Writers of Pro Football Prospectus 2008

Most Recent FO Features


» Four Downs: AFC West

There's a serious need for defensive help in Kansas City, Los Angeles, and Oakland. In Denver, meanwhile, the Broncos must determine whether or not Case Keenum can really be a long-term solution at quarterback.

13 Jul 2005

The I-Formation: Offensive Bread and Butter

By Michael David Smith

Offensive formations are rapidly evolving, with three-receiver sets, four-receiver sets, and multipurpose backs who can line up anywhere and catch the ball as well as they run it. But the way Pro Bowl rosters and media depth charts show each team's starting lineup hasn't changed in decades. Every team is still shown with two receivers, a fullback and a running back, the formation that became standard thanks to the development of the I-formation.

The I-formation has been the dominant offensive formation in the NFL and has stayed effective for decades because of the way it balances the running and passing attack. We'll get into just how that balance is achieved, but first a little bit of history.


To cap the 1940 season, Chicago Bears coach George Halas used the T-formation (with the fullback directly behind the quarterback and the two halfbacks on either side of the fullback) to beat the Redskins 73-0 in the NFL championship game. Before long other teams saw that the T was more versatile than its predecessor, the single wing, and they adopted the T for themselves. By 1952, the Steelers (the last holdouts) switched to the T, and the single wing was out of pro football.

But around the same time, Tom Nugent, the football coach at Virginia Military Institute, had a new idea. Instead of having halfbacks on each side of the fullback, he thought it made more sense to line up the halfbacks behind the fullback. With the offensive backfield in a straight line, they looked more like an 'I' than a 'T', and a new formation was named.

After five years at VMI, Nugent brought the I to Florida State in 1954, and it began to spread, most notably to John McKay, who used the formation to win the national championship at USC in 1962. Before long, NFL teams took notice, and the I became the most common set in pro football.

Coaches who frequently run the I-formation often extol its virtues with statements like, "We just want to line up and come right at 'em." But there's nothing inherently tougher about the I than there is about any other formation. What set the I apart from the T and many of the other formations of the early days of football is that it lent itself to a versatile offensive attack. That versatility is still valuable in the modern game.

The Playbook

A typical bread-and-butter play from the I-formation might have a name like "I right 32 iso". That means an I formation with the right side being the strong side -- the tight end and the flanker (the wide receiver a yard behind the line of scrimmage) are lined up to the right, with a split end (the wide receiver on the line of scrimmage) on the left. The 32 means the halfback (or the 3-back) is getting the ball and running through the 2-hole (between the center and right guard). "Iso" is short for isolation, meaning the fullback is leading the running back through the hole and engaging the middle linebacker in an isolation, one-on-one block. (You know, the kind of block where the TV audience hears Ray Lewis complain that he was double-teamed.)

The offensive linemen block the players in front of them, with each tackle taking a defensive end, a guard blocking one defensive tackle alone and the other guard and the center double-teaming the other defensive tackle. The tight end blocks the strongside linebacker, and the two receivers might run a flag or a hook route to take the cornerbacks out of the play.

A play like I right 32 iso won't produce a lot of long runs, but that's not what it's designed for. If the right personnel executes it properly, it will accomplish one of two things: Either it will consistently allow the offense to gain four yards a pop (and an offense that can do that will march right down the field) or it will force the defense to bring the strong safety close to the line of scrimmage for run support. When TV announcers describe this, they call it "eight in the box, " and it means the defense is susceptible to the other bread-and-butter play of the I-formation: the play-action pass.

When the offense sees the strong safety playing run, it becomes clear that it's time to run the play action. The linemen and fullback have the same blocking responsibilities and try to make it look like a running play. The quarterback fakes a handoff to the running back, but instead of giving up the ball he goes into his seven-step drop. With the strong safety playing run support, the cornerback has no help in coverage on the flanker, who becomes the primary receiver on the play.

Let's talk a little more about the fullback, who usually lines up about four yards behind the quarterback, with the running back usually about three yards further behind. At seven yards behind the line of scrimmage, the running back can easily survey the field to see where the defensive players are lining up, and he can get a full head of steam so that he's at full speed when he hits the line of scrimmage. The running back also has time to watch blocks develop and can find cutback lanes.

The flipside, of course, is that a handoff to the running back in the I takes longer to develop than a handoff to the fullback, and that gives the defense more time to react. That's why in the early days of the I-formation, the fullback regularly carried the ball, especially in short-yardage situations. Today, however, football is a much more specialized game, and I-formation fullbacks are blockers almost exclusively and rarely carry the ball.

Which Fullbacks do it Best?

But even though he doesn't run much anymore, the fullback is still one of the most important pieces of the I-formation. If he isn't able to effectively use that isolation block on the middle linebacker, the play won't work.

The best game I've ever seen by a blocking fullback was Cory Schlesinger's performance against the Bears on October 20, 2002. The Lions had 39 runs for 192 yards, and nearly all of them were on I-formation runs with Cory Schlesinger providing the lead block on Brian Urlacher (James Stewart got most of the carries). When a fullback dominates a middle linebacker as completely as Schlesinger dominated Urlacher that day, an I-formation offense is nearly impossible to stop.

Unfortunately, football statistics -- even the most sophisticated ones like those we use at Football Outsiders -- are greatly lacking in their assessment of which fullbacks do the best job of providing lead blocks for their running backs. About the best thing we can do is examine each running back's I-formation splits and infer what those splits tell us about the fullbacks who block for them. Stats Inc. keeps track of how running backs do in the I-formation, although they don't tell you the context of those carries.

Here are the leaders in a number of I-formation categories, starting with total carries:

RB Carries Fullback
Shaun Alexander 193 Mack Strong
Domanick Davis 145 Moran Norris
Rudi Johnson 128 Jeremi Johnson
Willis McGahee 127 Daimon Shelton
Warrick Dunn 126 Pritchett/Griffith/Duckett
Jamal Lewis 126 Alan Ricard
Kevan Barlow 123 Fred Beasley
Ahman Green 120 Henderson/Luchey
LaDainian Tomlinson 109 Lorenzo Neal
Kevin Jones 104 Cory Schlesinger

Even though the Texans use the three-receiver set as their base offense -- Andre Johnson, Corey Bradford, and Jabar Gaffney are all starters -- Norris is the prototypical I-formation fullback. Last year he had one carry and four receptions, which sounds like no contribution at all until you realize how often the Texans counted on the 254-pounder to lead their running game.

