We use cookies. You have options. Cookies help us keep the site running smoothly and inform some of our advertising, but if you’d like to make adjustments, you can visit our Cookie Notice page for more information.
We’d like to use cookies on your device. Cookies help us keep the site running smoothly and inform some of our advertising, but how we use them is entirely up to you. Accept our recommended settings or customise them to your wishes.

Merkle Congratulates Big10sBasement for Winning 2018 Bracket-lytics Challenge

The 2018 College Basketball tournament began with one narrative being pushed by many pundits: there are no dominant team and this is anyone’s year. Villanova decided to take that narrative and incinerate it, much like Donte DiVincenzo did to Michigan’s defense last night. The Wildcats put on a clinic throughout March, winning every game by more than double digits and covering the Las Vegas point spread in every tournament game - a feat replicated in the past 20 years by only the 2009 North Carolina Tar Heels.

Though the national media may have pushed the “no-dominant-team” narrative, the answer may have been clearer if an analytic approach was taken. Sixteen of the 46 student teams who submitted models for the Bracket-lytics Challenge predicted that Villanova would win the national championship before the start of the tournament.

Villanova’s offense was scorching throughout the regular season and the reason for that is shown in the graph below. Effective field goal percentage is a method of measuring field goal percentage while also considering that 3 pointers are worth 1.5 times more than 2 pointers and turnover rate is turnovers per 100 possessions. ‘Nova was king this season when it came to shooting the ball effectively while also maintaining possession efficiently. Many of the games Villanova played in the tournament consisted of 6- to 10-minute stretches where the opposing team would go cold or get flustered by the Wildcats’ constantly switching defense. But Jay Wright’s team hummed like a fine-tuned machine, producing crisp pass after crisp pass until they found an open 3-point shooter to knock down an easy shot.

Effective FG% vs Turnover Rate

The winners of the Merkle Bracket-lytcs Challenge utilized this logic in the model they entered into the competition. Big10sBasement used a model very similar to the “Four Factors of Basketball” developed by Dean Oliver for the NBA game, however, they appropriately fitted it for the college game. The basic premise of the model was: What happens in a possession of basketball? The team can either:

  1. Turn the ball over
  2. Get fouled and shoot free throws
  3. Attempt a shot, in which case
    1. They make it
    2. They miss it, in which case
      1. The opponent grabs the ball
      2. The shooting team grabs its own rebound

Quantifying the story of a basketball game and focusing on what can happen during each possession and how likely it was to happen allowed Big10sBasement to make accurate predictions for the College Basketball Tournament.

Congratulations to the Villanova Men’s Basketball team for winning the NCAA Men’s National Basketball Championship and to Big10sBasement for winning the inaugural Merkle Bracket-lytics Challenge.

Team Members include Charles Ruff and James Micheli from the University of Iowa and Jacob Sigmund and Nick Girdis from the University of Wisconsin. These winning team members will receive paid internships at a Merkle Americas location plus a team prize of $1,000.

As for the rest of us, there are some conclusions that can be found when analyzing the basketball film along with the basketball numbers. What matters when winning basketball games comes down to one thing: buckets.