We’ll be exploring this new technology in more detail at Think 2018. Join us.
Whether you’re a member of the Patriots Nation or an ardent Bills Backer, or even if those references have no significance, you can’t make it through a football season without hearing about fantasy football. It gives fans a personal connection to the game, the players and the actual plays.
Every week, fantasy football team managers dig through articles and predictions from sports analysts to select a starting roster. The breadth of options supported by the vast amount of available data is both intriguing and daunting. On average, fantasy football team managers can only consume 3.9 sources to base their decisions. However, for perspective, for every 60 seconds throughout 2017 there are 46,200 Instagram posts, 452,000 tweets, 900,000 Facebook logins, and 3.5 million Google searches. That leaves a lot of valuable information on the table.
There are typical scores and stats — structured data – but what about all the unstructured data, such as the millions of words printed daily in articles and blog posts? These articles contain rich details and insights but tracking them all requires more time than most can spend in a given day.
Highlights of some of the exciting work from IBM Sports and IBM iX.
The Next Era of Fantasy Football
To leverage the breadth of unstructured data to empower fans with more information, IBM and ESPN have teamed up to apply a novel artificial intelligence (AI) system with Watson to the fantasy football decision-making process. The ESPN Fantasy Insights with Watson is powered by dozens of machine learning techniques, helping uncover which player is going to boom, bust or play with an injury. Each of those factors, along with unstructured data, predictors combines to produce probability based point spreads for each player.
IBM iX designed this AI-enabled experience for millions of users. Working side-by-side with ESPN, the team designed a system that would provide impactful insights and enhance the current ESPN projections in a user-friendly format for fantasy football players. Additionally, a user experience was developed that integrated seamlessly into the current ESPN Fantasy Football ecosystem.
Teaching Watson the Playbook
To train our AI system with Watson, it read more than six million articles just to learn the game of football. Every day during the season, the system reads and comprehends millions of news stories, opinion pieces by experts and reports on player injuries from more than 3,000 sources every hour. The resulting insight then correlates with traditional, statistical data on more than 1,900 players across all 32 NFL teams to help fantasy owners decide which players to start. Now owners can compare players using conversations around the Internet as well as game statistics.
The insights generated from our AI system with Watson’s comprehension of unstructured data are a complement to traditional stats and the expertise of fantasy football players at all levels. Even the ESPN team supporting fantasy football guru Matthew Berry is using IBM Watson to distill millions of articles into insights. An IBMer recently went undefeated and won the league championship as a result of our insights.
Over the last decade, mobile technology has grown to play a pivotal role in business and in society as a whole. Starting from the early days of the mobile web, mobile technology has developed to the point that so many of us couldn’t last one day without our favorite apps. However, technology never stands still, […]
Self-checkouts are supposed to save us time, but crumpled barcodes in unpredictable packaging locations requiring individual scanning, and the dreaded ‘unexpected item in the bagging area’ warning message can quickly turn a trip to the shop into an extreme test of one’s patience. The longer and more arduous the process to check out, the more […]
With recent advances in AI, including natural language processing (NLP), voice will be the next big paradigm in how we interact with digital devices. With the big tech companies racing to establish dominance of their respective voice platform, it’s only a matter of time until spoken dialogue will become the standard for digital interfaces. Whether […]