Warrick Dunn's 126 carries in the I-formation were a lot, especially when you consider that T.J. Duckett had another 67. But when we look at the top running backs in terms of their yards per attempt in the I-formation, we see some interesting things, especially about the Falcons:

RB Yd/Att Fullback
Duckett 5.5 Pritchett/Griffith
Green 5.4 Henderson/Luchey
Bell 5.3 Hape/K. Johnson
E. Smith 5.3 Ned/Berton
Droughns 5.2 Hape/K. Johnson

Duckett was No. 1 in the league in yards per attempt in the I, and it's striking to compare the numbers of Dunn and Duckett when they're in the I and when they're not in the I. Duckett averages 5.5 yards a carry in the I and 3.8 yards a carry when he's not in the I. Dunn averages 4.3 yards a carry in the I and 4.1 yards a carry when he's not in the I. That demonstrates that Duckett thrives when he's running straight ahead into a pile, but is less effective when there's more open space around the backfield. Duckett never once got stuffed for no gain or a loss on his 67 carries in the I-formation.

Two of the top five running backs in yards per carry are Broncos running behind Patrick Hape or Kyle Johnson. For some odd reason Hape is listed as a tight end, and the TV announcers dutifully call him a tight end, but when you watch the Broncos you see that he actually plays fullback. He clearly does a good job of it, although if you don't already know that the Broncos do a good job of run blocking, this probably isn't the right website for you.

Next let's look at touchdowns in the I-formation:

RB TD Fullback
Holmes 14 Richardson
Davis 12 Norris
Martin 10 Sowell
McAllister 8 Karney
Johnson 7 Richardson
Tomlinson 7 Neal

Priest Holmes only took a handoff in the I-formation 75 times in his injury-shortened 2004 season, and yet he scored 14 times. Obviously, the Chiefs love using the I in goal-line situations, and Tony Richardson does a great job of leading Holmes and his backups into the end zone. On the opposite end, Marshall Faulk had 74 carries in the I and didn't score on any of them. Mike Martz generally doesn't like the I in goal-line situations, subscribing to the theory that it's better to spread out opposing defenses with an extra receiver in the red zone than to have them all crowding the middle of the field when a fullback is in.

Because the point of plays like I right 32 iso is to get four yards when the team needs four yards, if I had to choose one statistic to use to judge a fullback's effectiveness, it would be the percent of plays on which his running back got a first down from the I-formation. So let's look at that:

RB Carries FD FD Pct. Fullback
Martin 63 25 40% Sowell
L. Johnson 39 14 38% Richardson
Wheatley 35 13 37% Crockett
James 26 9 35% Mungro
McAllister 59 20 34% Karney

Sowell and Richardson have made plenty of All Pro teams, so it's no surprise to see them on the top of the list. Perhaps Zach Crockett, James Mungro, and Mike Karney deserve more credit for the roles they play.

Wrapping it Up

The I-formation has been the dominant offensive formation of the post-merger era, but it's far from universal. Don Coryell had success with the I early in his coaching career but later moved away from it with his high-scoring San Diego Chargers teams. Joe Gibbs, who played in college for Coryell at San Diego State, followed his mentor's lead and eschewed the fullback for an H-back who was more a threat as a receiver but not as useful as a straight-ahead blocker. Last year, Gibbs' Redskins rarely ran out of the I, and two of the other teams that most scrupulously avoided the I were run by Gibbs disciples: The Panthers with Dan Henning as offensive coordinator and the Vikings with Mike Tice as head coach and Scott Linehan as offensive coordinator. Most teams, even teams that prefer single-back sets like the Colts, still keep a full slate of I-formation plays in the playbook.

Although the offensive innovators (and the rule-makers) have made the NFL a more pass-happy league in recent years, the I isn't going anywhere. Fullbacks of today aren't as important to their teams as they were 30 years ago, but when a team needs a couple of yards, there's no better way to get them than having a lead-blocking fullback isolated on the middle linebacker.

Posted by: Michael David Smith on 13 Jul 2005

79 comments, Last at 08 Oct 2006, 10:49pm by Andrew


by JasonC23 (not verified) :: Wed, 07/13/2005 - 3:49pm

Bears fan here. Nice article...except for the mention of October 20, 2002. I can still see Schlesinger knocking Urlacher all over the field, at will. Ugh...

by Richie (not verified) :: Wed, 07/13/2005 - 3:54pm

Do you have a breakdown of percentage of plays run from the I overall, and by team?

It just doesn't seem like I see all that many I-formation plays.

by Richie (not verified) :: Wed, 07/13/2005 - 3:57pm

Also, I wish NFL Films (or somebody) would put together a piece on the history of offensive formations and strategies with film clips of the various strategies in action.

Watching things like the single wing, and wing-T in action are interesting. A few years ago, a local high school was still running the wing-T, but no more. The plays develop so fast. QB gets the ball, turns and hands it to a RB who lined up just a yard away.

My grandpa even talked about old formations where the QB really didn't even handle the ball that often. I guess the snap went straight to a running back. I've never seen this in action.

by michael (not verified) :: Wed, 07/13/2005 - 3:58pm

Excellent article. As a Chiefs fan, I am glad to see Tony Richardson lauded. He would probably be an outstanding feature back. Instead, he is perhaps the best fullback in the league. When the day comes to replace him, KC might find it more difficult than they know.

by mistamaxwell (not verified) :: Wed, 07/13/2005 - 4:01pm

I'd be very curious to see the overall effect that the play-action has on passing numbers (ypa, comp %) versus a 'regular' pass out of the I-formation (or other formations). My guess is that, overall, it's probably minor, if not negligible, due to overuse. Do STATS or FO chart such a thing? It's good to see the fullbacks being isolated and quantified here, though. It's about time there is some objective study done that shows us why Richardson or Sowell or Howard Griffith or Tom Rathman are making the Pro Bowl continuously.

by MDS (not verified) :: Wed, 07/13/2005 - 4:04pm

Richie, I don't have that kind of a breakdown. STATS probably owns that stuff, but they don't make it public. From the numbers I was able to get, I'd guess that about one-third of carries by a running back are in the I.

by zach (not verified) :: Wed, 07/13/2005 - 4:20pm

i've only ever seen a direct snap to anyone besides the QB as part of a punt fake.

by Jared (not verified) :: Wed, 07/13/2005 - 4:21pm

I am officially a FO junkie...keep it coming!

by Mike (not verified) :: Wed, 07/13/2005 - 4:23pm

Great article!

