Predicting the upsets for the NCAA Men’s Basketball Tournament using machine learning

Contemporary Analysis (CAN) and Cabri Group and have teamed up to use Machine Learning to predict the upsets for the NCAA Men’s Basketball Tournament. By demonstrating the power of ML through our results, we believe more people can give direction to their ML projects.
Machine Learning (ML) is a powerful technology and many companies rightly guess that they need to begin to leverage ML. Because there are so few successful ML people and projects to learn from, there is a gap between desire and direction. 
We will be publishing a selection of games in the 2017 NCAA Men’s Basketball Tournament. Our prediction tool estimates games where the lower seed has a better than average chance of winning against the higher seed. We will predict about 16 games from various rounds of the tournament. The historical baseline for lower seeds winning is 26%. Our current model predicted 16 upsets for the 2016 tournament. We were correct in 7 of them (47%), which in simulated gambling gave the simulated gambler an ROI was 10% (because of the odds). Our target for the 2017 tournament will be to get 48% right.
Remember, our analysis isn’t to support gambling, but to prove the ability of ML. However, we will be keeping score with virtual dollars. We will be “betting” on the lower seed to win. We aren’t taking into consideration the odds in our decisions, only using them to help score our results.
We will be publishing our first games on Wednesday 15th after the first four games are played. We won’t have any selections for the first four games as they are played by teams with identical seeds. Prior to each round, we will publish all games that our tool thinks have the best chance of the lower seed winning. We’ll also publish weekly re-caps with comments on how well our predictions are doing.
Understand the technique that finds a group of winners (or losers) in NCAA data can be used on any metric. Our goal is to open up people’s minds onto the possibilities of leveraging Machine Learning for their businesses. If we can predict things as seemingly complex as a basketball tournament (Something that has never been correctly predicted), then imagine what we could do with your data that drives your decisions?
If you have questions on this type of analysis or machine learning in general, please don’t hesitate to contact Gordon Summers of Cabri Group ( or Nate Watson at CAN (
Those interested in the detailed description of our analysis methodology can read the technical version of the article found here.
**Disclaimer: Any handicapping sports odds information contained herein is for entertainment purposes only. Neither CAN nor Cabri Group condone using this information to contravene any law or statute; it’s up to you to determine whether gambling is legal in your jurisdiction. This information is not associated with nor is it endorsed by any professional or collegiate league, association or team. Machine Learning can be done by anyone, but is done best with professional guidance.

Rethinking Business Intelligence Software

People don’t care about business intelligence software, they care about what it can do for them.  CAN is built on this idea.  Instead of focusing on business intelligence software, we are focused on providing answers directly to our clients.  We are improving this process by launching the CAN Portal.  The Portal is how we work with our clients.  It will allow you to get better answers faster and more securely.
What are your objectives?

How CAN Takes a Different Approach

At Contemporary Analysis (CAN), we take a completely new approach to helping companies and organizations get more out of the information they have access to. At our core is the idea that businesses should be working smart and hard.  At CAN, we are different because we always keep the human element, actionable impact, and added value at the forefront of our development process.

We start the entire process with keeping the human element in mind. Everyone has gone through the frustrating process of being passed off from one person (if you’re lucky enough to reach a real human) to the next throughout a customer service or sales process. In most of these situations, half the time speaking with a new person is catching him or her up on things you have already said to other members of their organization.
At CAN, we understand the importance of having one contact throughout the entire process. This contact, known as a Navigator, takes the time to understand your specific business and helps you distill problems with big impact solutions. Navigators understand the majority of managers and executives don’t have the time to learn about predictive analytics. Navigators take the time and effort to understand the problem or issue from the end user’s point of view and then strategize to reverse engineer an efficient solution. It is our job to couple your expert knowledge and historical data to give you a solution with impact.
At CAN, we understand businesses outsource services for added value. The value of using predictive analytics is only as great as the actions and changes made with the information provided. You could have a GPS system in your car, but if you never turn it on, it doesn’t do you any good. From step one in our customer process, we work on finding helpful insights into areas in which you can TAKE ACTION or MAKE CHANGE, not just look at the report and think “Hmm, that’s interesting.” If the information we provide doesn’t induce change on at least some level, we didn’t properly do our job.
At CAN, the solutions we provide make sense financially. We use predictive analytics to answer questions in about 30 days. Think about that for a second. In just 30 days you could have an analytical model which, while not being perfect, will allow you to make much more informed decisions. Whether it’s having a better understanding of up-sell, cross-sell, or customer loyalty. It is important to remember, the goal of predictive analytics is to be LESS WRONG, and models continually become less wrong by using current information to test and re-test.
Compare 30 days with CAN to the alternatives – doing nothing or creating an in-house predictive analytics department. In smaller companies, the alternative to CAN is to do nothing. Smaller companies don’t have the resources to create in-house predictive analytics, but have a lot of the same issues as large companies.
The other alternative is in-house analytics. I believe in-house predictive analytics departments are something every large company should invest in. Properly managed and financed in-house departments can change organizations in ways never thought possible, in ways which only the future will show us. No longer would justifications for decisions be based on gut instinct, or worse yet, “because that’s how we’ve always done it.” This business is your passion, and nothing proves a point faster than quantitative justification. However, what if you’re not a large company?
In-house departments require management, direction, and resources. A company looking to develop the smallest possible in-house predictive analytics department will pay for the following:

