College Basketball Analysis and Predictions
With SAP Lumira and SAP Predictive Analysis
After an intense 3 weeks of crunching stats, analyzing results, and running predictive algorithms on data from 64 college basketball teams, our SAP Data Viz team ended up correctly selecting 11 out of 16 sweet sixteen teams and 2 out of the final four (Florida and Wisconsin). Thanks to everyone who participated!
How’d we do?
- Final Four: 2 out of 4 teams correct
- Sweet Sixteen: 11 out of 16 teams correct
- Overall: 40 out of 63 teams correct
- Finished in top 15% of all NCAA brackets
The Analysis and Data Stories:
- March Madness Predictions with SAP Lumira by Craig Powers
- BI2014 and HANA2014 Takeaways by Holger Mueller
- Insights from SAP Insider BI 2014 – BI for the Business User by Cindy Jutras
- March Madness 2014 using SAP Lumira by Sam Ko
- My 8 year old is mad about SAP Lumira by Ryan Oneil
- UK vs UCONN – You Care by Tammy Powlas
- 11 of 16 sweet sixteen predictions with SAP Lumira and SAP Predictive Analysis
SAP Data Viz Team's Bracket
Using SAP Lumira and SAP Predictive Analysis, the data viz team at SAP analyzed data from 64 teams to fill out the 2014 bracket. Each selection was made based on data, removing any gut feel or school bias from the equation. Here’s a screenshot of the bracket:
What data was used and how was it analyzed?
The team used public data from several sports websites and looked at a few key statistics when analyzing the teams. The key statistics included: 3-point shooting percentage, 2 point shooting percentage, offensive and defensive rebounds, strength of schedule, and RPI (ratings percentage index). Each individual match-up was compared and analyzed on a team by team basis. Watch this video showing data visualization with SAP Lumira Cloud.
Here’s an example of how we analyzed the team match-ups.
Using SAP Lumira the team was able to ask questions of the data and investigate key statistic individually or compare with other statistics to get an overall picture of each match up. Next, the team applied a predictive algorithm to the data to uncover groupings (clusters) of teams who might make it to the finals. Here’s an example of a predictive model run against the data:
The example shows teams by cluster and allows analysis of the characteristics which could help them advance to the final four. Here’s another view showing a scatter matrix identifying the top clusters for 3-point shooting percentage and 2-point shooting percentage.
Watch this sports fan data geek video with SAP Lumira and SAP Predictive Analysis
The final step is the share the analysis and let the trash talking begin. Visualizations can be shared to http://cloud.saplumira.com allowing others to investigate the data and come up with their own insights. See how we analyzed the data using SAP Lumira Cloud by viewing this public interactive visualization and see what insights you can come up with.
*Disclaimer: this analysis and predictions are an attempt to showcase SAP technology not accurately predict sport outcomes.
Get involved and share your analysis and selections for your favorite teams
Here's how you can join in on the action:
1. Download SAP Lumira via www.saplumira.com
2. Download SAP Lumira College Basketball sample data
3. Publish your viz to SAP Lumira Cloud
Join the SAP Data Viz Teams pool and let the trash talk begin! Not sure how to participate? Read this tutorial for how to publically share visualizations. Follow @SAPLumira on Twitter for details and tweet-away prizes.