Question: It seem like (in the case of a strong side run) no one is taking care of the weak-side linebacker. Is that because Will typically can't get to the ballcarrier or the hole he is attacking in time to be a factor? Could you defeat the I by using more of a 4-2-4 type of alignment, where you have a good coverage weak side LB (almost more of a safety type), give him responsibility for helpin the weak side CB with the split end, move the free safety over to help the strong side CB with the flanker, and bring the strong safety up to stop the run?

It seems to me that, in addition to an effective fullback, two other key requirements for the I to be successful are at least one WR who can beat single coverage (otherwise there's nothing to stop the strong safety from moving up to stop the run all day), and good guards who can keep the hole from collapsing. Otherwise, the FB has to block a DT or a DE, and no one picks up Mike.

by Mike (not verified) :: Wed, 07/13/2005 - 4:27pm

I've seen a number of teams use the direct snap to a RB on running plays, but it's more of a gimmick than a regularly called play. For example, the Pats direct snap to Faulk about 1-2X per game.

by Sean D. (not verified) :: Wed, 07/13/2005 - 4:43pm

Re: 7 Actually, last year there were at least two direct snaps done on non-punk-fake plays. Twice the QBs tried to pull off the old "I'm gonna pretend like its too noisy and go yell at my WR while my RB gets a direct snap" move. Peyton did it once and Bulger did it once. When the Colts did it the play was called back on a false start by Manning who had already gone under center and thereby could no longer wander around yelling at people. The Rams were slight successful in that they got 4 yards from Faulk. Both attempts were against the Patriots. I also remember the Chargers attempting a direct snap to Tomlinson. This had worked in preseason, but it failed to produce yardage in the regular season, although I can't find any internet articles on it. Also, in Atlanta they direct snap to their running back all the time, what's that guy's name? Oh yeah, Ron Mexico.

by Pat on the Back (not verified) :: Wed, 07/13/2005 - 4:57pm

Great article. I do have a question (or possibly a nitpick), though, about 32 Iso. I always played that the fullback was the 3 back, both in HS and my limited experience in college. As the fullback I always loved "Dodger 31 Dive" (pro set formation, hand off to the FB running to the weakside off gaurd).

Basically, on any given running play, the offense has 9 guys to block the defenses 11 (because the QB and the ball carrier aren't generally blocking anyone). Since a run is usually designed to pick up only 4+ yards, then the pursuit of the Will LB usually can't get there until it is too late (that is, 4 yards downfield). For him to get there before the 4 yards on a sweep, he has to fight through the 5 OL, coming out to block, plus the TE and the FB. On a run to the strongside gaurd gap, the weakside gaurd will often try to chip the Will as the center takes the weakside DT. Of course, if the offside DT gets in the gaurd's way, then the center is usually off clean and can block the Will while the FB isolates the Mike LB (because the DT basically took himself out of the play). Either way, when executed properly and all the blocks hold up, one of the safeties will have to make the stop.
hopefully this lines up right:


You can see what I mean about the offside gaurd getting a clear shot at the W if the run is up the middle.

by Johonny (not verified) :: Wed, 07/13/2005 - 4:57pm

It seems to me some teams stress full backs that can be used in the passing game more than others. I wonder how the ussage of say Sowell and the Jets compares to a typical Gibbs style hybride TE.

by PerlStalker (not verified) :: Wed, 07/13/2005 - 5:03pm

re Patrick Hape

Hape is a bit odd. The catch is that Denver regularly lines up with their TE in the FB spot and shift him out to the TE slot. They also do the reverse. It's a particularly new thing there though as I recall them doing that a bit with Shannon Sharpe too.

It does give them even more flexibility in an already flexible formation. With TEs that play well at HB, they can line up with a single back and two TEs and shift one back and now they're in the I for a regular run. Or go the other way and your I has become a 2 TE set with a single back where the former FB is now faster into his pattern for passes.

It's fun watching Carswell shift back and forth between TE and FB. He's huge (especially since last year when he started training to move to tackle) but has pretty good hands so, unlike many lineman who line up at FB for power plays, he can shift out to TE and actually catch passes thrown his way.

I think the way Denver treats the FB/TE spots as a combined possision may become more common around the league because of the increased flexibilty.

by Sean D. (not verified) :: Wed, 07/13/2005 - 5:17pm

Re: 14

I think this how the Vikings use Wiggins/Kleinsasser as well. However, as we can see, they don't use the I formation quite as much, or I guess they never had one RB who ran out of the I as much. All the info above is either a rate stat for a particular RB or a counting stat for a RB. So it's hard to tell where the Vikings stand with their tag team attack of Williams/Moore/Bennett/"High-as-a-Kite" Smith. All we can tell is that none of them avg more than 5.2 yds/attempt from the I. But I would have to think that with a healthy Kleinsasser and their new 5 man attack (enter stage left Ciatrik Faison) they are fairly similar to the Broncos.

by MDS (not verified) :: Wed, 07/13/2005 - 5:20pm

"It seem like (in the case of a strong side run) no one is taking care of the weak-side linebacker. Is that because Will typically can’t get to the ballcarrier or the hole he is attacking in time to be a factor?"

Good question. I think you've got the right explanation. As I said, a play like the one I mentioned is more designed to get four yards dependably. The Will might be able to make the tackle, but usually not until the running back has gained four yards or so.

As for the name 32 iso, I looked at the few playbooks I could find, and although the numbering of the holes is just about universal, the numbering of the backs changes from team to team. So it could just as easily be 22 iso.

by ElJefe (not verified) :: Wed, 07/13/2005 - 6:12pm

I believe you alluded to this at the beginning of the article, but didn't the I formation originate with three running backs? Making it more of a variation on the T/wing-T/(later)wishbone?

If there were three backs, did they set in a straight line, or were the (2) halfbacks set side-by-side (if so, why isn't it called the L formation?).

You still occasionally see a play out of a three back I formation, but that formation uses two fullbacks. I've even seen an NFL game where a team used that formation on almost every down (NO vs. DAL, 1982). Oh, where have you gone Guido Merkens!

by RichC (not verified) :: Wed, 07/13/2005 - 6:17pm

Re: #7

The 2-pt conversion the Pats scored against Carolina in SBXXXVIII was a direct snap to Kevin Faulk who ran it right in.

