The NFL is either at the start of an option revolution or in the throes of an option fad. It's impossible to determine which without a crystal ball. Many offenses are either installing option packages or perfecting the ones they already put in place, and every defense is scheming to stop them. Everybody has an opinion, but few have any data. Just how many option plays did teams execute last year? When did they execute them? How effective were they?
With the help of the Football Outsiders Game Charting Project, we can answer a few of those questions. Option plays were very effective last year, though they were not as prevalent as you might think, and the wave of the future -- or perhaps the craze of the moment -- often amounts to little more than a spiced-up handoff.
Option to Hand Off
The Football Outsiders Game Charting Project recorded 463 plays last year that were categorized as some kind of "option:" a read-option, quarterback option, triple option or some other tactic in which the quarterback has the choice of giving the ball to a teammate or keeping it himself.
Right away, we run into classification issues. Option plays are a little like pornography: We know them when we see them. A game charter must sometimes make a judgment call about whether a play was really an option at all. If Charlie Batch jogs a few paces to his left after a handoff, does that really mean he had an option to keep the ball himself? Separating the many species of option plays causes more problems than it solves, because they can start to look like snowflakes to even the trained eye, and nobody has both Greg Knapp's and Kyle Shanahan's playbook on the shelf. "When in doubt, leave it out" is the best principle for making judgments based on game film. There were 463 plays that were obvious examples of options.
There were 32,883 offensive plays in the NFL last year, so 463 plays are not a heck of a lot: 1.4 percent. Even if we account for some option plays that were mislabeled, or passing plays built from the option (more on those later), it's impossible to make options add up to as much as five percent of NFL offense in 2012. That said, the percentage of option plays in the NFL hovered around zero for decades, which makes the sudden spike noteworthy, as does the fact that three option-heavy teams reached the playoffs.
Here's a basic overview of how successful option plays were, and just who ended up with the football:
|Total Plays||Yards Per Play|
|Runs by QBs||157||6.81|
|Runs by RBs||277||5.41|
|Runs by Others||29||6.53|
Many of the "others" were Redskins receivers, most notably Brandon Banks, who either lined up temporarily in the backfield or went in motion to get involved in the option play. A few others were Wildcat players who could in no way be considered quarterbacks (cornerback Patrick Peterson of the Cardinals, for instance).
On a per-play basis, options were very successful. The average NFL rushing play gains 4.3 yards. Option plays fare 38 percent better than typical rushes. In fact, rushing averages are at their highest point in decades and up about one-tenth of a yard in the last two seasons. There are multiple factors causing that, but the rise of the option is one of them.
The other surprise in the table above is that options result in a run by the quarterback only one-third of the time. That percentage is probably even lower, because any misclassified option plays are probably runs by a running back (the quarterback keepers are easy to spot). Most modern-read option offenses are designed to make the handoff the go-to decision: If the quarterback has any questions about his read, or if the defense is hanging back to stop a pass, the quarterback is supposed to give the ball up. The 33 percent ratio suggests that a team could run 30 options per game and expose its quarterback to only 10 hits, and the quarterback would have some control over when he is hit, and by whom. That doesn't eliminate the "options will get quarterbacks killed" theory, but it does show that the stats are not as stacked against quarterbacks as they seem.
Options by Team
The next table breaks down option plays by team.
|Team||Options||Yards Per Play|
The Redskins, of course, embraced read-option concepts and were very successful running them. The Panthers were more successful on a yards-per-play basis, and ran nearly as many option plays. The Panthers have been running options for two years, and opponents were very wary of their option package after the team averaged more than 25 points per game in 2011, so the evidence that defenses can quickly adjust to stamp out the option is not very strong.
The four teams that used the option regularly, with a starting-caliber quarterback pulling the trigger, all experienced success with the system. Six yards per carry is a lot, especially when the other features of modern option plays are factored in. Actual pitches at the line of scrimmage (a staple of the old collegiate triple option) are rare, and the quarterback does not carry the ball often, so many option plays are just handoffs with the window dressing of a quarterback acting like he is up to something. An extra 1.65 yards per play is a heck of a reward for a pretty small risk.
Here's a breakdown of option plays by down and distance:
|Situation||Options||Yards Per Play||First Downs/TDs|
|Goal to Go||21||2.94||7|
Options by Defense
You may be curious about which defenses fared well or poorly against option plays. The answer is that no defense faced enough options to produce a sample size that any number cruncher would be comfortable with. The next table breaks down options against each team's defense. As you can see, no team faced more than 45 options last year
|Defense||Options||Yards Allowed Per Play|
All the table really shows is that teams that often faced the Redskins and Panthers had to cope with lots of options. The Saints faced the Panthers twice, plus the Redskins and 49ers. The Giants got a double dose of the Redskins, plus the Panthers and 49ers. It would be great to draw conclusions from their relatively large samples, but let's see. The Saints were dealing with Bountygate and drew the Redskins in the season opener, when no one knew what to expect. The Giants beat the pudding out of the Panthers and walloped the 49ers when Colin Kaepernick was still a Wildcat gadget, then had a love-hate relationship with Griffin for two games (lots of yards, but three forced fumbles).
The Saints defense performed relatively well against the option, which may be significant because NFC South teams knew it was coming. (The Buccaneers did great in a too-small sample, the Falcons poorly in the same circumstances.) There is no evidence that defenses can pounce all over options if they know they are coming, but there is evidence that many teams were facing something for the first time last year. As options increase, defensive attention will increase, and it only makes sense that yards-per-play will stabilize. As of last year, no team devoted enough defensive attention to demonstrate that they had some kind of "anti-option" capability.
