A Wednesday night in May, a little after eleven. The quality control office smells like dry-erase
In This Chapter
- The Hook: The Empty Arrow
- 38.1 A Play Is Half a Sentence
- 38.2 The Engine Room: Base Runs and the Plays That Protect Them
- 38.3 The Passing Core and Its Own Constraints
- 38.4 The Floor Under Everything: Protection, and the Ball as a Blocker
- 38.5 Tempo: Not a Play, a Multiplier
- 38.6 The System Map: Every Answer Has a Price
- 38.7 The Four Anchors as One System
- 38.8 Closing the Empty Arrow: The Middle-Read Tag
- 38.9 Keeping the System Honest: Self-Scout, Tendency-Breaking, and What to Cut
- 38.10 The Minimum Viable Offense, at Every Level
- Spaced Review
- Project Checkpoint: System Map
- Chapter Summary
- What's Next
Chapter 38 — The Complete Offense: How Run, Pass, Screen, Play-Action, RPO, and Tempo Connect into One System
The Hook: The Empty Arrow
A Wednesday night in May, a little after eleven. The quality control office smells like dry-erase markers and the second of my two coffees. Teo Arriaga is standing at the whiteboard with a red marker, and I'm standing next to him with a blue one, and we're playing a game I've played with every young coach who has ever worked for me. I call it the Answer Game.
The rules take one sentence. I name a play; he names the defense's best answer to it; I name the play that punishes his answer; he answers that; and we keep going until somebody runs out. Teo plays Russ Kimbrough. He's earned the right. He tagged every snap Sacramento played against us this season.
I write Inside in the middle of the board. Inside zone, from Eleven, Gun Doubles Right. The run our whole offense is built on.
"Holloway fills the alley," Teo says. He doesn't even look at his notes. "Two-high, but the field safety triggers on the guard's first step."
I draw an arrow out of Inside and write Inside PA, Z Post. "Then Trey runs the post through the grass Holloway just left. Thirty-five yards. You've seen it."
"Fine. Holloway stops filling. He sits at ten and plays the pass first."
"Then you just took a man out of the box." I draw an arrow back to Inside. "Six in the box, Jalen has five on five, and we run it until you put him back."
"Okay. I walk Holloway down to eight yards before the snap. Eight in the box."
"Seven or fewer, run it. Eight, take the throw." Arrow to Inside Slant / Inside Glance. "Your Will has a run fit and a pass window on the same side. He can't have both."
Teo smiles. He's been waiting. "Ames sits. He plays the window."
"Then he isn't in the box, and Darius hands it off." Arrow back to Inside.
"Man," Teo says. "Everybody's got somebody."
"Mesh. Stacks. Lace. And the quarterback runs when every back is turned."
"Zero. Six coming, nobody deep."
"The zero bin. Two hot throws, then the max-protect deep over. Fifty-two yards."
"A mug. Two in the A-gaps, one of them drops, the corner comes."
"Center on the Mike, back on the other backer, right tackle widens, and Trey sight-adjusts to the hitch when his corner leaves. Plus ten, first down at the Sacramento 49."
By now the board looks like a spider web. Every play has arrows coming in and arrows going out. Teo caps his marker, uncaps it, and writes one more line in red, low on the board, by itself:
Third-and-nine. Two-high pre-snap. Twenty-nine straight times. At the snap, I rotate.
I reach for the board. I draw the arrow. And I stop, because there's nothing to write at the end of it. On January's call sheet, there wasn't. We had Mills in that box, the post and the dig, and the post was the right throw against two-high, and Darius threw it into Ike Royce's chest at 2:41 of the divisional round. The arrow on the whiteboard points at white space.
"That's the one, Coach," Teo says quietly.
"I know."
"Every other answer Russ had, we had a play that made him pay for it. That one, we had a feeling."
He's right, and that empty arrow is the reason this chapter exists. Not because one interception ended our season. Because it showed me, better than any win ever did, what an offense actually is. It isn't a list of plays. It's a web of promises: for every answer the defense can give, a play that makes the answer cost something. Where that web is whole, a defense can only choose which way to lose. Where it has a hole, a smart defensive coordinator will find it, and he'll save it for the snap that matters most.