Speaking of which, do QBs really think their (bad) attempts at "oops, the shotgun snap went over my head" fakes fool people? (or maybe from a DL/LB point of view they actually do....)

by B (not verified) :: Wed, 07/13/2005 - 6:24pm

I thought Manning's "confused" act for the patriots looked pretty convincing, cause that's what he always looks like. Of course, it turned out the play wasn't actually legal, but still, it was very convincing.

by dman (not verified) :: Wed, 07/13/2005 - 7:10pm

good article. I think this'll help my madden game alot. I like to run out of the I alot.

by Johonny (not verified) :: Wed, 07/13/2005 - 7:53pm

Another question. Anyone know if the Dolphins will feature more of a Viking look to their offense?

by Starshatterer (not verified) :: Wed, 07/13/2005 - 7:53pm

do QBs really think their (bad) attempts at “oops, the shotgun snap went over my head� fakes fool people?
I doubt the QBs believe that, nor do I think it's meant to genuinely fool anyone.

Motion tends to draw the eye, so any dramatic, unusual motion by the QB has great potential to distract at least a few defenders, for the fraction of a second it takes for such a play to succeed. Suckering a defender to chase the phantom botched snap would just be an added bonus.

by MikeT (not verified) :: Wed, 07/13/2005 - 7:55pm

Pat's explanation of the blocking on 12 sounds right. In HS and college, the weakside DE is sometimes left free on an iso or smash while that tackle climbs out to find the Will backer or a safety. In the pros, that DE will be fast enough to make the tackle in the backfield.

by David Keller (not verified) :: Wed, 07/13/2005 - 8:14pm

A poster asked earlier, "Do you have a breakdown of percentage of plays run from the I overall, and by team? It just doesn’t seem like I see all that many I-formation plays." I think I can address that. A breakdown happens when someone becomes overwhelmed with their circumstances and act out emotionally.

by Vince (not verified) :: Wed, 07/13/2005 - 8:16pm

Speaking of which, do QBs really think their (bad) attempts at “oops, the shotgun snap went over my head� fakes fool people? (or maybe from a DL/LB point of view they actually do….)

Hey, it's not like the QB has anything better to do. If it works once in a hundred plays, it's worth doing every time.

I agree that first downs is a better metric than total yards for grading fullbacks (or really, any blocker.) So if you were to design a "fullback rating," you'd want to include:
* The team's total first downs rushing out of I formation plays. (I would use totals as opposed to percentage because totals will reward FBs who are always on the field)
* Short-yardage run success (perhaps by 3rd down conversions?)
* receptions and/or receiving yards.

Does that sound reasonable to people?

by Basilicus (not verified) :: Wed, 07/13/2005 - 8:49pm

I'd be interested to know if there are any teams that semi-regularly run a fullback out of the I-formation. It'd be interesting to see how successful that is these days.

"do QBs really think their (bad) attempts at 'oops, the shotgun snap went over my head' fakes fool people?"

I've seen it work more often than not. Defenders key on the QB more than any other offensive player on the field. This is one reason why the play-action works so well...I'd love to see a half-hour highlight film of the greatest play-action fakes ever. Chad Pennington would have to be on there a few times, holding onto the ball and just wandering towards the sidelines with his back to the D.

Oh, and nice article by the way...you guys are really doing impressive and interesting work this offseason.

by Mike (not verified) :: Wed, 07/13/2005 - 8:56pm

#24--Hmm. I guess this was some kind of an attempt at a joke.

by PerlStalker (not verified) :: Wed, 07/13/2005 - 9:35pm

When Denver moved Mike Anderson to FB after TD returned, I figured they would run the ball with Anderson more. It's kinds scary for a defense, I think, to look in the backfield as see to 1000 yd backs. You could do a lot of fun misdirection things. Ah well.

by Steve Taylor (not verified) :: Wed, 07/13/2005 - 9:50pm

I didn't really start watching football until we got a pro team here in Nashville. Most sites I go to assume that I already know all this stuff. Thanks for putting out this series of articles; now I understand a lot more of the commentary and analysis I see. Keep 'em coming!

by Reinhard (not verified) :: Wed, 07/13/2005 - 10:52pm

For this analysis, was an I-Formation basically any formation with two players behind QB, one of them the fullback? Strong and weak I would seem to offer similar advantages.

Even if the weakside LB is unblocked, he may only reach the RB once he has crossed the LOS or maybe even gained a yard or two... and then dragging him down will take another yard or two... general rule of thumb for offence, block the most dangerous guys first. The opposite side linebacker and DBs... often it is better to block the other players harder rather than go for all one-on-one blocks; very effective if succesful, but more likely to break down.

by chris miller (not verified) :: Thu, 07/14/2005 - 2:43am

kordell stewart's tenure in pittsburgh was a source of constant frustration for me. double wings, wishbones, i salivate at what could have been...with stewart, ward, and bettis in the backfield, no formation is off the table. adding randle el to this mix would have increased the effectiveness of these 'throwback offenses' yet another order of magnitude. imagine sending your 'quarterback' in motion, leaving your 'halfback' in shotgun. mmmmm.
im not advocating an entire offense designed around the wishbone, but a couple packages of plays, maybe a dozen or so of these 'trick plays' every game. if nothing else, your opposing defensive unit will have to at least gameplan for them.

by israel (not verified) :: Thu, 07/14/2005 - 3:40am

An entire piece on fullbacks plus thirty-one comments, without a mention of a certain Pro Bowl alternate who wears 35 in Black and Gold?

MDS, what do your stats say about Mr Dan Kreider please?

by Vince (not verified) :: Thu, 07/14/2005 - 5:32am

I think, to look in the backfield as see to 1000 yd backs. You could do a lot of fun misdirection things. Ah well.

I don't think it would be as effective as you might think. John Madden wrote in one of his books (I think it was One Knee Equals Two Feet) that defenses had figured out that if they keyed on the blockers and not the runners, they wouldn't be fooled by misdirection plays. And that was about 20 years ago. Defenders have only gotten faster since then.

by Reinhard (not verified) :: Thu, 07/14/2005 - 5:40am

So hand off counter to the blocking scheme on misdirection plays... ?

by somebody (not verified) :: Thu, 07/14/2005 - 5:47am

RE: 33
Misdirection plays don't work huh? Then why did the patriots score 41 points against the steeler d by running end arounds, sweeps, and hb counters.

by Starshatterer (not verified) :: Thu, 07/14/2005 - 10:14am

Somebody (#35 )--

The misdirection runs had some success against the Steelers, but so did just about everything else: short passes, long bombs, long field goals, defensive scores. The Steelers (on both sides of the ball) were knocked off their game early, didn't even partly recover until the second half, and then only long enough to make it interesting for a while.

