What Are Analytics Anyway?

The Lions trail the 49ers by three in the NFC Championship game with a little over seven and a half minutes left in the game. They’ve blown a historic seventeen-point lead to get to where they’re at now and the momentum has swung in San Francisco’s favor. They need to get points on the board; they need to stifle the season-threatening bleeding. It’s fourth and three at the San Francisco 30 yard line. While this is no chip shot for Detroit kicker Michael Badgley, he’s made 77.1% of kicks from this distance, 85.2% if we look at only the past three years.

As we know and some painfully wince at, Dan Campbell and the Lions decided to go for it on fourth and three. A quick pass from Jared Goff to Amon-Ra St. Brown resulted as an incompletion and turnover on downs. They failed to convert and it likely cost them the game.

But analytics agrees with the decision.

What the hell is this analytics that’s infiltrating our game like a popular female pop-star (see Taylor Swift story here)? Why should we leave football decisions up to a gaggle of pimple-assed nerds that have never played a down of football? Well, before we ask those questions and grab our pitchforks, let's dive into what analytics even means.

Basically, analytics is the process of discovering, interpreting, and communicating significant patterns in data. More basically, analytics is looking at a bunch of random shit (data), then trying to make sense of said random shit. We use analytics and analytical thinking every day but don’t realize it. Deciding where to order pizza from before the game and ultimately chose the place that didn’t give you explosive diarrhea the last two times you ordered? Analytics. Deciding which one o'clock game to watch and decided on the competitive Dolphins-Cowboys game over the suck-bowl Patriots-Broncos game? Analytics. 

Let's jump back to that hellacious fourth down decision by Motor City Dan Campbell. Referencing the ESPN Analytics Fourth Down Decision Analysis Tool (espnanalytics.com/decision), we can see that going for it gives the Lions a .3% higher chance of winning the game (pushes glasses up from nose). But how do these neck beards get to this percentage? 

The recommended decision is determined using a model (the organization of various data) that considers various factors. The model incorporates the score, time remaining in the game, distance to the first down, yard line, number of timeouts, pregame win probability, and the relative strength of the offense and defense on the field. They take into account the win probability expected given a fourth-down success and fourth-down failure and weigh it. However, it should be noted that this model is more aggressive than Dan Campbell after consuming ten Red Bulls. It will tell you to go for it on fourth and one anywhere on the field. Literally. This is why it is to be used as a tool and not the final gospel in determining game time decisions. 

Analytics has been around the game of football ever since we began keeping score. We could argue that analytics is actually how we keep score. Team A crossed the goal line this many times as opposed to Team B crossing this many times which means Team A wins the game. That’s analytics, my football-loving friend! What has changed, however, is the complexity of the analysis being performed. We have more data than ever at our fingertips and the compute-power needed is accessible to essentially anyone with a laptop. 

Every NFL team uses analytics in one form or another and has for years upon years. Sure, more teams may rely more heavily upon it, but it’s there among all 32 teams. According to a survey, every NFL team now has at least two analytics staffers, and 21 franchises have at least three full-time staffers dedicated to analytics. While the NFL has been slower to embrace analytics compared to other sports, such as baseball (more on this later), many teams now use analytics in various aspects of the game. Use cases include player evaluation, roster construction, in-game decision-making, training, and injury prevention. It’s there and it’s nothing new.

One thing that is new is how we view analytics and how the media hypes it up. I’ll take a moment to echo the points I made in a previous article discussing stats surrounding the DPOY award (https://shorturl.at/foqI5). I’ve talked about how we all love stats and feeling smart by using advanced analytics to get our points across. Just tune into a Thursday Night Football game on Amazon Prime for Christ's sakes. There are squiggly lines, numbers, and a litany of other bullshit spewed across your screen, with some asshole popping into frame getting a big ole nerd boner over analytics and predictions. I get it. I’m one of those nerd assholes who are data and analytics infatuated. I’ve even held a title of Director of Data Analytics in my previous job. But are we going to let the revenge of the nerds happen? Are we going to let unathletic dweebs who never played a snap of football (I was a special team "specialist" in high school) tell us how, where, and when the game should be played?

A company called Pro Football Focus, or PFF, has come under scrutiny as of late. Owned by our boy, Chris Collinsworth, PFF provides in-depth analysis for the NFL and other sports. It provides player grades, rankings, and other expansive stats. They’re challenging the traditional metrics that we’re accustomed to, such as sacks, completions, and rushing yards and instead offering advanced statistics such as pass-rush win rate, adjusted completion percentage, and breakaway percentage. It tracks outcomes of obscure situations like a team’s overall conversion rate when facing a second and eleven while being the away team. Some folks scoff at these insights and dismiss it as a bunch of useless garbage. Why? Because people don’t like when you piss around with their traditional ways of football and how it’s judged.

