There are many ways to tell when autumn is arriving. Days start to shorten, nights start to lengthen, the air starts to cool, children start to go back to school, and college football fans start to call for the heads of coaching staff.
And so it came to pass, in the dark days of October, in this Year of Our Lord Two Thousand Twelve, that the fans of my alma mater became restive, being disheartened in particular by a peculiar lack of defensive success, and a resultant lack of victories in conference play. And these fans began to exclaim, in the most agitated of manners, "rabble rabble rabble."
Some demanded blood. Some pleaded for patience. Statistics were hurled back and forth, quoted, dismissed, emphasized, reviled.
In the midst of this firestorm, I found myself combing through drive logs on ESPN and Yahoo Sports, plucking at this and at that statistic, and wondering if offensive performances could be examined in isolation from the rest of the team's performance, and likewise defensive performances, and whether or not such examination might allow comparisons between teams in the conference, and thus yield information of prognostic value. So I decided to put together a spreadsheet, courtesy of Microsoft Excel, into which I could input stats, allow the program to crunch the numbers, and obtain some quick and dirty analysis.
At first this seemed a relatively innocuous task - how often does an offense score? How many points per scoring drive? How many scoring drives does a defense allow? How many opponent points per drive? But then, insidiously, the realizations of how intertwined and entangled offensive and defensive performances began to creep in - how much blame can be borne by the defense for giving up points if the offense turns the ball over deep inside their own territory? How much by the offense if they turn the ball over deep in the opponent's territory, and then the defense allows a lengthy drive? How much easier did the defense make the offense's job by returning a turnover for touchdown? How much impact did special teams have on field position? How badly did that missed field goal hurt?
And the project veered alarmingly and rapidly in the direction of Excessively Complex.
Fighting desperately to keep things under control, while not simplifying the analysis to the point of uselessness, I ended up with a myriad of columns interconnected by a web of cross-references and containing a proliferation of Excel formulas I'd had no idea even existed before starting the spreadsheet. By the numbers, there are 14 input fields per team per game in the master input sheet, and a total of 54 additional columns of cells pulling info from the input fields and crunching data in various ways. These feed into separate offensive and defensive analysis sheets, which run to some 30+ columns themselves and which are finally funneled into two more offensive and defensive sheets devoted to reporting the data in the visual form of charts.
With the spreadsheet more or less in a state of completion, though, the work is only partway done. Over the next few weeks, as more matchups play out and more data is compiled, I hope to get a clearer picture of whether or not this thing is actually a useful predictive tool, a curious snapshot of team performance at a given point during the season, or a gigantic waste of time. At this point you know as much as I do, so stay tuned...
No comments:
Post a Comment