16311

I was at a conference last week and one of the organizers asked the crowd of about 100 CIOs if they are working on a Big Data project. Only about 10 percent raised their hands. Why is the adoption rate so low?

One major barrier in the adoption of Big Data is understanding how the Big Data concepts can be applied to each corporation. As I thought about the changes made to the ways we can make decisions driven by the flood of  data, I thought of one of my favorite hobbies - fantasy football.

I fell victim to the fantasy football bug about 15 years ago and formed a league with several of my co-workers. The Big Data of fantasy football is the vast volumes of structured statistics and the unstructured layers of news, analyses, videos, radio shows, Tweets and Facebook posts. It’s the judgment applied to assess the quality, dare I call it veracity, of the provider, author or analyst.  But it’s also about the velocity and timeliness of the information.

In the early days of fantasy sports, information about teams and players was available only by watching or listening to the games and reading the newspaper, requiring a lot of skill or dumb luck to pick the right players in the best situations.  But, that’s all changed in the era of the Web and Big Data.

To play the game, you draft players from all the professional football teams to create a winning team. Upfront knowledge of how important each player is to his individual team is important. How many touchdowns will he likely score? How many yards will he gain? How many field goals will he kick and passes will he catch? But picking a team before the start of the season is only the beginning of this data-driven decision-making.

Each week of the football season, you are faced with several decisions regarding which players you want to “start” and which to “sit” (see, you draft more players than you can play each week). In addition, player injuries are a factor, overlooked rookies emerge as significant, and many other dynamics in the real world of football impact your ability to score points and win fantasy games in weekly match-ups against other teams in your league. New information sparks new questions that require more decisions.

For example, consider Willis McGahee, the running back for the Denver Broncos. He injured his knee a few weeks ago and I was left with a decision to make. Do I replace him as one of my weekly go-to guys with someone already on my bench or do I try to pick up another running back in the player pool not on anyone else’s team? I am faced with literally hundreds of sources of data, stats, video replays and analysis to use to help make the best and most informed decision:

• Traditional media reports on fantasy football, in print, TV or online, USA Today, CBS, NBC and ESPN (all have dedicated columns and reports and even an occasional column in the Wall Street Journal)
• Free or subscription content from any of the dedicated fantasy football news and analysis sites (remember, this is a $1B industry)
• Reports and stats reported by over 300 fantasy football apps in the iTunes store alone
• College and pro game logs for each of the players and game logs for teams the Broncos will face over the next few weeks
• Condensed replays of the last few Broncos games through a special satellite TV or online subscription
• Twitter feeds of other Bronco players, coaches, equipment managers and front office staff on Twitter or any of the other hundreds of NFL players who tweet for some insights, not to mention all of the analysts, insiders and reporters who are active Twitter users…

After scanning the aforementioned big data available, I decided that either Lance Ball or Ronnie Hillman would be possible fill-ins for McGahee.  Some thought Ball would get the nod as a veteran pass blocker in Peyton’s offense while others suggested that Hillman, the rookie, was more explosive and athletic. Surprisingly, we were all wrong.

Ninety minutes before the game against the Chiefs, Broncos coaches named Knowshon Moreno as the replacement for McGahee. He didn’t even suit up for the Broncos for 10 games this year! In the early fantasy days, this fact would be lost on fantasy players until the game started. Today, it was reported online, tweeted and retweeted, emailed and popped as mobile app alerts across the fantasyverse.

Fantasy football is just one example of the impacts that Big Data can have a huge impact on decision-making, including how sometimes all the data in the world won’t serve as an effective predictor. Hopefully this story stimulated your own thinking about how you can get started with some Big Data projects of your own.

Image shared by Parker Knight

16311