In this week's Varsity Numbers, Bill Connelly revisits some measures and concepts: Adjusted Scores, Covariance, and momentum (or whatever you choose to call it).
01 Dec 2004
By Michael David Smith
Listening to sports radio after the Colts dismantled the Lions on Thanksgiving, I heard someone say, "Peyton Manning is great at the play-fake." This ought to join "Barry Bonds walks a lot," "Shaq gets fouled a lot," and "There will be no hockey this year" on the list of things that need not be said on sports radio, because no one who listens to sports radio could possibly lack this knowledge.
What I wanted to explore, as I watched and then re-watched every play of the first three quarters of the Colts' 41-9 win -- before Manning headed to the bench and I headed to the dinner table -- was why the play-fake works so well for the Colts. Certainly, other teams have quarterbacks who could fake handoffs to the running backs. What do the Colts do differently?
The answer was summarized quite nicely by regular Football Outsiders commenter Trogdor, who sent me an e-mail saying, "Something I haven't seen discussed is how well Indy's line helps sell the play action. They go all-out to make it look exactly like a run, even doing things like pulling guards and trap blocking. It has to be incredibly hard on a defense when the whole team is selling the fake like that."
So today I'll analyze just what the Colts' offensive line does so well. Let's start with the first drive, which was a thing of beauty. On a second-and-one run by Edgerrin James, the whole line shifted to the left as James ran to the left sideline. Nothing unusual there, except all five offensive linemen hesitated before crossing the line of scrimmage. Because of that, the Lions' linebackers couldn't tell whether it was a pass or run by looking at the Colts' line, the way linebackers usually do. That wasn't particularly helpful on the James run (in fact, it prevented any linemen from blocking the Lions' James Davis, who made the tackle after a four-yard gain). But it perfectly set up the next play, in which Manning faked the handoff to James on the left side, all five linemen made identical blocks to the previous play, and Manning found a wide open Marvin Harrison for a 23-yard gain. That's what the Colts do: Their offensive line prevents the defense from making reads and forces defensive players to guess whether a run or a pass is coming. The drive ended with another play-action pass, this time from the Lions' four-yard line. The offensive line moved left, Manning faked left to James, and the play action brought Davis, the linebacker, toward James. Manning threw directly into the spot vacated by Davis for the touchdown. If Davis had stayed back and prevented Manning from throwing, you can bet the Colts would have noticed that Davis was looking to stop the pass, and the next play would have been a run.
Some have criticized the Colts for throwing in goal-line situations, but it seems clear that the Colts are simply ahead of the opposing defenses. They'll keep throwing until it stops working. That doesn't appear to be coming any time soon. In fact, if anyone should be criticized for throwing at the goal line on Thanksgiving, it's the Lions on their subsequent possession. After a five-yard gain by running back Kevin Jones brought up second-and-goal from the two-yard line, the Lions called two passes, both fell incomplete, and they settled for a field goal.
The other accusation that has surfaced against the Colts, that they run up the score late in games, isn't borne out by the evidence. The Colts have outscored their opponents this season by 39 points in the first quarter, 55 points in the second quarter, 35 points in the third quarter, and eight points in the fourth quarter. Does that sound like a team that runs up the score to you?
On the first play of the second drive, the Colts went right back to their old tricks. Center Jeff Saturday pulled left as Manning faked left to James, then threw to Harrison, who was open to the right. It was Manning's only bad pass of the day, and it fell incomplete. But it illustrated the way the pulling linemen set up the play-fakes, and, no surprise, it worked as the Colts continued to march downfield for their second touchdown.
Offensive linemen are often described as if they're merely interchangeable parts, and that isn't fair to them. But the Colts' line really does look like it could operate with just about anyone. Manning's fifth touchdown pass was also undrafted rookie center Trevor Hutton's first play. Manning lined up in the shotgun and the line didn't protect very well, with the Lions getting plenty of pressure up the middle. But Manning quickly released the ball to Harrison streaking across the middle of the field, underneath the secondary, at the seven-yard line. The play illustrated why Manning is so hard to sack -- it's not that the offensive line protects so well, it's that Manning releases decisively and the Colts' offense always has a receiver available on a hot route.
The Indy line's technique of not blocking downfield doesn't only explain why the play-fake works so well. It also explains why James is so effective at picking up a few yards consistently, but so rarely breaks long runs. Consistent rushing success is about passing the defensive line, which the Colts' offensive line always allows James to do. But long runs are about passing the linebackers and secondary, and James rarely has blockers in front of him once he crosses the line of scrimmage.
As much as selling the run seems to work, the Colts' offensive line is smart enough to adjust on the fly, especially veteran tackles Tarik Glenn and Ryan Diem. On Manning's sixth and final touchdown of the day, left tackle Glenn read the play beautifully. It was yet another play-fake, this time with James running to the right. Ordinarily that would call for Glenn and the four other linemen to go right with James, but Glenn saw that a backside rush was coming, so he abandoned the play fake and sealed off the backside while his linemates pretended to run block. The Lions' linebackers and strong safety all bit on the run fake, leaving Fernando Bryant one-on-one against Harrison. The word "mismatch" is overused in sports, but it couldn't be used more appropriately than in describing Bryant one-on-one against Harrison.
As Trogdor mentioned, pulling guards and trap blocking are an important part of the pass protection on play-action passes. The Colts have played the past two weeks with a pair of rookie guards, Ryan Lilja and Jake Scott, but it hasn't seemed to limit what they can do offensively. Scott, particularly, is light on his feet but strong enough to block at the point of attack when he has to. He was a 280-pound tackle at Idaho a year ago, and last week he was winning the battle with the strength of the Lions' defense, the huge tackles Shaun Rogers and Dan Wilkinson. Much credit should go to Colts offensive line coach Howard Mudd, who was a Pro Bowl lineman with the 49ers in the 1960s. He clearly figured out that the way to beat a big line like Detroit's with small linemen like Lilja and Scott is to use trapping and pulling instead of running right at them.
The Colts' offensive line doesn't overpower anyone, the way the Chiefs' and Packers' lines do, but it's a unit that deserves a bit of the credit as Manning re-writes the record book.
Finally, a closing observation: Manning picked on Dre' Bly all day, starting with the first play of the game, a 20-yard completion to Reggie Wayne. Bly was also beaten on two of Brandon Stokley's three touchdowns. Bly made the Pro Bowl last year because he's one of the best in the league at capitalizing on opposing quarterbacks' mistakes. But when the opposing quarterback doesn't make any mistakes, Bly gets beaten. It was an odd spectacle last weekend as the two worst corners I saw were the two NFC starters in February's Pro Bowl: Bly and Champ Bailey, who was repeatedly torched by Oakland's Jerry Porter.
(Ed. note: The Indianapolis off-tackle draw was also analyzed in this guest column from 2003.)
Each week, Michael David Smith looks at one specific player or one aspect of a team on every single play of the previous game. Standard caveat applies: Yes, one game is not necessarily an indicator of performance over the entire season. If you have a player or a unit you would like tracked in Every Play Counts, suggest it by emailing mike-at-footballoutsiders.com.
1 comment, Last at 31 Jan 2007, 9:15am by betting online