I left the arrow on the board all summer. It's still there. Under it, in blue, it now says middle-read tag. You'll see what that means before this chapter is over.
🏃 Fast Track: If you've coordinated an offense, you already believe in constraint plays. Read the threshold concept in 38.1, the System Map in 38.6, the four anchors as one system in 38.7, and the middle-read tag in 38.8. Then do the Project Checkpoint. 38.9 (what to cut) is the section most veteran staffs skip and shouldn't.
🔬 Deep Dive: This chapter stands on four earlier ones: the RPO conflict in Chapter 13, the play-action sequence in Chapter 19, the distribution idea in Chapter 25, and tempo in Chapter 32. The worked 60–70 play example of a whole system lives in Appendix L.
📺 Film Watcher: This Sunday, pick one defender, usually a safety or a weak-side linebacker, and write down what he did on every run. Then watch the next pass from the same formation. If the offense threw into the space he left, you're watching a system. If it didn't, you're watching a list.
38.1 A Play Is Half a Sentence
Every chapter so far has taught you a piece. Inside zone. Smash. The RPO. The boot. The screen. The zero bin. If you learned them well, you can draw each one against three coverages and tell me who blocks whom. That's real knowledge. It's also the knowledge Curtis Bale had.
Curtis was the most gifted play designer on Walt Dupree's staff when I was there. His first season as a coordinator, his offense was the most talked-about in the league, and his conference's leader in explosive plays. His second season, defenses played deep, he abandoned the run, his pocket quarterback took sacks, and he was fired before the next season. He carried 61 formations, and 38 of them carried one play each. By the middle of his second year he had more than 40 pass concepts in a game plan, and on 70 long-yardage downs he called zero screens. Every one of those plays was good. Not one of them was connected to another one.
That's the difference this chapter is about. A play, by itself, is half a sentence. It says, "Here's what I'm going to do." The other half of the sentence is, "...and if you stop it the obvious way, here's what that costs you." Until you've written the second half, you haven't finished the thought.
Three words
Let me give you the vocabulary, and then the idea.
A base play is a play you want to run a lot because it's cheap, it's sound against most of what the defense does, and your players can execute it in their sleep. For the Outlaws, inside zone from Eleven and outside zone from Twelve are the base runs; our quick game and a handful of dropback concepts (Smash, Mills, Stick, Drive) are the base passes.
A defensive answer is what a good defense does once your base play starts working. It almost never means "play it better." It means giving one defender a second job, or adding one more defender where the play hurts. Holloway fills the alley on inside zone. Ames sits in the slant window. The backside end chases outside zone flat down the line. The corner squats on the hitch. Each answer is correct. Each one also leaves somewhere it can't be.
A constraint play is a play built to punish one specific answer: it looks like the base play for as long as possible, then goes to the place the answer emptied. Play-action punishes the safety who fills. The RPO slant punishes the Will who fills. The boot punishes the end who chases. The corner route punishes the corner who squats. The screen punishes the rusher who pins his ears back. A constraint play doesn't have to be your best play. It has to make the defense go back to defending your best play honestly.
Many coaches and writers call this way of thinking "constraint theory." Chris B. Brown's Smart Football writing is where a lot of coaches my age first met the phrase. The idea is much older than the name. Lombardi's power sweep had its companions. Every wing-T coach in Ohio could tell you which play came after the buck sweep and why.
🚪 Threshold concept: An offense is a system of constraints. Every play is the answer to the defense's answer to another play.
Before this clicks, you think of a playbook as a menu: a collection of good plays, and the play caller's job is to pick the best one for the down and distance. You judge a play by its own average. You add plays when you see a cool one on film. When a play stops working, you replace it.
After it clicks, you can't look at a play alone anymore. You see every play as a link: it exists because the defense did something to another play, and it changes what the defense can do to the next one. You judge a play by what it does for its family. You add a play only when you can name the defensive answer it punishes and the base play it protects. When a play stops working, your first question isn't "what's wrong with it?" but "what answer did they find, and which of my plays was supposed to punish that answer?"
The tell that you've crossed over: when someone asks you, "What's your best play?" your honest answer is, "Best against what?"
