Football Statistics are Severely Overrated
August 18, 2010 § Leave a comment
In 2006 the Dow Industrial Average hit 12,000 for the first time in history under George W. Bush; it went on to hit 14,164 in October of 2007 the all-time highest close in history. Therefore George W. Bush was the greatest president for the economy. The stats do not lie, right? Of course they do. Just like the economy, the game as a whole cannot be pinned down to a stat, or series of stats, neither can football. Further the economy and football cannot be to a relevant point for the same reason, there are too many variables, and most of the contributing factors go unknown or are otherwise muddled to a point of being incalculable. It is a failure of football fans and analysts to give too much weight to stats; they are at best minute indicators of isolated contributing factors.
In other sports stats are more relevant because they are much simpler games. This all boils down to the fact that they have a very rigid structure, narrow ways of scoring, and are otherwise comparatively limited in functions of players. The rigid structure of innings in baseball and the short possessions of basketball does not allow for the teams to dictate the pace of the game and control of the clock, which is among the most integral parts of football. The narrowness of scoring is rather obvious. The “limited functions of players” is actually ambiguous. First it applies to the specialization of talents, where the different positions have entirely different skills, goals, and responsibilities from their teammates. Then it also applies to the differences in the way players are used from team to team, and the way the individual talents are not interchangeable within a position, but subject to overall systems with the skills, goals, and responsibilities open to coaching preferences.
Case Study: DeAngelo Hall
When he was an Atlanta Falcon he was amazing with pro bowl appearances in ’05 and ’06 and was otherwise a top tier guy in ’07. In ’08 the Raiders sent a 2nd round pick and a 2009 5th to Atlanta and made Hall the highest paid defensive back in the NFL with a 7 year 72 million dollar deal. “In his eight games in Oakland, Hall was beaten 40 times for 552 yards on 66 passes thrown his way, according to data compiled by STATS. He gave up more yards than any defender this season and was tied for third worst in catches allowed.*” So Hall got cut half way through the ’08 season by the Raiders and was picked up for the rest of that year by the Washington Redskins. Washington went on to sign him to a six year $54 million contract the following February. So what in the world happened to go from 72 million dollars for seven years, to getting cut and paid the 8 million of the guaranteed portion though jobless, to picked up for the rest of that season, to averaging 9 million a year for six years, oh yeah, and this time 22.5 million is guaranteed, almost triple?
The answer is simple; his skills fit what Atlanta and Washington do, but not Oakland. Both ATL and WAS primarily played zone coverage, where OAK preferred man, and well they should with one of the best man CBs in Nnamdi Asomugha. Another thing that the Falcons and Redskins have in common is they have their corners in off-man, several yards deep giving the wide receivers a cushion, where the Oakland Raiders like to play press (bump and run) coverage. Teams that play zone generally play-off man because it is easy to disguise coverages, so they are related. Basically from this it is clear that DeAngelo Hall is better at reading quarterbacks and receivers, than he is playing the more physical game against receivers; that is all.
You’re the Head Coach; Choose Your Running Back:
There is just under a minute left, and you have all your time outs. You are trailing by 3 at the opponent’s 32 yard line, and right on the verge of being comfortable kicking for overtime, a few more yards and you would be confident in it. It is third down and 3 yards. Your QB jammed a couple fingers on the last play and does not want to throw for it, the trainers say he’ll be fine in a couple minutes; your back up QB is injured, so you are forced to run it. You have two options for the running back, Steven Jackson or Chris Johnson.
Steven Jackson – Rams^
Stats Overview Rushing Fumbles
ATT YDS AVG LNG TD FUM LST
2009 324 1416 4.4 58 4 2 2
Career 1548 6707 4.3 59 41 20 13
Chris Johnson – Titans^
ATT YDS AVG LNG TD FUM LST
2009 358 2006 5.6 91 14 3 3
Career 609 3234 5.3 91 23 4 4
^ Stats taken from ESPN.com
“Obviously”, you want Steven Jackson. Chris Johnson is a higher risk/reward per handoff, where Jackson is more consistent with less break away runs. Jackson is a more downhill runner; he will not get caught for a loss too often, which would hurt your already shaky field goal range. Also consider that Chris Johnson’s record setting amount of long runs skews that average higher than what a typical run would be; the average is the mean, but the mode is more important and unseen by any stat.
This questioning of stat relevancy extends to every position on the field. Another example would be that Don Coryell, of the Air-Coryell offense, would probably prefer Randy Moss over Jerry Rice if he could draft either, even with the hindsight of knowing Rice’s career numbers. In his offensive system, receivers stretched the field vertically and Moss is regarded among the fastest to ever play the game, while Rice was, even in his earliest days, never the fastest receiver. Moreover, while Rice had better hands and was a better route-runner, Moss’ leaping ability would be more beneficial in the deep passing style game, simply outrunning and out jumping defenders. Conversely, Rice was the absolute perfect fit in Bill Walsh’s high efficiency, ball control passing game. Further, one could make a case that Walsh’s system helped extend Rice’s career, by removing speed as a top concern, and focusing on timing and route-running.
Basically the stats are vital and telling. They do not indicate nearly as much as people like to ascribe to them. There is a reason why teams do not seek out mathematicians to hire in their personnel departments. You have to watch what is happening, see how the players do in each of the almost unlimited amount of scenarios, and develop theories of what should happen. The key word is theory; there is no finite linear way of devising any plan in the game. There is too much happening on every play that all has to be taken into account, and cannot be translated into numbers to indicate an overall picture.