Evolution of Sport (EOS) is an opportunity to present a message, an idea or a revolutionary thought that could someday change the face of sport.
EOS Edicts: *Be Bold *Be Unique *Be Inventive *Be Analytical *Be Concise *Be Respectful *Be Curious *Be Humorous *Be Honest *Be Inspiring
Below is the list of the current confirmed 2014 conference EOS Talks.
It is now possible for anyone to broadcast a live video stream over the web. As a result, niche sporting events are now accessible to fans anywhere. Although recent technology has made the distribution aspects of broadcasting affordable, creating high quality content is still expensive. In this talk, we show how sporting events can be filmed automatically using player tracking technology to drive robotic pan-tilt-zoom cameras. In order to mimic the anticipatory skills of human camera operators, our proposed approach includes a virtual camera operating on a short delay. Smart venues, incorporating technology such as this, may be part of the future of sports broadcasting.
Despite the exponential growth in interest and enthusiasm for sports analytics, we have not seen a corresponding increase in the rate and quality of discovery. This is largely due to a lack of quality data being available to passionate, talented people. Vantage Sports wants to change this reality.
Vantage has combined NBA minds (headed by Ryan Blake) with world-class technical talent (ex-Google engineer) to create technology that aggregates more than 16,000 unique and relevant data points/game. It successfully mixes traditional scouting with the tenets of rigorous analytical analysis. This presentation will introduce Vantage’s data set as well as a few of the new metrics that we can derive for every phase of the game and lay the groundwork for future discovery.
The Wild, Wild West of Analytics” will introduce college football into the analytics conversation, discussing how college football differs from its pro counterpart and where we’ve come thus far in the exploration of college football stats and analysis. The sport is a strange combination: It’s huge (either the second- or third-most popular sport in the country, depending on the survey), and it’s relatively untapped from a data perspective.
Despite the randomness inherent in a sport with a pointy ball, the college football analytics community as a whole has made progress, both in terms of scope and size. This presentation will discuss the work that has been done thus far and the direction research can and should go moving forward.
Football front offices and an increasing number of fans want answers to questions such as: How much is each player worth (on a balanced scale throughout the league)? Who has the deepest rosters in the NFL, and how did they develop them? Does my quarterback contain the critical elements required to win a Super Bowl? The answers to these types of questions hold obvious strategic advantages, but to this point few known analytical techniques provide them.
In the search to providing comparative analysis, the Elitics Model was created, introducing the Player Efficiency Rating (PER) for the NFL. Similar to John Hollinger’s PER for the NBA, the goal of the Elitics PER Model is to provide one metric in reviewing all players from all positions throughout the NFL
Professional mixed martial arts (MMA) is primal, yet simultaneously highly evolved. Regional dialects of combat sports coalesced within the Ultimate Fighting Championship (UFC), popularizing new syllables in the human language of aggression like “arm bar” and “spinning back fist.” Modern fighting now requires fluency in both striking and grappling arts, forcing a rapidly evolving hybrid sport. MMA, ultimately, killed the Kung Fu star.
Utilizing decades of detailed MMA performance data, basic advantages were quantitatively tested for the first time. Does size really matter in fighting? How does the Southpaw Advantage work? How is the “Tale of the Tape” misleading us? Additional (hidden) performance factors were also tested and weighed. It turns out there’s no such thing as a “fair” fight, not if you’re using analytics.
We are experiencing an analytics revolution in soccer, made possible by the collection of player- and spatially-tagged event sequences that occur during games. Hypergeometric Enrichment Analysis
(HEA) was invented in the biomedical research realm to help researchers discover genes and patterns related to cancer biology. In partnership with the San Jose Earthquakes, we apply HEA to Major League Soccer event sequences. Through this approach, we identify patterns of players and player-interactions that significantly contribute to goals for or against. We identify strong players and weak players that might not be noticeable on a stat sheet via traditional metrics such as goals and assists. We present two new statistical metrics, Enriched in Goals For
(EGF) and Enriched in Goals Against
(EGA). We identify and discuss the contribution of MLS players with higher than expected EGF and EGA scores. Finally, we show how HEA can be used to educate coaches and players in advance of a match, highlighting an opponent’s recent strengths and weaknesses in front of goal.