Once a team with as many offensive weapons and tricks as the Patriots do gets rolling, things get pretty ugly for any defense.

by zach (not verified) :: Thu, 07/14/2005 - 12:05pm

#11: Also, in Atlanta they direct snap to their running back all the time, what’s that guy’s name? Oh yeah, Ron Mexico.

lol... but does that qualify as a punk fake?

Actually, last year there were at least two direct snaps done on non-punk-fake plays.

by Pat on the Back (not verified) :: Thu, 07/14/2005 - 12:22pm

Misdirection plays and playfakes are two different things nowadays.

Playfakes (like playaction pass and fake hand-offs) are meant more as a way to freeze the strong safety if the defense requires him to aid in run support. This is the advantage of establishing the run, in that the SS has to make a split second decision about which responsibility he has on whatever play, and if he decides incorrectly initially that can make for a bigger gain. This usually doesn't effect the LBs. Essentially, think of a playfake as a watered-down flea-flicker.

Misdirection, on the otherhand, is more about blocking. More often than not (at least in a 4-3) the interior linebackers key off on the center/gaurds, which is why you will occasionally see a play where one of the gaurds pulls against the play direction and runs out to the middle of nowhere. The idea here is that the linebacker (or both, possibly) will follow the gaurd and take himself out of the play, and maybe even get in the way of the outside linebacker, thus negating their both of their pursuits. Basically, if the line blocks like the back is coming through a certain hole, but he ends up going somewhere else, that is a misdirection play. This happens on a lot of gadget plays, because you can get away with it once in a while, especially against aggressive defenses (see Philly and Pittsburgh). Indy also uses counter-intuitive blocking schemes a lot to sell Manning's fakes.

And actually, if memory serves, the Pats end-arounds had one gaurd pull to the left and the other pull to the right, essentially freezing the linebackers. I'll check up on that when I get home.

by MDS (not verified) :: Thu, 07/14/2005 - 12:26pm

"the I formation originate with three running backs?"

Yep, that was how Nugent originally lined it up, but teams quickly discovered that moving one of the halfbacks out as a flanker (or wingback) opened things up more.

Regarding Kreider, Staley was quite good out of the I (66 carries, 321 yards, 4.9 average, first downs on 24% of the carries). Bettis wasn't as good even though he got the ball more out of the I (94 carries, 345 yards, 3.7 average, first downs on 21% of the carries).

by Kibbles (not verified) :: Thu, 07/14/2005 - 12:43pm

Re #14:

I don't know about in the past with Shannon Sharpe, but I know that Denver didn't start using its TEs and FBs until a bit into the season. First, Anderson moved from FB to RB and then was lost for the season, and then eventually Droughns had to move from FB to RB, leaving the Broncos without a FB on their active roster (if I'm recalling the sequence right, that is). They signed Kyle Johnson to their squad, but in the game in the interim, they developed the shifting TE/FB look. Hape performed well there, so they kept him doing it for the season. Despite lots of playtime at the FB, though, he is strictly a blocking TE, and will likely not see much time at FB at all now that Johnson has a year of experience under his belt.

Speaking of Carswell, though, I can't wait to see the Broncos have Carswell start reporting to the referees that he's lining up at Tackle, but he's one of the eligible recievers on that play. They did it once with Cooper Carlisle 2 years ago against SD, but it'll be so much more effective with Carswell on the field. Plus Carswell as an offensive lineman would have to make the funniest decoy I've ever seen. I'd laugh to see an OLB break off and try to cover a guard or a tackle as he's rumbling down the field.

For all the flak Shanahan gets, he's still one of the most creative offensive minds in the game.

by James, London (not verified) :: Thu, 07/14/2005 - 1:03pm

Another Fine Minicamp. Assuming you have a good FB/H-Back, what are the 'cons' of the I Formation? If you have a credible play- action threat and a TE with good hands then you can approximate a 3WR set from the I.
Why isn't the I used more often, particularly as a good FB is so much easier on the cap than a good 3rd WR?

by MDS (not verified) :: Thu, 07/14/2005 - 1:41pm

I think it has to do with offensive coordinators thinking they can get better mismatches in the passing game if they line up three receivers and therefore make the opposing D bring in its nickel package. Even on running plays, I think a lot of coordinators believe they'll have more success if they spread the field and leave their running back alone in the backfield than they will if they have a fullback blocking but also have an extra defensive player in for him to block.

by Sean D. (not verified) :: Thu, 07/14/2005 - 1:55pm

RE: 40
If Carswell is declared as an offensive lineman then he can't be "rumbling down the field" unless he wants to get the Broncoes penalized, which I'm fine with, but I don't quite understand the strategy.

by ZasZ (not verified) :: Thu, 07/14/2005 - 2:10pm

Sean (#43),

Any player can go downfield, if they declare to be an eligible receiver. The ref announces it, so it can't be used as a trick play much, but if you have a lineman with great hands, it could be worth it, just to spread confusion on the defense.

by Chris (not verified) :: Thu, 07/14/2005 - 2:18pm

Absolutely terrific article. Exactly the type of thing I wanted to see.

by Mike (not verified) :: Thu, 07/14/2005 - 2:35pm

Re 38,

That's a nice clarification. Thanks. However, isn't there's some risk associated with such misdirection plays, because your pulling guard leaves a hole that a clever defense will recognize and exploit?

I remember a Patriots game last year or the year before where the opposing team was on like the Pats 4 yard line, and they tried to do just that. The defense wasn't fooled, and Bruschi came right up the middle and got a crucial sack-fumble that essentially put away the game. Does anyone remember who the opponent was? (I ask because I know there are a number of Pats fans on this site, and also because I recall TMQ mentioning it and I know a lot of folks here read TMQ). My gut says it was either Buffalo or Indy, but I'm not sure. Anyway, the unsurprisingly uninformed sports media was all over this poor guard, saying he blew his blocking assignment, but it really, really looked like he intended to pull because it was supposed to be a misdirection play.

by person (not verified) :: Thu, 07/14/2005 - 2:43pm

Question above:

If you have a credible play- action threat and a TE with good hands then you can approximate a 3WR set from the I. Why isn’t the I used more often, particularly as a good FB is so much easier on the cap than a good 3rd WR?

