Sports Analytics Innovation Summit 2014 San Francisco: Day 2 Highlights
Highlights from the presentations by Analytics leaders from San Francisco Giants, New York University and LA Dodgers on day 2 of Sports Analytics Innovation Summit 2014 in San Francisco.
The Sports Analytics Innovation Summit (September 10 & 11, 2014) was organized by the Innovation Enterprise at San Francisco. From the bleeding edge of performance enhancement to fan engagement & business analysis, a wide spectrum of topics were discussed. Experts shared their thoughts on how analytics are constantly evolving the way the sports industry operates, in every sense.
We provide here a summary of selected talks along with the key takeaways.
Highlights from Day 1.
Here are highlights from Day 2 (Thursday, September 11, 2014):
Russ Stanley, Managing VP, Ticket Sales & Services, San Francisco Giants gave an exciting talk on "Using Analytics to Define the Ticketing Experience". Data and analytics have become the new buzz word in ticketing. Dynamic pricing took all the data and produced information that allowed us to properly price our tickets. Now our analytical team is providing us with recommendations to better run our business. We rely on this information for promo dates, special events, season ticket, risk assessment and keeping the sell-out streak alive.
Based on his over 25 years of experience with San Francisco Giants, he remarked that it has been astonishing to see the change from a world where decisions were predominantly made by gut instinct to today's Analytics era - where there is number crunching behind even the small decisions. Today, San Francisco Giants' analytics architecture comprises of a data warehouse, Salesforce customer management tools and Tableau reporting.
SF Giants has been doing dynamic pricing for about 6-7 years. Currently, they use Qcue as the front-end management tool for pricing decisions. There has been substantial benefits of modifying price dynamically based on real-time monitoring of demand and several factors that impact demand. Sometimes, the urge to continue the sell-out streak obstructs revenue maximization through dynamic pricing. Sell-out streak has some reputation factor, and thus, one needs to be careful about trading it off for greater revenue.
All games are not created equal. Real-time monitoring of demand can help determine the effective ways to add value to soft games. The increased price from big games allows the firm to have softer prices on smaller games. Monitoring and measuring the effectiveness of promotions and special events can deliver unprecedented insights to make the most of marketing dollars. Analytics plays a key role in improving fan engagement, which is a top strategic priority, to maximize season ticket retention.
Philip Maymin, Professor, New York University delivered a highly informative talk on "NBA Optical Analytics in the Real World". He shared his half a decade of hands-on experience with the SportVu data to present practical insights on how best to extract value. The four main cross-dependent components are: asking the right questions, computing the answer, presenting the results, and automating the process. He also talked about the tools, programming languages, and skills are needed to store and compute the data. It's important that projects be simultaneously quickly prototyped and yet managed to build an extensible system with real value for the organization rather than a collection of one-off results.
Talking about the current scenario, he described the hierarchy of optical adoption. The zero level comprises of the standard public reports which includes data on speed, distance, touches, passes, drives, catch-and-shoot and pull-up. The first level comprises of the standard team-only reporting. This is very basic level of reporting, and mostly, the teams are stuck here since it is hard to get started on deploying analytics effectively. The second level comprises of cases where custom one-off reports are being generated through external contracts. The third level, the ideal state, comprises of custom in-house tools and reports, nightly runs, entire data and analytics system incorporate all traditional analytics as well, live visual interactive tools, cloud-deployed, extensible and production-level system.
Talking about the size of data, he mentioned that SportVu data per game is 40 MB and per season it is 50 GB. These numbers might seem drastically low compared to the data processing at Walmart or Facebook. So, the data at hand is not "big data". A better approach is to treat SportVu data as binary data, and not as relational data.
There are a lot of questions that cannot be answered satisfactorily without optical analytics. Number crunching can provide deep insights to answer the questions on player evaluation, coaching strategy, shot selection, team chemistry, etc.
Stan Conte, VP, Medical Services & Head Athletic Trainer, LA Dodgers shared his insights on baseball injuries in his talk "Professional Baseball Injury Analytics". Baseball has a rich history of statistics and with the advent of Sabermetrics, the world of baseball performance analytics has exploded. However, research and analytics of injuries falls far behind even compared to other professional sports. He discussed the epidemiology of baseball injuries, the trends, patterns and costs of injuries in baseball. Using the example of specific injury trends such as Tommy John Surgery, he showed where injury analytics presently stand and what can be done in the future.
Injury analytics plays a very important role in structuring prevention programs, monitoring outcomes and assessing risk. In 1996, it took him 2.5 years to get the data on past injuries. The only reliable source from information was the disabled list, which is a roster management tool, and is used to replace injured players. Even today, we lack clean data on injuries in baseball.
His paper published in American Orthopedic Society of Sports Medicine in 2001, revealed that the injuries had increased sharply from 1989 to 1999. The MLB Injury Surveillance System was established in 2010 by MLB. It is an event based system that does data collection and reporting. Over 180 people enter data every day, which is a good news (greater quantity of data) as well as bad news (poor quality and reliability of data). This has enabled a lot of data analysis such as severity of injury vs days lost, pitcher injury rates, average injury rate by position, etc. Today, this system enables rigorous medical risk assessment based on factors such as medical history, demographics, biometrics, performance metrics, and usage.
|Top Stories Past 30 Days|