How Big Data Could Change Football Forever

From the sabermetrics used in baseball to the NBA's motion-tracking cameras, innovation abounds in the world of sports. Football is no stranger to these developments, and there's one little-known company that might just change the game forever.

I'm talking about Competitive Sports Analysis, or CSA for short. I was given an exclusive look at the Atlanta-based company, which uses "Big Data" to analyze sports for everyone from head coaches to fantasy players. 

A tool for any fantasy owner

CSA's flagship products are its scoutPRO Fantasy Football and Fantasy Baseball editions, both of which are part of Gannett's  USA Today Sports Media Group. Each allows users to analyze fantasy sports from an objective and subjective standpoint, combining commonly used stats with expert opinions. I was given access to a demo version of scoutPRO Fantasy Football for Weeks 16 and 17 of this year's NFL season.

The tools work as add-ons to any pre-existing fantasy service. For example, I used it in conjunction with a league on Disney's  ESPN platform. It's also compatible with CBS , Yahoo , and most other mainstream fantasy providers.

As the company is quick to point out, its fantasy football software has better predictive capabilities than many of its competitors in this space. Its rankings are accurate more than 60% of the time, which CSA says is higher than Time Warner's  Sports Illustrated, FOX Sports by 21st Century Fox (NASDAQ: FOXA) and NFL.com itself.

A couple examples: In Week 16, I used scoutPRO's weekly rankings to determine which two running backs to start out of Eddie Lacy, Knowshon Moreno, and Dennis Johnson. I was tempted to sit Moreno because he was coming off a measly four-point outing the week before. CSA's algorithms, though, projected that Lacy and Moreno would score 15 points each, more than Johnson's 11-point estimate. I wound up starting Lacy and Moreno at my RB1 and RB2 spots they combined for 29 points, just one off scoutPRO's total estimate.

In Week 17, the software came through for me again. Its rankings listed Kansas City's Knile Davis -- the backup to Jamaal Charles -- as the 17th highest running back, better than stalwarts like Alfred Morris, Frank Gore and Zac Stacy. ESPN, meanwhile, had Davis at 24th, while Yahoo had him as low as 30th overall among RBs.

Thankfully, I picked him up and started him at my flex spot. On Sunday, Davis rushed for 80 yards and two touchdowns as the Chiefs rested Charles, making my pickup the third best RB, and the highest scoring player on my team that week. 

I ended up winning my league championship partially because of these moves, in addition to dominant efforts from players I had been behind the entire season. I don't give scoutPRO all of the credit for my victory, but it did help me make most of the tough decisions correctly.

Twenty years in the making

But that's not even what excites me. The company is also developing a version of this software for football coaches, called scoutPRO Coaching Edition, which can be tailored for NFL or college teams. CSA's founder, Diane Bloodworth, shared the story behind its long road from the classroom to the football field.

The original scoutPRO Coaching Edition was designed as a game planning system for college football coaches. I had the idea more than 20 years ago...in a class on entrepreneurship [at the University of Miami] with the star running back and developed a business plan to predict the opponent's strategy and develop a winning game plan.

As Bloodworth told me, however, the first prototype of this software "was too early and the world, much less the coaching industry, wasn't ready for next-generation analytics." Instead, she focused on developing a fantasy football product until, in her words, coaches were "ready." 

 

Image via John Martinez Pavliga, Flickr. 

This is where scoutPRO Coaching comes in. In addition to giving college football coaches more statistical tools to analyze high school players, CSA's software aims to track off-field metrics too. The company says it's currently in talks with a testing company to measure what it calls "default skills" like intelligence and football aptitude. In many cases, colleges are forced to rely on ACT and SAT scores to gauge football IQ, a proxy that isn't always accurate. At least the NFL has the famed Wonderlic Test, which, despite its imperfections, is better than nothing.

There's more

Another area where scoutPRO Coaching can revolutionize college recruiting is in the stats themselves. While the majority of NFL recruits come from the same level of play (Division I-A), and thus, can be easily compared, the same can not be said for high schoolers.

Depending on the state, high school players compete against a wide range of talent. In Florida, for example, there are eight levels of play, from 1A to 8A. In Georgia there are six classes. CSA told me they're having internal discussions on how to normalize this data, and Jon Deming, a product marketing manager with the company, shared his thoughts on the subject in a recent blog post:

If a [running back] playing for an "A" team was to rush for 2,000 yards, does that automatically make him better than an "AAAA" RB who rushed for 1,000 yards?  If you were only interested in taking the raw data as is, you would say the RB who rushed for 2,000 yards is better.  But when you factor in the increased level of competition that the "AAAA" RB faces every week, the 1,000-yard difference may not be as much as you think.

Looking toward the future

Because scoutPRO Coaching allows coaches to customize how players are analyzed, it's useful for every type of program, from a spread offense to one that runs the wishbone. According to Bloodworth (emphasis mine), "The longer-term goal is to evaluate...existing players." In other words, it's quite possible the software can eventually have a major effect on front office decisions like contract extensions, free agent signings, and more. 

Mashable has compared scoutPRO Coaching to the MLB's "Moneyball" phenomenon, and even that may not give Competitive Sports Analysis enough credit. Its algorithms have helped fantasy owners win their league championships, yes, but they can also change how coaches view football from the ground up. As Jon Deming has said in the past, "Too much money and time goes into the recruiting process for coaches to be making educated guesses on players," and I couldn't agree more.