I believe that developing the potential of the I-formation in this way is the idea behind the original "West Coast" offense designed by Sid Gillman and exemplified by Don Coryell's Chargers, then, later, the Jimmy Johnson Cowboys. (Not to be confused with the short-passing offense of Bill Walsh's SF 49ers, which is now what is usually meant by "the West Coast offense".) Lots of play-action, and the FB is more of a threat to recieve than to rush, though still playing a role as lead blocker. And the TE is also expected to be more of a reciever than a blocker; play-action makes up for his shortcomings. This has been around for a while and has a reputation as a high-scoring scheme. (Corrections/confirmations welcome.)

Why bother with a 3wr set when you can just use your TE as the extra receiver by lining him up closer to the edge of the fields? Well, the TE doesn't only require sure hands in this case, he also must have the ability to consistently "get separation" from the defender just before the ball is thrown. There are many more mediocre WRs who have these skills and might also blossom into starting-calibre recievers if given more playing time, compared to the relative shortage of sure-handed TEs who also have the agility required to be anything more than a safety valve. So, a scheme which lets you develop your mediocre recievers is beneficial to the club, in addition to its other merits (novelty, forcing the defense to substitute, etc.)

by bravehoptoad (not verified) :: Thu, 07/14/2005 - 3:16pm

Back in the day, when Denver had Terrel Davis, Olandis Gary, and Mike Anderson, and all three were 1000-yard backs, I kept wanting them all to just once line up in the wishbone. Would it have been the only time in the history of the NFL where three 1000-yard backs were in the backfield at the same time?

re #31: Yes I, too, am disappointed that Pittsburg keeps drafting these versatile atheletes and then not versating them enough. If you've got three quarterbacks on the field at once then why not do something cool with them?

by Vern (not verified) :: Thu, 07/14/2005 - 5:03pm

Re: 46
Yep. It was Buffalo. Not only did the guard pull, leaving Bruschi open to go straight up field, but Travis Henry decided to continue his fake too, watching Bruschi go right by. Strip sack and Seymour returns it for a TD the other way.

Re: 42
This suggests the shift in thinking is that its easier to get players to execute clear-out routes than it is to get them to execute great blocks. From that, you can also see how critical it becomes on defense to be able to disrupt these routes, rather than say, being able to overpower a good block.

This tends to align with increasing trend for good defenses like the Pats to be better at versatility and decision making. I wonder if strong I formations with good old fashioned overpowering fullback blocks do better against them? I remember the Colts running all over them in last year's opener. Can't recall the formations.

by Tim (not verified) :: Thu, 07/14/2005 - 5:03pm

Re #44:

That's not really true. If I understand the procedures correctly, the refs get a list of eligible receivers from each team at the beginning of the game. Then the guys who are announced during the game as eligible (the Kevin Barrys, Jumbo Elliots, and Refrigerator Perrys of the world) are the ones who line up in an eligible position but who were not announced as eligible at the beginning of the game. I believe there may also be a stipulation about the jersey number a permanently eligible player (such as a WR or RB) must have, but I'm less sure about that.

The other requirement for being an eligible receiver is that the player is uncovered. (It's like Mahjong.) In other words, you can't have a guy lined up at the same depth as you on both your right and your left. That's why the wide receiver on the strong side is lined up a yard off the line of scrimmage - it allows the TE on that side to be an eligible receiver.

The other thing is that there have to be at least (or maybe exactly, but I think at least) 7 offensive players on the line of scrimmage. Typically that means the 5 offensive linemen are in the middle with an eligible receiver (WR or TE) on either side, with additional receivers perhaps lined up off the line. Conceivably, if you really wanted your left tackle to be eligible, you could line up with a TE on the right and a WR on the line outside of him, thus making the TE ineligible and the tackle on the other side of the field eligible. Teams sometimes get penalized for having an ineligible receiver uncovered, which is what that would be if you didn't declare your tackle eligible before the play. I think if you did declare him, and proceeded to cover up your tight end, the defense just might sniff out a trick play.

I can't believe I just wrote that much about rules. My apologies to anyone who read this far down.

by Vince (not verified) :: Thu, 07/14/2005 - 5:19pm

RE #40 and a few others: Let's clarify who the eligible receivers are on any given play:

* The three men in the backfield who don't take the snap (usually RB, FB and a flanker)
* The QB is eligible, but only if he takes the snap from shotgun. I swear that's the rule, don't ask me why.
* The men on either end of the offensive line (usually one TE and the split end)

And that's it. In addition, offensive linemen are declared ineligible no matter where they line up, UNLESS they declare themselves eligible to the referee. HOWEVER, even after reporting, they must still line up on the end of the line or in the backfield to be an eligible receiver. Under no circumstances can the center or guard be an eligible receiver.

(Unless you do something truly whacky like putting the center on the end of the line and all the other linemen on one side of him, and I don't even know if that's legal anymore. I know it used to be.)

by Pat on the Back (not verified) :: Thu, 07/14/2005 - 5:31pm

Why isn’t the I used more often, particularly as a good FB is so much easier on the cap than a good 3rd WR?

The good FB may be cheaper than a good 3rd WR, but the 3rd WR is probably cheaper than the good 3rd CB needed to cover him.

numbers 50-79 are considered ineligible, as are 90+. Usually, Numbers are given by position as follows:
K,P,QB,RB,H-Back, etc (in the backfield): 1-49
WR,TE (eligible on the LOS):80-89

Also, you are right about the covered/uncovered thing. In HS, as a FB, I wore 63 (and was permanently eligible). One gadget play was to have me line up at the normal weakside tackle position, and move the weakside tackle to the usual TE position. Basically, it was like the gaurd snapped the ball and I was the TE. Nobody ever covered me, easy 40 yards.

by PerlStalker (not verified) :: Thu, 07/14/2005 - 5:36pm

re 51

I think it's still legal but it is a bit unsafe, I would think. Linemen are too fast these days and would probably be all over the QB before anything had a chance to happen.

