years. One and two seeds were slightly softer than their counterparts from 2004 to 2011, and three seeds were considerably less efficient. Is it any wonder that not a single three seed reached the Final Four—and only one (Baylor) made it to the Elite Eight.

The interesting thing about this curve is that three of the seeds that we markedly weaker than their counterparts all faced each other in opening round mini-bracket. The three, six and 11 seeds last year were very soft…and the results of those match-ups confirmed the weakness. Instead of four three seeds emerging from the 3|6|11|14 brackets, we saw two threes, a six and an 11 (North Carolina State).

Meanwhile, the strong 12 seeds split with the comparatively weaker five seeds in round one. And solid 13 seed Ohio squad took down a soft four seed in Michigan. How to explain the two 2v15 matchups? No numbers pointed toward the Norfolk State upset of Missouri. Yes, the Tigers were a weaker than usual two seed, but the Spartans were a dreadful 15 seed. Their .3808 Pythag was worse than the average 16 seed. On the other hand, Duke was a markedly weaker two seed with just the 17th best Pythag in the dance last year, which means they could’ve been a five seed. And Lehigh owned the 56th best Pythag in the tourney field…so they could’ve been a 14 seed. In hindsight, that upset doesn’t seem quite as shocking as it did when it was going down.

Duke and Missouri aside, most of the ones and twos moved on, not so much because they were strong, but because their competition was so weak compared to their historical quality. And the winner of the dance? Kentucky wasn’t just the best one seed last year; they we’re also solidly better than the typical one seed.

My takeaway from this 2012 chart—which is similar to the 2010 and 2011 charts—is that Pythag can help you determine up front whether a tourney field is weaker or stronger than average. And if the one-through-six seeds are predominantly weaker than their historical counterparts, it’s more likely that the tourney will be unpredictable. That will at least guide you in whether you take a chalky approach to your bracket picks or go out on a limb.

The 2013 tournament could be the same story. I’ll monitor the top teams throughout the season to see if they’re as dominant as top teams in the past. If we don’t see any squads separating from the pack…while the pack gets more bunched up in terms of the efficiency numbers, I’m betting that we’ll have another mad March Madness.

This entry was posted in Measuring Madness, Tourney Trends. Bookmark the permalink.

5 Responses to Will the tourney keep getting crazier?

  1. Gary Diny says:


    I was wondering if you have the abiltiy to load in the Pythag number into Bracketmaster. Looking at the last few years that you have been tracking the Pythag numbers (7 years I think) and if there is any connection in predicting upsets in the seed by seed matchups for respective rounds. Just hoping to gain more predictive insight into where those upsets are coming from.

    Thanks for all the great work.


    • ptiernan says:


      I have input nine years of Pythag info (including Def and Off efficiency) into my database…but it hasn’t found its way into the Bracketmaster. With all the changes we made to the site this year, that will probably have to be a post-2013 upgrade. But I will a series of articles assessing the value of possession-based stats in predicting both upsets and toss-up games (where the seed gap between opponents is less than four).


      • Gary Diny says:


        Thanks for the update. As long as the GB Packers are in the NFL playoffs, I have some other things to distract me from NCAA hoops. But I will keep up to date on the happenings here as well.


        • ptiernan says:

          Well…I am (or was) a Lions fan. March Madness can’t come soon enough.

          • Gary Diny says:


            My condolences on the Lions, I just hope they keep the coach…means one less team for Green Bay to worry about in the division.

            March Madness will be here soon enough, currently waiting out 12+ inches of snow forecasted for tonight/tomorrow. Joy of winter.


Leave a Reply

Your email address will not be published. Required fields are marked *

* Copy This Password *

* Type Or Paste Password Here *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>