Option to Throw
Nearly every option play in the data above was a rushing play. Football Outsiders game charters only list a passing play as an "option" if there is absolutely no doubt about it. Otherwise, we would be scratching our heads at every play-action pass and making too many judgment calls. But teams that use the option also have passing plays built off those options, and all of that play-faking in the backfield should, theoretically, have an impact on the passing game.
One brute force way of examining how teams like the Redskins and Panthers used the option to set up the pass is by counting the number of shotgun play-action passes each team attempted. Many of these passes are built from the team's option formations. It's a crude instrument, but it's the best we can do without pretending we can read Rob Chudzinski's mind or watching every single play of the 2012 season a dozen more times.
Our final table shows all of the shotgun play-action passes attempted by the "big four" option teams. The league average is also given for comparison.
The amount of Panthers and Redskins play-action shotgun passes tells us that we are on to something. Only the Broncos used such plays as often as the top two option teams, and Peyton Manning was probably using them as his own sort of "option" (a pre-snap audible instead of a post-snap keeper play). The Seahawks and 49ers were closer to league average, which is still relevant: Many NFL teams attempted fewer than a dozen shotgun play-action passes, so 25-35 attempts, while technically close to the league average, are quite a lot.
Our four option teams used shotgun play-action passes between 33 percent and 44 percent as often as they ran options. The Redskins' and Panthers' ratios are so similar to each other that it is spooky, and it suggests the shape a team's stats will take if it makes a serious commitment to the option. The Redskins and Panthers each executed about 15 plays per game that started with a quarterback in shotgun preparing to handoff to a running back. Around seven of those plays became passes, five or six became handoffs and two or three became quarterback keepers. Fifteen plays out of 60 or 70 does not seem like a lot, but even the most committed college option team does not run an option-style play every time, and it is hard to imagine any NFL team making a both-feet leap into such a system. If Chip Kelly is hedging and adjusting, then everyone will.
The data above shows that the Redskins, Panthers, Seahawks and Niners were very effective play-action shotgun passing teams, and their option rushing games almost certainly helped. Keep in mind that the "average team" contains not just tons of Newton and Griffin but a lot of Manning and Tom Brady (another fan of the shotgun play-fake), so there is not a lot that is "average" about it. Also, the QB Ratings in the table are as useless as ever in small samples, overreacting to the fact that the Niners threw zero touchdowns, the Redskins threw three interceptions and the Seahawks completed 81.5 percent of their passes. Four teams with inexperienced quarterbacks got substantial productivity from this type of pass -- another byproduct of the option that will catch the eyes of both offensive and defensive coordinators. Best of all, the sack and scramble totals were manageable, so the quarterbacks are not getting brutalized.
Ravens general manager Ozzie Newsome said on Super Bowl Media Day that he believed the NFL option would soon go the way of the run 'n' shoot, the passing offense that took the NFL by storm in the late 1980s. Newsome was voicing the opinion of many skeptics: Defenses would figure the option out and pounce all over it, and offensive coordinators would decide that it wasn't worth the trouble.
Ozzie Newsome is not right all the time, but he is Ozzie Newsome all the time, and his opinions are not to be taken lightly. Defenses around the league will be devoting more practice and meeting time to stopping the option. The element of surprise will be long gone for teams like the Redskins. There is going to be some statistical settling: six yards per run and 11 yards per pass attempt are not sustainable long-term figures.
But here's the interesting question: What really happened to the run 'n' shoot? It clearly did not die. Teams used four or more wide receivers on 1,249 offensive plays last year. Before the run 'n' shoot, teams spread the field with four receivers only on third-and-30. Now, it's just another tool in the toolbox, a tool that a team like the Packers uses on 10 percent of snaps. (Other teams used four receivers more often, but the Packers were less likely to use it as a play-from-behind strategy.) The run 'n' shoot seeped into the DNA of NFL offense. There are no teams that employ eight 5-foot-9 receivers and zero tight ends anymore, but when a team spreads the field so a tiny receiver can run a hitch route against soft coverage, we are watching a run 'n' shoot staple.
The current option trend is nothing like the Wildcat fad of 2008-10, although the Wildcat did help blaze the current trail. The 2008 Dolphins used the Wildcat 84 times and averaged 5.7 yards per play. That was the high-water mark for both usage and productivity for that scheme. Current option plays are better integrated into the offense ("surprise" quarterbacks are rare), more versatile and have already been more widely adopted. The Wildcat is essentially a series of about a half dozen plays. Modern options have the potential to be full systems, like the run 'n' shoot was.
My hypothesis is that the Redskins, Panthers, 49ers and Seahawks will continue to use option tactics this season, at about the rate they ran them last season. (The 49ers may up their usage a bit with Kaepernick as a full-year starter; the Seahawks are committed to being as multi-faceted as possible.) The Eagles will use options at a rate similar to the Redskins and Panthers. The Jets, Bills and Dolphins will continue to tinker: Rex Ryan has talked about a read-option system for Geno Smith, both coach Doug Marrone and quarterback EJ Manuel have some experience with it in Buffalo, and Ryan Tannehill can run and had a little option success last year. Other teams will continue to dip their toes now and then. There will be bad offenses that run options terribly, and great ones that run it well. Defensive tactics will adjust in the next season or two, but no foolproof strategy will emerge, because there really are no foolproof strategies in the NFL.
Option plays (runs and passes built directly from them) will eventually level out and become tactics that some teams use about 10-15 percent of the time, other teams not at all -- just like four-receiver formations. By 2017 or so, we will not think much of it when Johnny Manziel fakes a handoff and sprints to the edge. It will be something that some quarterbacks do five times per game, some zero, something that happens in playoff games and in meaningless December games between five-win teams. The option will survive because, under the hood, it's just some running plays and play-action passes with a little extra zest. It will survive because it is versatile, and it works.