Many of you will know the stuff I'm about to put up here, but for people who have watched football for years and had their heads filled with the nonsensical babbling of so-called "analyst" or are new to the game or just want to know more and be a better football fan, here's a quick FAQ to get you up to speed. Most of this research has been conducted by Football Outsiders, which was preceded by Bud Goode, Pete Palmer, John Thorn and Bob Carroll and the latter three's work: "The Hidden Game of Football".
How can 16 games be a meaningful sample size?
Aaron Schatz handles this question here. The short answer is that while a football season only contains 16 regular season games, those games are much more dense with information than a single baseball game. Over the course of a season most quarterbacks will record 500+ pass attempts, most running backs 300+ touches and so forth. Naturally, as with any statistical analysis, the less plays a player records the more unpredictable their future performance.
What's wrong with traditional stats (yards, TDs, YPA etc.)?
Yards, touchdowns and so forth aren't meaningless, just very incomplete. The perfect example of this is the 3rd and 13 draw to Max Strong for 11 yards. Traditional stats list this as 11 yards, a 11 ypc average--a very successful run. In reality, the utility of this run is very limited. It doesn't pick up the first down. It likely was only allowed because the defense was anticipating a pass and sent their linebackers deep. It certainly doesn't imply that Strong will be able to maintain a high yards per attempt going forward.
In football everything is contextual, and ever play must be adjusted for not only the game context but for the opponent's individual strengths. So, running an 11 yard draw against the St. Louis Rams on 3rd and thirteen isn't worth much, but running a 5 yard draw on 3rd and 4 against the Chicago Bears is.
In summary: traditional stats fail to account for game situation and strength of opponent, therefore, are not only inaccurate at assessing past performance but can be misleading when predicting future performance.
Is it better to be lucky or good?
In the long run it's better to be good, but luck plays an underappreciated part in determining any single contest. Here's a short list of common football occurrences involving luck:
- Fumbles: Causing fumbles/fumbling is not luck, but largely skill, but recovering fumbles/failing to recover fumbles is largely luck. Once a ball is loose, where it bounces and who ends up with it is a very meaningful, but also almost entirely random. Some fumbles, like a botched snap, are more likely to be recovered by the offense, but as a rule of the thumb: fumbles/forced fumbles are a meaningful stat while fumbles lost/fumble recoveries are not.
- Then again:
A defense’s tendency to force turnovers is fairly important to the team’s success, but it seems to be even more unpredictable. In general, a team’s ability to force fumbles seems to be almost entirely luck. There is a little bit more persistence in a team’s ability to force interceptions, though it isn’t clear how much of this ability is just a residual effect of general defensive ability. Again, this only pertains to the team level. Whether there are some individual players with a special ability to force turnovers significantly above average rates would be an interesting subject for further study.
- Third Down Success: Some teams will over/underachieve on third down offensively/defensively over the course of a few games and sometimes over the course of a season. In 2004, Seattle was consistently worse in third down situations than they were in first and second down situations. In 2005, their third down performance returned to the level expected given their first and second down performance and the offense surged. Because third downs are disproportionately important compared to first and second downs, a team can look better/worse than they actually are simply by being lucky/unlucky in third down situations. With few exceptions, third down success eventually falls in line with first and second down success.
- Completion percentage and interceptions: I haven't read a formal study WRT to completion percentage and interception percentage, but generally speaking a quarterback with a poor completion percentage but few interceptions will eventually see their interceptions "catch up" to an expected level given their completion percentage.
Does a team need to establish the run to win?
Short answer: No. In fact, what people are reacting to is the simple fact that rushing attempts and winning have a high correlation. With those attempts often comes big rushing numbers, but the attempts themselves, however successful, correlate better with wins than the rushing yardage. The reason, simple enough, is that when a team is ahead they tend to run the ball more. And that's it. A team does not need to establish the run to win, they need to outscore their opponent, in the big picture the passing game/the pass defense is a much more important indicator of a team's success. If Peyton Manning throws for 500 yards and 6 scores in the first half, no one cares but his fantasy owners that Joseph Addai didn't record a single rush.
Here's a potpourri of facts dispelling other football truisms of dubious value:
- Third year receivers aren't much more likely to break out than 2nd or 4th year receivers.
- Pace, i.e. how long a team takes before snapping the ball in between plays, is all but meaningless to the outcome of a game.
- QB rating isn't perfect, but it also isn't worthless. Still, DVOA >> Qb Rating.
- Offensive holds may be frustrating, but they are far less damaging than sacks. If a hold truly prevented a sack, all-in-all it's a net positive.
- For a kicker, accuracy is less important than simple kicking power. Accuracy is erratic, a player like Neil Rackers can be a shaky kicker for years before posting one of the great kicking seasons in recent memory as he did in 2005. In other words Hawk fans, don't expect Josh Brown to ice every last second field goal attempt again, but he'll continue to be valuable for his strong leg.
That's all I have for now, I'll clean this up, add links and add more info in the next couple of weeks.