MARCH MADNESS | NCAA BRACKET | ANALYTICS

VIEW HIGH FIVE  

How BigData is the Key to Winning Your Office NCAA Bracket Pool

March Madness is a tournament steeped in numbers. Sixty-four teams compete. Americans fill out 70 million NCAA brackets each year. There are 67 games played. And just one team wins it all. Yet, what makes the Big Dance so exciting is its unpredictability. Only once have all four No. 1 seeds gone on to the Final Four, and every year, a “Cinderella team” seems to make an unexpected run to the National Championship. The tournament is so unpredictable that Warren Buffett once put up $1 billion for anyone who could accurately predict the outcome of all 67 games. Seems crazy, but consider this: People have a better chance of winning the lottery, than they do submitting a perfect bracket. The odds are something like 1 and 1.6 billion. Yet, Big Data is disrupting the NCAA tournament. Now, data scientists are applying advanced algorithms and statistical analysis to determine what teams are best positioned to win it all. And the results of these data science projects are promising. The Dance Card, for example, a formula that predicts which teams will win at-large tourney bids has been near-perfect in the last four years. Another, a simple algorithm using three data sources provided a reliable probability calculator for each game’s winner. Although Big Data hasn’t found the secret to the perfect bracket; the point is clear: Data can help weaken and eliminate unpredictability. And whether you’re using it to fill out your NCAA bracket or to predict which product will sell the most during the holidays, your decisions are more well-informed when they’re backed up by right data.

The Math of March Madness via The New York Times

The Math of March Madness via The New York Times

In 2014, nearly 200 data scientists competed to see who could most accurately predict NCAA tournament winners. The winning project, though, was much simpler than the rest. It used just three data sources – the Las Vegas spreads, and measures of each team’s offensive and defensive efficacy – to determine the probability of each team winning a game. Simply put, accurate models aren’t necessarily those that use the most data – they’re the ones that use the right data.

Read More
Inside Sports Analytics: 10 Lessons for Business Leaders via SAS Analytics Blog

Inside Sports Analytics: 10 Lessons for Business Leaders via SAS Analytics Blog

Sports analytics is a burgeoning field, but analytics are being used much differently in sports than in business. Here’s one example: Sports teams use data to determine the best, highest performing athletes to draft and trade for. But in business, data is rarely used to look inwards to the organization. Turning the data onto individual performance can help businesses quickly determine who their most valuable players are.

Read More
The Real March Madness: Using Statistics to Fill Out Your NCAA Bracket via The Denver Post

The Real March Madness: Using Statistics to Fill Out Your NCAA Bracket via The Denver Post

What are the best statistics to use to accurately predict your bracket? There’s no right answer. But what you see is that data is a powerful tool for addressing the unpredictability of college basketball. One scientist quoted created the Basketball Power Index (BPI), and in 2012 and 2013, the team with the top BPI entering the tournament went on to win it.

Read More
Duke's Analytics Advantage Keeps Growing via ESPN

Duke's Analytics Advantage Keeps Growing via ESPN

Almost every NBA team has installed military-grade cameras in their arenas to collect tiny pieces of data. For example, the number of passes, where players are on the court, and shot selection are all captured using these cameras developed by SportsVU. Duke University rolled out the program in 2013 – one of the first colleges in the U.S. – before going on to win the championship in 2015. Coincidence? Probably not.

Read More
Your NCAA Bracket: A Classic Big Data Problem via Data Informed

Your NCAA Bracket: A Classic Big Data Problem via Data Informed

In any situation – whether business or sports – data scientists face challenges. The NCAA tournament presents its own unique set of issues: It’s a tiny sample size, there’s a lot of noise, and injuries and illnesses can create incomplete information. But the key is preparation. If you know what challenges might be there, you can adjust your models on-the-fly to improve learning. This article shows an important lesson: You must prepare for the unexpected in your Big Data experiments.

Read More

TEAM CLUTCH PROMISE

Understanding data and analytics is TEAMCLUTH’s specialty & Our expertise is in making Business Intelligence/BigData technology simpler, stress free & actionable for you to make things happen in your business.

Subscribe to our High Five to receive curated story on Data & Analytics.





Subscribe