  • Find and hire a highly sought after Ph.D. or Masters at a cost of 75-150k.
    • Add half the salary again for employee benefits, taxes (FIFA, FICA, etc.), office space, physical equipment, and HR resources.
  • Purchase a Tableau, SAS, or SPSS license costing 10’s of thousands of dollars for just one year

This cost of 175-250k is just the initial investment. You haven’t solved one problem yet. Taking into account the months of required training and corporate acclimation before any useful insights can be made by this new hire, it can be 12 to 18 months before you have a solution to just one of your smallest problems. 30 days and a cost of at least a zero less sounds much better to me.
Like I said before, I would encourage all large companies to take the plunge into predictive analytics. Even with well developed predictive analytics solutions there is a high possibility the department would suffer the same downfall as some IT departments, the disconnect between who designs or provides the technical knowledge and who actually uses it on a day to day basis.
At CAN, we take your expert knowledge and use your data to provide valuable insights you never thought possible. Throughout the entire process, we never lose track of the importance of the human element. Whether you’re a large company with an in-house analytics department or smaller business with no means for self analysis, we care about your business and can give you value in the form of information from which actions can be taken.

Missiles, the NBA and Predictive Analytics

Ask any Cubs fan. Sometimes life just isn’t fair. Conventional wisdom in sports says that the teams with money to buy talent win more games. For example, The Yankees, they will always have the best talent money can buy. In basketball it’s no different. The teams stacked with expensive talent always seem to have the advantage. Well, they use to anyway.
A few innovative NBA teams asked themselves a question: How can we win more games against teams that have more money? The answer: predictive analytics and missile tracking technology. Yep, you heard me right, math and missiles. (more…)

New Pew Survey on Big Data | Big Data is the new Oil | 'Minority Report' software hits the real world

Why data trumps experience in trial conversion: “Using predictive analytics to qualify trial users and focus on those that are most likely to convert can double conversion rates. In a 2012 study, the Aberdeen Group published a finding that companies using predictive analytics have a 73% sales lift versus companies that did not. … Publishers should use predictive analytics to develop trial scoring rules. These scoring rules can constantly prioritize trials in their likeliness to convert which increases close rates and sales productivity. These same predictive analytics are useful in design of trial parameters such as length and access limits.” (Business Insider)
Connect Big Data With Customer Behavior to Improve Social, Email, and Web ROI: “Since we have lots of data, we have lots of integration challenges. … Mastering that flow of data between the places that generate it (click-stream, communities, sentiment analysis, email and SMS messaging, and portals) and the systems that utilize it (marketing automation, messaging delivery, and social publishing) is creating complexity, as well as opportunity.” (more…)

Analytics Market Grows in 2012

Analytics Market Grows in 2012

“The global market for business analytics software grew roughly 14 percent in 2011, fueled by pervasive hype about ‘big data’ as well as new technological innovations, according to a report unveiled by analyst firm IDC yesterday. Between now and 2016, the business analytics market will have a compound annual growth rate of 9.8 percent, reaching US$50.7 billion, IDC said.” (Global Financial Network)
— “As part of that overall business analytics segment, the data warehousing platform software portion represented the fastest growth, at 15.2 percent in 2011 compared with 2010. IDC also pegged analytic application growth at 13.3 percent last year from 2010, and BI and analytic tools at 13.2 percent last year from 2010.” (Information Management) (more…)

Featured Posts – Click the Brain
CAN Jewels