I have seen in recent years the "trick" FG formation of just the center and two guards lined up in the box with everyone else lined up outside the hashes. I don't think I've ever seen a play work that used that but I have seen the formation a couple of times.

by Basilicus (not verified) :: Thu, 07/14/2005 - 6:46pm

"The misdirection runs had some success against the Steelers, but so did just about everything else: short passes, long bombs, long field goals, defensive scores."

I think it isn't that trick running and other trick formations don't work against the quicker, smarter defenses, it's just that most teams don't have both the discipline to have 1: all their players on O (starter and backup) know the trick plays backwards and forward and execute them perfectly and 2: the coaching wherewithall to not just plan trick plays for their offense, but to also plan these offensive trick plays specifically against particular opposing defenses (something that Belichick and company are excellent at.)

by Kibbles (not verified) :: Thu, 07/14/2005 - 8:13pm

Re #48: Back in the day, when Denver had Terrel Davis, Olandis Gary, and Mike Anderson, and all three were 1000-yard backs, I kept wanting them all to just once line up in the wishbone. Would it have been the only time in the history of the NFL where three 1000-yard backs were in the backfield at the same time?

I don't really see the point, since only one of them can run the ball on any play. That said, Denver did play a significant portion of the game against the Chiefs 2 years ago in the wishbone (the one with the huge Dante Hall return, iirc), and the Chiefs looked like they didn't have a clue how to react. I know two of the backs were Portis and Anderson, and there's a good chance that the third was Droughns, so if we're talking revisionist history, then there were 3 1,000 yard backs at the field, although at the time one only had 80 or so career yards.

by Trogdor (not verified) :: Thu, 07/14/2005 - 8:17pm

An example of a misdirection play is one that I've seen a lot the last few years. The line (and FB if applicable) block as if it's a run up the middle, while the QB pitches the ball to the HB running wide off tackle. Since most D-linemen key off the blocks instead of watching the QB or backs, they will crash the interior, leaving the back essentially 1-on-1 with the LB with contain responsibilities. A talented back can bust some big runs from this play, and if the D overcompensates to guard against it, it opens the middle of the line even more.

As for the unblocked defender on an iso play, he can crash the interior to make a tackle, but as mentioned it will often be a few yards downfield. In addition, the offside LB or DE (opposite the direction of the run) will usually have contain responsibilities. If he crashes the interior too often, the offense can respond by running an end-around. If the defender loses contain, it's big yardage for the offense. Just an example of the cat-and-mouse game that goes on every play.

Also, the Bills were the team I most remember using the direct snap to a RB. Thurman Thomas was incredibly dangerous running from the shotgun, which we know is not a baseball formation.

by J (not verified) :: Thu, 07/14/2005 - 9:07pm


I am guessing the stats provided for the I-formation are only the base I-form; not the weak I, the strong I, the big I (extra TE).

Is my guess right?

by vince (not verified) :: Thu, 07/14/2005 - 9:41pm

#57: Actually, the I think the splits from stats are single back, split backs and I-form. So I think "I formation" would include ALL I formation plays -- base I, strong I, weak I, power I, etc.

by Vince (not verified) :: Thu, 07/14/2005 - 10:58pm

Clarification: My last note should have read "splits from STATS, Inc." Just in case any of you don't know who STATS are.

by Reinhard (not verified) :: Fri, 07/15/2005 - 12:02am

probably none of us are right :)

by Israel (not verified) :: Fri, 07/15/2005 - 9:48am

#50 writes "The other requirement for being an eligible receiver is that the player is uncovered. (It’s like Mahjong.) "

Someone thinks footballoutsiders.com is a place for Mah Jong as a point of reference? What are we coming to?

by MDS (not verified) :: Fri, 07/15/2005 - 11:19am

The Bills were very good using the direct snap to the running back. I remember Kenneth Davis had a big run for a first down on third-and-long in the Super Bowl using that play. Those Bills offenses were something. It's odd how they really haven't spawned any copycats the way most great units do.

by Scott de B. (not verified) :: Fri, 07/15/2005 - 11:36am

"I am guessing the stats provided for the I-formation are only the base I-form; not the weak I, the strong I, the big I (extra TE)."

Some info on how these variations are used would be very nice.

by Wanker79 (not verified) :: Fri, 07/15/2005 - 12:28pm

“Iso� is short for isolation, meaning the fullback is leading the running back through the hole and engaging the middle linebacker in an isolation, one-on-one block. (You know, the kind of block where the TV audience hears Ray Lewis complain that he was double-teamed.)

I just couldn't let that slip by without anyone taking note of it. lol

by zlionsfan (not verified) :: Fri, 07/15/2005 - 4:24pm

FWIW, the versions of NCAA Football I have include the three-running-back I formation as the Maryland I. I remember seeing Indiana use it (in the game, not in real life). Watching the plays unfold on-screen, you do get the feeling that the original I is slower than the modern I.

Having typed that, I'm looking at it and thinking it's obvious. Either you've got three RBs in the space where you had two before, or your HB is lining up another three yards behind the line of scrimmage. At least in the 'bone you spread the backs out a bit so they're not that far from the QB.

by MDS (not verified) :: Fri, 07/15/2005 - 5:36pm

Tom Nugent, the inventor of the I, coached at Maryland in the 50s and 60s, and that's how he lined up his running backs. So that's why it's called the Maryland I.

I really don't play video games, but from what everyone says here, maybe I should start. It sounds like they're a great way to pick up on the Xs and Os of the game.

by Kibbles (not verified) :: Fri, 07/15/2005 - 5:53pm


Video games don't really give any clues about how line play really winds up working out, but it's a great way to get a good feel for WR routes, zone and man defense, and the whole concept of "creating mismatches". All of the tricky line concepts, such as stunts and twists, tend to be ineffective, as linemen pretty much pick a direction for their first step and then engage in a block. You also never see any double-teams on blocks, either.

A lot of defensive concepts, such as containment and the QB Spy, also don't seem to play out the same in games and in real life, either.

Still, there is some element of reality, and it's always fun pretending you know what you're doing and designing your own playbook. I have to admit that Romeo Crennel stole the concept of a 2-5 defense that he opened the SB with from me. I've been using it with a lot of success since Madden 2003. The pass rush options are almost limitless, and zone defenses are really potent. Best of all, great OLBs are a lot more cap friendly than great DEs.

