Wednesday, March 19, 2014

NCAA March Madness Time!

Tomorrow is one of the best sports days of the year as the NCAA men’s basketball tournament begins. Usually I just focus on upsets, good games, picking the best bracket, etc. However, this year this year there is an extra incentive, as UVa capped off their ACC regular season championship with an ACC tournament title, earning a #1 seed in the East region. It was the first ACC regular season championship outright and #1 seed for the Cavs in over 30 years, and the first Tournament title since 1976. Now, do I follow my heart and pick the Hoo’s to win it all or do I listen to the experts and try to be more accurate with my bracket predictions?

First let’s take look at the trends from the past 28 seasons by round.

First round:

  • #12 seeds upset the #5 seeds 36% of the time (last year it was 75% with #12 seeds Mississippi, California, and Oregon pulling off the upsets) 
  • 21% of #4 seeds go down (#4 Kansas State went home early last year) 
  • Meanwhile only 14% of #3 seeds lose their first game (last year #3 New Mexico was upset though)

Second round:

  • Only 12% of #1 seeds were upset (#1 Gonzaga was upset by eventual Final Four team Wichita St, who happen to be undefeated this year)
  • Meanwhile 36% of #2 seeds have been upset by the #7/#10 winner (#2 Georgetown was upset by the #15 last year, not even getting to the #7/#10) 
  • Only 12% of the time has a region gone #1, #2, #3, #4 (which actually did happen once last year)

Sweet 16:

  • 70% of #1 seeds make the elite eight (was 25% last year with only champion Louisville advancing) 
  • On the other side of the region, 72% are either #2 or #3 seeds (100% last year)


Elite Eight:

  • Only one time in 33 years of the 64 team bracket has all four #1 seeds made the Final Four (which was no exception last year) 
  • 41% of #1 seeds make the Final Four (only one #1 seed made it last year) 
  • #2 seeds make it 21% of the time (no #2 seeds made it last year)

If you add up all the seeds that make the final four, the average total is 10. Meaning if you pick all #1 seeds, that’s less than half the average, or if you throw in a #11 George Mason with a #3, #2, and #1, that’s double the average. (Last year’s total of 18 (#1, #4, #4, #9) was way above average.)

If those are the yearly trends, then who did the experts pick this year? Here is a breakdown of 22 “expert” brackets.

Championship:

  • 32% of the experts chose #1 seed Florida to win the entire thing (compared to 75% for eventual champ Louisville last year)
  • UVa was the only #1 seed not picked of the six different teams
  • 27% picked #4 Michigan State with 8 losses while only 9% picked undefeated #1 Wichita State

Runner Up:

  • 7 different teams were chosen to lose the title game
  • 31% #4 Louisville was the highest
  • 22% have Florida over Louisville

Final Four:

  • 86% picked #4 Michigan State to reach the Final Four, while only 1 expert picked UVa
  • Only 27% picked #1 Wichita State, while 55% picked #4 Louisville, out of the Midwest Region
  • #1 Florida 68%, #2 Kansas 32% in what looks an entertaining South regional Final
  • Sorry Wildcats fans, Villanova was the only #2 seed not picked
  • No teams outside of the top #4 seeds were picked, how boring

Elite Eight:

  • 100% of brackets picked Florida in the Elite Eight
  • 90% have #4 MSU facing #3 Iowa State, which is absurd considering only 16% of #4 seeds and 26% of #3 seeds even make the Elite Eight
  • #4 Louisville and #1 Wichita State are split 60/40
  • Only 9% picked UVa, the lowest #1 seed, and 0% for #2 seed Villanova


Sweet Sixteen:

  • 23% have Kentucky beating #1 Wichita State, but only one bracket had #4 MSU or #3 Iowa State losing
  • All 20 experts picked #4 Louisville to make the Sweet 16, stunning considering half the #4 seeds do not make it there
  • #2 Kansas 55% vs. #7 New Mexico 45% looks like a great game
  • Same with #2 Villanova 55% vs. #7 UConn 40%

Conclusions

  • Since 30% of #1 seeds do not make the Elite 8, UVa or Wichita St. are the choices to falter. 
  • Since 54% of #2 seeds do not make the Elite 8 either, Villanova and Kansas have the toughest opponents.
  • Only 52% of #3 seeds make it out of the first two rounds, so experts pick against Creighton as a #3 seed to continue this trend. 
  • Since only one time out of 34 has a Final Four been all #1 seeds, experts think UVa and Wichita St. look the most vulnerable. 

After looking at the trends and expert picks I am even more inclined to pick UVa to win it all. If the entire pool picks MSU, Louisville, etc, I gain an advantage by picking UVa if they win, while not gaining anything by picking against Virginia.  The consensus of the expert picks is staggering though. Using the Ken Pomeroy rankings (the en vogue college basketball sabermetrics rankings), UVa is the #4th best team in the country while MSU is #10. Granted Michigan State is now fully healthy and coming off a BIG 10 tournament title and they may be underseeded as a #4. However, they have gone 8-7 in their past 15 games, while UVa has only one loss in a meaningful game since New Year’s.  Historic trends have #1 seeds advancing 70% of the time in this matchup, while only 16% of #4 seeds, but the 22 experts picked 9% #1 seed/91% #4 seed.

Meanwhile on the other side of the region, which team do you think got 95% of the picks to make the elite eight? The #7 ranked team or the #22, #25, or #26 ranked teams? Iowa State, the #22 ranked team, was predicted to win in 21 of the 22 brackets, while #7 Villanova got zero.

Either the experts or on to something that I do not know about, or maybe I found out a key to who I am picking this week.  That’s why they play the games.