Can I say something without everyone getting mad? I truly enjoy Pro Football Focus and the advanced analytics available to us. Having as much information as possible at the fingertips of teams, coaches, and fans is a good thing and makes the game of football better and more enjoyable. I believe that having these insights available to coaches makes the game more competitive. More and more coaches are embracing analytics in their strategy, most notably the younger fleet of coaches like Mike McDaniel in Miami, and he has an offense that’s damn fun to watch. But that suave, goofy little son of a bitch said it best. "It’s a tool, like anything else, but you use it with every other piece of information to make the best decision," said McDaniel. The guy has it right, it should be considered, not the end all be all as a decision maker.

Analytics are present in more sports than just football. There’s a little film that came out in 2011 (and an excellent book before that) called Moneyball. The story made us all salivate and romanticize analytics in sport. It had us all rooting for that dreamy anti-hero Brad Pitt to shake up the unfair game of professional baseball using new, unpopular methods.

If you haven’t read the book or watched the film, it basically goes like this: The GM of the Oakland A's faces the challenge of assembling a competitive baseball team with a limited budget. To overcome this, he adopts a radical approach to player evaluation, relying on statistical analysis and computer-generated data rather than traditional scouting methods. He builds a lineup of overlooked misfit toys and challenges the highest payrolls in baseball. Wow, my calculator is bursting right out of my pocket just thinking about it.

But there’s much that happened before these theatrics of the 2002 baseball season. In the 1980s, there was a growing community of baseball nuts that decided the traditional metrics and statistics in baseball were misleading. An error is subjective and a hit and run is a bunch of bullshit that means nothing, they claimed. So they looked to collect and analyze their own statistics. Does this sound familiar?

What ultimately came of this were stats such as OPS (on-base slugging percentage) that measures a ball player’s offensive performance by adding a player's on-base percentage (OBP) and slugging percentage (SLG) together. This provides an overall measure of how effective a hitter has been at reaching base and hitting for power. But the geekery didn’t stop there. They created OPS plus. A related statistic that normalizes the number across the entire league and adjusts for external factors like ballparks, with a score of 100 being league average and 150 being 50 percent better than the league average. Sports analytics are in constant evolution and as inevitable as Mahomes making the playoffs or the Pirates shitting out young talent to the Yankees.

Often thought mythical barometers of measurement such as momentum are even calculable. Yes, momentum is a real, measurable thing. Brett Yarris, Pro-Athlete Behavior Coach and Founder of Football Behavior, explains it perfectly by saying, “Momentum is a measure of behavior’s resistance to changes in the environment. It occurs when a series of reinforcers or punishers are presented without disruption.” Brett adds, “The more the behaviors compound together without disruption, the more resistant to change they become, good or bad.”=

The competitive edge comes not by merely acknowledging that momentum is there, but by measuring and acting upon it. This is where Brett and Football behavior come in, with Brett saying, “we’ve been mathematically measuring it for over a half a century. Typically, in longer term waves (I do so “week to week” and it plays a major role in my predictive modeling), but I’ve also done it play to play and rep to rep when I’m training players.” It’s relatively newfound realizations such as this that are evolving as time and technology progresses forward. As the world changes around it, so do the sports we dearly admire.

Sports are often subjective. That team is bad, that quarterback sucks, that fan base is full of assholes. Statistics and analytics look to take the subjectiveness out of the conclusion findinging. They offer objective, tangible insights to make reliable data-informed decisions. It offers up overall records, completion percentages, and the number of assaults surrounding a fan-bases stadium. It’s a record of the things the human brain can not capture, store, and analyze on its own. It is a tool, not so unlike a JUGS machine for a wide receiver. Although it may remove the human, traditional aspect from athletics, it ultimately improves both the sport and the competitors.

People accept change about as well as Brett Favre accepts Taylor Swift. They hate that shit with a burning, unfounded passion. We like things we’re used to, we’re romantic about the traditions and the way it’s always been. But if there’s one thing that’s inevitable in life, its change. Does this mean we abandon all past knowledge and methods? Absolutely not. I believe it would serve us well to approach these new, presently viewed as uncommon, methods with an open mind.

You know, goal posts used to be at the front of the end zone. The two-point conversion was added in 1994. I’m thankful for the evolution of the game and am excited to see where it goes next.

Dan, a bona fide sports and data geek, hails from the wilds of Western Pennsylvania with an undying passion for the Stillers, Pens, and Buccos. Dan has embarked on an exciting sports writing journey, ready to subject the world to his unique blend of enthusiasm, questionable insights, and yinzer homerism. Find him and his laptop in the corner of a Starbucks near you.