Statistics and Their Relevance to Football

This is going to be a long one.

Sometime around 2005, I was surfing Amazon.com for some football books, and came across something interesting. It was called the "2005 Football Prospectus" written by a group of guys calling themselves the Football Outsiders. I bought the book, read, and thoroughly enjoyed it. They had put together a group of people to study film and record statistics far beyond what the NFL normally records, and came out with two interesting statistical numbers, DVOA and DYAR.

Now, I'm not going to get into what these are, as you can go to their website, www.footballoutsiders.com and read to your hearts content. What I do want to get into, is the effect of these statistics on football, and what statistics really mean to football.

When I saw these statistics, and read this book, I thought it was incredibly profound. It was something new, fresh, and interesting in a world weighed down by tired sports cliches and former-player analysts who still cannot seem to grasp anything about the game beyond the position they played. I thought it was going to change the way we looked at football.

And it did.

Slowly, as FO becomes more popular and more people discover them, you're starting to see people who use these statistics because they are so new and fresh as a way to "fight the man" of the mainstream, and try to break out their own originality. All of a sudden, some of these people think that DVOA and DYAR are the greatest things since sliced bread, and is the be-all-end-all of what defines a good player.

Now, I want to take a second here and make it very clear that GatorPhan and PhinPhan are NOT these kinds of people. Their posts may use statistics heavily, but they are well informed and think beyond the numbers. I do really enjoy reading their posts.

However, it is still a growing fad, and there are plenty of people who are trying to ride in on the coattails of it like they are some kind of football expert because they know how to run a calculator.

So let me get one thing straight with everyone right now. Statistics are not the answer. People have fought and tried for YEARS to make statistics make sense of sports, and to some extent it actually works, especially in baseball. However, in baseball on any given play, there is absolutely no more then 10 players that can POSSIBLY have an effect on the play, and in reality it is very rarely more then 4 (pitcher, batter, fielder, baseman). The pitcher throws the ball either strike or ball, the batter either hits or misses, and if he hits it then someone catches it or doesn't, then makes a throw or misses it. This obviously isn't the full list, but the point is that there is at the very least a finite amount. Because there is such a small number of variables in any given play, statistics works.

That is because statistics is very good with a small number of variables. Coin flips, number of colored M&Ms in a bag, things like that, statistics are very good at. However, the more variables you add into any certain condition, the less accurate statistics are. Anyone who knows anything about statistics will at least agree with this basic fact.

Now consider the average football play. There are 22 players on the field at any time, and any ONE of these players can change the entire dynamic of the play. Each lineman can hold their block or not, and even that can be a grey area depending on if they give the QB 3 or 4 seconds. The QB has options to throw to, each of those options can make or miss the catch, then if they make the catch, they can get more yards. The defensive players either make the right read or they don't, or the safety can fake a move to the inside or he can't. Essentially, you get the point here, there is an INFINITE amount of possible variables on any given football play. There is absolutely no possible way to account for all of those variables.

We also have to look at how each player is being used, the offensive or defensive schemes. For instance, is Larry Fitzgerald necessarily a better WR then Hines Ward? Would Hines be as useful to the Cardinals as Larry is? Would Larry be as useful to the Steelers as Hines is? The answer in both situations is a resounding NO. Hines is a wonderful all-around receiver, in that he makes the high-pressure catches, is a great blocker, and will fight his ass off for every yard. Larry is a great WR as well, don't get me wrong, and there are a lot of things that he does better then Hines does, however, he's not as well rounded. Besides, the point is not whether one is better then the other, rather the point is who does better in the situation they are expected to perform in. By that measure, both receivers are spectacular.

There are a TON of things that happen on any particular play in football that are not recorded at all, yet can have a monstrous effect on the way the play develops. For instance, you guys should all remember Jake Scott, who is in my opinion one of the best safeties to ever play the game, and is certainly one of, if not the, best Dolphins safety ever. He was very good at finding opposing QBs secrets and determining how they read the play. For instance, he figured out that Joe Namath would always watch the Free Safety’s break to read the coverage, and he would take a false first step to fool him. This would throw off Namath’s read, and would lead to bad throws, interceptions, deflections, or tackles that may be credited to other players, but it was Scott’s first step that changed the whole dynamic of the play. There is no statistic to measure that first step.

There is simply too many variables for statistics to really mean anything to the game of football beyond sparking some interesting debate.

Think about it for a second. Ever wonder why there seems to be such a clear divide between people in the NFL, and people out of it? Why coaches always roll their eyes at mentions of statistics? Why GMs worry more about "measurables" and "potential" then college statistics?

Think about every interview with a non-TO or non-Ocho Cinco player you've ever heard when they are asked about statistics and records. 95% of the time you'll hear something like "While it's an honor to be mentioned with (insert the guy who held the record before), i'm not really that worried about, i'm just here to help the team win."

This is because football players and coaches know the reality of the game, and the reality is that there's only two statistics that matter.

Wins and Losses.