I'm still waiting for him to give me the credit I'm due. :(

by Basilicus (not verified) :: Fri, 07/15/2005 - 10:08pm

Wasn't Crennell's arrangement the 2-4 or the 2-3, not the 2-5?

by Reinhard (not verified) :: Fri, 07/15/2005 - 10:29pm

Strong I - the fullback lines up over the guard on the strong side (the side the TE is on)
Weak I - the fullback lines up over the guard on the weak side
Power I - instead of a flanker (Z reciever?) there is an extra fullback in the backfield. I think this is the NCAA "Maryland I" mentioned above.

The reason I said that it would make sense to include the weak and strong I is because they have very similar pros and cons to a regular I.

by J (not verified) :: Sat, 07/16/2005 - 1:44am

Earlier (post #57), I asked what I -formation variations were included in the stats. The reason, I found it surprising to see the Steelers only ran out of the I-form 160 times + a few others by Haynes (55 total rushing attempts), Parker (32 total rushing attempts), Kreider (4 total rushing attempts), and Brown (1 total rushing attempt). Haynes was primarily a 3rd down back, so most of his 55 rushes would have came from the shotgun. Parker saw some rushing attempts from the I-form.

The Steelers ran a very limited playbook last year. They seemed to run alot out of the I-form (or variations of the I - weak, strong, big, 3 WR). They used some singleback-two TE formations, some singleback trip bunch form (mostly passing, but some rushing), goal line form, and variations of the shotgun (3, 4, and some 5 wide...mostly passing). If my memory is right, that is about all the formations they used, atleast on a regular basis.

My point, according to the numbers, the Steelers ran 160 times out of the I with Bettis or Staley, plus maybe 30 additional rushing attempts out of the I by Parker, Haynes, Brown, and Krieder. The Steelers had 618 rushing attempts. Subtracting the attempts by WRs or QBS takes that total to 534. So only 35% of the Steelers rushing attempts came out of the I formation. This just seems low to me, I guess I am wrong. I would like to see more stats on the Steelers rushing attempts out of the I-form.

by Kibbles (not verified) :: Sat, 07/16/2005 - 3:13am

Re #68:

I'm pretty sure it was 2 DL, 5 LB, 2 CB, 2 Safeties.

It really wouldn't make sense for NE to field a 2-3 defense. That would mean they opened with 6 DBs on the field, which I deem highly unlikely if Troy Brown is your nickle CB already.

by dryheat (not verified) :: Mon, 07/18/2005 - 4:30pm

#71 -- correct. Although sometimes it was a 1-7. The strength of Belichick's defense is the LB corps. Those are his versatile athletes, so the more you can get on the field, the greater confusion and athletic mis-matches you give the offense. When the Offense breaks huddle, sees Richard Seymour over the ball, and behind him in a two point stance sees Vrabel, McGinest, Colvin, Bruschi, Phifer, and Johnson, two or three of which are going to be blitzing and others dropping into coverage, it's easy for the O-linemen to flub their blocking assignments and the QB can't do a pre-snap read.

by Phoebe Buffay (not verified) :: Tue, 07/19/2005 - 1:14am

I'm totally mad at #38, 56,60, 70. I never kissed ross. He kissed me!!

by cjm (not verified) :: Tue, 07/19/2005 - 3:43pm

re 51, 52: the Colts ran a center-eligible fake punt once last year - I think it was against the Pats. They lined up with the center and two guards over the ball, with the guards both a step back from the line, and then a bunch of other players over on the right side, and a player or two over on the left as well, off the line. HS teams sometimes use this formation for punts, mostly because of the added fake possibilities, but usually the throw goes to one of the players over in the bunch, not to the center. The Colts did attempt the pass to Jeff Saturday, but it was incomplete.

I had never heard this about having to declare players eligible to the refs, so I don't know how they worked that. Seems like they must have put Saturday on the eligible list before the game, else the declaration right before the play sort of telegraphs your intentions.

re 71, 72: The Pats used a 0-6 or 0-7 alignment a number of times against the Colts in the playoffs, and I thought they used it again once or twice in the SB. No down linemen at all.

by Zac (not verified) :: Tue, 07/19/2005 - 10:22pm

Regarding declaring eligible players. I think some times get around the fact that it telegraphs your play by doing it whenever possible. i.e. they'll declare someone eligible whenever possible when he's not doing anything special, in the hopes that when they do declare him eligible for the purpose of throwing to him late in the game, he will get ignored.
The Packers always declared Kevin Barry eligible in their Jumbo set, and I don't think they every threw to him.

by lobolafcadio (not verified) :: Thu, 07/21/2005 - 6:24am

About Will and SS being not blocked, all has been said about Will, and there's usually a big difference between a RB and a SS. Even the SS known as "big hitters" can't be compared with RBs. Another thing, coaches usually teach RBs to go straigth forward the first 4 yards and then outside (kind of corner route) to avoid the pursuit, doing this, you avoid the Will and arrive on the safety you can quietly beat with a fake (cadrage-débordement) or most simply hit to teach him to respect you. If you spend all the afternoon to send your big RB against their SS, you will gain a lot of yards and put their SS completely out the running and passing game.

by Rudy Pyatt (not verified) :: Wed, 01/25/2006 - 11:48pm

Nice article, great site! A couple of points:
The "modern" two-back "I" is also known as the "California 'I'" because John McKay took the up-back out of Nugent's three-back set. This put the halfback deeper with the line reading advantages already described.
The true MdI set is three RBs inline behind the QB, w/two TEs on the line. Great for power off-tackle: you don't need to pull linemen when you have two leads in the backfield. And the backs are tougher foor Ds to read than in the CalI. Like the T, the misdirection (AND play action possibilities) can and should be done QUICKLY. That's why FB dives don't work well anymore: they don't hit quickly. But that's another rant...

by Joe Pisarcik Magnet (not verified) :: Sun, 08/20/2006 - 2:11am

I seem to recall seeing a diagram of the Power-I where the fullback lined up behing the QB, the tailback lined up behind him, and there was another back to the FB's right.

Whither the pro set?

by Andrew (not verified) :: Sun, 10/08/2006 - 10:49pm

Regarding direct snaps: many college teams that use the shotgun as a primary formation use the direct snap as a quicker hitting run than the standard delays and draws, especially when they lack a running quarterback.
Also, NCAA has two three back I formations: in the Maryland I, the fullbacks are lined up straight, while the power I offsets one like the strong/weak I.