Okay, truth be told, this blog is really about big data analytics and the Cleveland Browns. However, if you’ve read even this far, then you’re probably either a Browns fan or a Browns hater. It really doesn’t matter to me which though, so long as you keep reading because it really is an interesting intersection of the two. Moreover, the rest of the story is more about the use of knowledge anyway (or in this case data).
To set the stage, prior to the 2014 NFL draft held with such great fanfare but a few weeks ago, it was widely reported that the Browns spent north of $100,000 to have a big data study conducted on the characteristics of a quarterback in this year’s draft most likely to be successful in the NFL. Now it would seem to me that in the big money league that is the NFL, $100K to them is more like change lost in the sofa cushions, but for some reason it was big news. So given that, what the Browns did after spending that kind of coin surprised a lot of people. Maybe that was really the big news.
According to one source, the Browns commissioned an “expensive and thorough analytics study” of quarterbacks stretching back decades using a variety of variables. Now in the interest of transparency, I know something about “expensive and thorough analytics study using lots of variables” because my Master’s thesis was on the subject of predicting the winning betting strategy for NFL games. Suffice it to say that since I’m working in knowledge management today and not at some sports book in Las Vegas it’s an indication of my success at that (although thankfully I did graduate).
So what’s the point—especially if you’re not a Browns or even an NFL fan? It’s that after all the machination Teddy Bridgewater was considered by the study to be the quarterback who will have the most success among the QBs in this year’s draft class. Why does that matter you say? It’s because at the start of the draft the Browns had both the #4 and the #26 picks overall, and---used neither to pick Bridgewater! More specifically, they traded their #4, ending up at #8 instead—and didn’t pick Bridgewater. Later, they traded their #26, ending up at #22 instead—and didn’t pick Bridgewater.
Who they did pick was Johnny Manziel, whose legal name in some parts of the country is apparently Johnny Football. WTH? Johnny Football? They just threw $100K down the NFL drain? So what did the (presumably) knowledgeable leadership team of the Browns know that they big data analytics study didn’t? I don’t know—I’m just a fan that keeps buying tickets and “waiting for next year.” What insights perhaps did big data not have that the Browns leadership have in their “gut”? Does this portend that human expertise and wisdom is still valuable, even in this onslaught of big data analytics? Or does it mean that big data analytics isn’t quite ready for prime time yet?
Certainly there have been celebrated cases of big data success—like the Target example of a few years back. What was different in this case? Well, most of us may never know unless we can cough up $100K ourselves to commission our own study to find out what advanced analytics was used. Nevertheless, one must assume that this was a very serious study, by very smart people, using highly sophisticated techniques for arriving at their conclusions.
In the end however, we may not know for several years. And, even then intervening variables may have come into play that confound the original predictions. As a Browns fan, I definitely hope that man triumphed over big data analytics this time, but I also wonder if IBM’s Watson had been given this task (at a cost probably several times this study) what the answers might have been? In defense of big data analytics, who has a success story that they can cite?
A final note: not only was Bridgewater not the first quarterback taken, he wasn’t even the second, and he was actually the last pick in the first round. I, for one, can’t wait for the football season to begin!
You can follow me on Twitter @KM_dude.