I hope this isn't useless gibberish.
Yesterday, there was some conversation about advanced statistics, clutch, and a bunch of other things revolving around a link that OCC provided in his post. The post was by John Morgan of Advanced NFL Stats. After sifting through the post, and being completely honest I feel like I had to open up my dictionary to get some of his points, I found the gist of the article to be true. That message being that statistics in sports has evolved. We need to move on and stop using the big play to determine greatness. The smaller sample size that is a single great play is not worth more than consistent production over the course of a game, or better, a season.
This paragraph is the key point of the article.
But truth in the form of information has hunted fiction down, and in a desperate attempt to survive, mythmaking has turned to scandal mongering. Some scandals seem trivial, others seem life or death; some probe into the ever-shrinking possibility of a private life, others attempt to judge the quality of a person by their performance as an athlete. This last one, for all its presumption and prejudices and plain meanness, statistical analysis targets. And this last one, the realm of "clutch" and "overrated" and "the will to win," is vehemently guarded by otherwise smart and rational people, who deny new kinds of perspectives the way Rollingstone Magazine denies anything relevant happened in rock 'n' roll since 1975. But to those smart and rational people: Your time is passing. The recent kerfuffle over which hat Colin Kaepernick wears are death throes. Look around, the business of cooked up controversy and baleful character assassination is, day-by-day, more desperate, decrepit and irrelevant.
As a fan of advanced statistics I 100% backup John Morgan's article. There's something to being able to perform under pressure, but sometimes it's just luck and needs no conversation. I think that sports need more people who actually watch sports, rather than those who watch highlights on ESPN after the game. A highlight only tells so much of the story. The game is the entire story, but even then you have to dig deeper to understand it. That's where advanced statistics come in.
A brief history on me and advanced statistics. I am a huge fan of football, but I'm also a huge fan of baseball. Anyone following baseball recently has probably heard of sabermetrics. For years (and it's still going) there has been a divide in the baseball community. The sabermaticians, and the traditionalists. The traditionalists see the saber community as a bunch of nerds trying to ruin the game that they love. "Math? And baseball? A ratio can't hit a curveball. lol NERD." Yet, what the traditionalists fail to realize is these "nerds" love the game also. We love the game for different reasons, and we want to see the best product out on the field. Baseball has come up with a statistic that measures a players complete game. It measures hitting, defense, baserunning, pitching, and position value. It's called WAR. For a better in-depth understanding here's FanGraphs glossary's definition.
But baseball has a value system that we can weight. A single is one base, a double is two bases, a triple is three bases, and a home run is four bases, to put it very simply. It's more complex than this, but I don't think anyone wants to know all that badly. The scenarios can be mathematically adjusted in baseball, which is something I find very difficult for football to duplicate.
Like I explained in OCC's post, football has more scenarios and variables. My exact comment is below.
I agree. Football is much harder to find meaningful statistics for because the outcomes are valued differently and they rely on a teammate most of the time.
How can we value a 3rd and 10 pickup of 11 on a screen pass differently than a 3rd and 10 perfectly thrown and timed out hook route for a pickup of 11?
A sack for a loss of 3 yards is worth more than a sack for a loss of 1 yard. The scenarios are harder to predict because the many different outcomes. It’s just too hard to weigh.
Stats in any sport can be deceptive, particularly more in football. For the sake of example, Tony Romo could throw what seems to be a 75 yard touchdown on the stat sheet. When in reality, all he did was throw a three yard screen to DeMarco Murray who did all the work. Or perhaps Tony Romo could lead the Cowboys all the way down to the one yard line, only for Murray to punch it in on his first carry of the drive.
On plays where a quarterback or runningback succeeds, those players must rely on their teammates to do their job. If an offensive linemen doesn't block well enough, Romo can't throw it to Dez Bryant, who has to catch the ball to finish the play. If the line doesn't get enough push, then Murray is going to be stopped for a loss in the backfield. Baseball has it easier in that every player accounts for himself. Derek Jeter can't help Robinson Cano get a hit. Yet, Tyron Smith can help Romo complete a pass. There is virtually no way to value teammate production in a statistic.
My theory: Possibly I can help begin this process. I'm not a mathematician, so I'm just laying out the groundwork. Someone might have already posted this idea. In that case, sorry, you get the Nobel Prize instead. My idea is that football creates some replacement offensive line theory. This is for quarterback and runningback statistics. It will level the playing field. Every QB and RB is based on what they would do if they have a league average offensive line. If a quarterback has a weak offensive line he's more likely to turn the ball over more. So making it fair for the worst and the best would level the playing field, thus making it easier to say who's the best. I'm not sure how to compute this though, because there would be some crazy formula involved.
Here's another baseball reference. In baseball a player with a higher RBI total, but with other similar offensive statistics can seem to be the better player. However, the players ahead of the batter getting on base is out of the batter's control. Player A can have a guy with a .380 on base percentage in front of him, while Player B could have a guy with a .290 OBP in front of him. Player A is going to drive in more runs. Sounds a lot like if a guy on the defensive line has a lot of quarterback pressures, but no sacks. QB pressures are valuable, yet everyone only talks about the finished product. A pressure could lead to a sack, interception, or fumble. A sack could be the result of a teammate clearing the way for you to get that sack. *Cough* Aldon Smith and Justin Smith. *Cough*
Football is also a small sample size. Teams only play 16 games, a maximum of four more if they're lucky. That right there is an extremely small sample size. It's the reason we see guys be the best one years and duds the next. Was Chris Johnson really good, or was he just a product of a small sample size? Or maybe even a product of his teammates?
In the end, I find it incredibly impossible to find the perfect system. Heck, baseball isn't perfect in their stats, but they're close to perfect. Football will probably never get there because of the so many different outcomes. A first down is 10 yards, but you can pick up 20 yards. That's worth a lot more than a first down where you only gained 10 yards. It's a team game, so more analysis should be focused on the team. However, general managers want ways to comprise their teams with the best talent possible. Advanced statistics would be that step. Now, I'm not saying advanced football statistics are flawed. They help for what it's worth, but the whole story is still not being told. There's values in the game that cannot be weighted, so I can't see them every dominating the sport like in baseball.