Introducing a data-minded digest for Cowboys fans
A little over a year ago, I started the @CowboysStats Twitter account. At the time, I hoped it would grow into a platform to share my findings with what I expected would be a small but passionate group of nerdy Cowboys fans.
I had no idea how many of you were out there.
For years I had followed closely the work of Brian Burke, whose old website, Advanced Football Analytics, broke my brain and forever changed the way I thought about the game. Before ESPN hired Burke a few years ago to work on their Stats & Info team, he devised a couple of simple but powerful models estimating the value of each location on the field based on down, distance and a handful of other factors.
Burke wasn’t the first to look at the game of football this way—many of the core concepts were laid out decades earlier by former NFL quarterback Virgil Carter and in a book called The Hidden Game of Football—but Burke and another group of researchers at Football Outsiders did a lot to demonstrate the power of this approach.
The conclusions Burke and others drew from these expected-points models inspired a whole new wing of football analysis—from the widely accepted conclusion that NFL coaches are far too conservative on fourth-down, to the growing consensus that the passing game is underutilized in most situations, to a host of other public findings by a scrappy group of amateur and professional data analysts on Twitter and elsewhere.
I’m one of those amateurs. My professional and educational background isn’t in data analysis, but in newspaper reporting. I was drawn into this conversation by a relentless obsession with big data in sports. What limited skills I have in this arena—from applied statistics to basic coding—were mostly self-taught.
Yet you all embraced my place in the Cowboys conversation. Today, barely 15 months after I started posting regularly on Twitter, @CowboysStats has nearly 10,000 followers, an audience I genuinely never thought it would reach. I have formed friendships and learned a ton from the online communities of Cowboys fans and writers, Seahawks Twitter posters (,,,) and NFL numbers-crunchers.
I’ve also benefited enormously from the efforts of the incredible team behind nflscrapR, including Ron Yurko, Maksim Horowitz and Sam Ventura. By making nflscrapR a free and public series of tools for gathering and analyzing NFL data, these folks have made it possible for random enthusiasts like me to replicate and build on the work of Burke and so many others who came before us.
I’d be remiss if I didn’t thank Pro Football Focus’s Timo Riske—from whom I’ve learned so much about football and data analysis in the past year—and Ben Baldwin of The Athletic, who first made me aware of the resources available through nflscrapR. Without Baldwin’s evangelism, I’d probably still be copying and pasting EPA values into spreadsheets, 50 rows at a time.
I have a lot of ideas for what this newsletter could become. Most writers on Substack start a free publication with the intention of adding an optional paid subscription down the line. If the initial response is positive enough and I feel I can add enough value, I may eventually add such an option as well.
More realistically, though, I expect this newsletter will be a slow-developing side project that will be free to all readers for some time.
For now, I want to gauge interest and explore what types of content work best on this platform. My goal is to share cool research and walk that tricky tightrope of figuring out what hard data can tell us about our Dallas Cowboys and their place in a changing NFL.
What you can expect if you sign up:
longer, in-depth Cowboys analysis pieces featuring charts, tables and other visualizations;
running logs of meaningful playcalling tendencies for both the Cowboys and their opponents;
summaries of the biggest plays and playmakers from each game, through the lens of expected points and win probability;
recurring roundups of the latest findings from the NFL analytics community;
and other types of content as it occurs to me.
Let me know what other types of analysis you’d be interested in seeing here. If you’d like to support this project, the best way to help is simply signing up for the free mailing list at the big blue button below.
Thanks again!
—Daniel H.