Why it works: the defense has eleven people and a rulebook
🔍 Why this works: A defense is a finite set of people with assigned jobs. Every time it stops one of your plays by adding a job to a defender, that defender has less of himself for his old job. He can fill the alley or defend the deep quarter behind him, but he can't do both on the same snap at full speed. The offense's structural advantage, the one this whole book is built on, is that it knows where the ball is going and the defense doesn't. A constraint system weaponizes that advantage: if the first second of two plays is identical, the defender has to commit before he knows which one he's facing. Whatever he commits to, a play exists that goes somewhere else. That's not trickery. It's arithmetic plus time.
Here's the loop in its simplest form. Read it top to bottom, then notice that the bottom feeds the top.
THE CONSTRAINT LOOP
+----------------------------+
| 1. BASE PLAY WORKS |
| Inside zone: +5, +7 |
+-------------+--------------+
|
v
+----------------------------+
| 2. DEFENSE ANSWERS |
| one defender takes a |
| second job: Holloway |
| fills the alley early |
+-------------+--------------+
|
v
+----------------------------+
| 3. CONSTRAINT PUNISHES |
| the same look, the space |
| he left: Inside PA, |
| Z Post, +35 |
+-------------+--------------+
|
v
+----------------------------+
| 4. DEFENSE RETREATS |
| Holloway hesitates |
+-------------+--------------+
|
+----> back to 1: Inside zone, 4-yard TD
📊 Diagram (described): Four boxes stacked vertically with arrows running down. Box one: the base play works (inside zone gains five, then seven). Box two: the defense answers by giving one defender a second job (the strong safety fills the alley early). Box three: a constraint play from the same look punishes that job (play-action, post behind him, 35 yards). Box four: the defense retreats (the safety hesitates). An arrow from box four loops back to box one, where the base run scores a four-yard touchdown.
That's Anchor D, the four-play sequence that tied Game 1, drawn as a loop instead of a list. Notice the last arrow. The constraint isn't the end of the story. Its real job was to make the base play work again, and it did: on the fourth snap Holloway took one step and stopped, and Ruffin walked in from the four.
💡 Coach's point: A constraint play that hits and never sends you back to the base play has wasted half its value. Big play, then same look, base play. That's the receipt.
🔄 Check your understanding: A coordinator says, "Our screen game averaged only 4 yards a call against four-man rushes, so we're cutting it." Using the vocabulary of this section, what question should he ask before he cuts it?
Answer
He should ask what defensive answer the screen was built to punish and what happened to the base plays it protects. If the screen's job was to slow down rushers who pin their ears back on long downs, then its value includes every dropback where the edge rushers had to hesitate. Judging it by its own yards per call grades the constraint without its family. The Outlaws' own numbers show why: the screens gained about 8.4 yards per call against five or more rushers and 4.9 against four. Those are different jobs. A screen called into a four-man rush is a changeup, not a blitz-beater, and it should be judged that way.
38.2 The Engine Room: Base Runs and the Plays That Protect Them
Start where our offense starts: the run. Not because we're a run-first team (by the numbers we're not), but because almost every constraint in our system is a run's second half. In the season you've followed through this book, about 28 percent of our dropbacks were play-action and about 15 percent of our plays carried an RPO tag. Every one of those calls borrowed its first second from a run.
Each base run, the defender who stops it, and the play that punishes him
Our three run families are inside zone (with duo, which plays like it from the back's point of view), outside zone, and the gap runs, power and counter. Each one gets stopped in a predictable way by a good defense, and each predictable answer has a predictable punishment. That predictability is the point. You want to know, in July, which play you'll call when the safety starts filling in October.
| Base run | The defense's best answer | The defender who carries it | Our constraint | Where it lives |
|---|---|---|---|---|
| Inside (inside zone, Eleven) | The field safety triggers into the alley | Strong/alley safety | Inside PA, Z Post (or Y Dig) | Ch 19 |
| Inside | The Will fills fast through the backside A/B gap | Weak-side linebacker | Inside Slant (H slant into his window) | Ch 13 |
| Inside | An eighth defender walks into the box | The overhang safety or nickel | Inside Glance; the quick game ("eight, take the throw") | Ch 9, 13 |
| Inside | The backside end crashes the dive | Backside end | Inside Read (QB keeps) | Ch 12 |
| Inside | Scrape exchange (end takes the dive, LB scrapes to the QB) | The scraping LB | Arc (Harlan arcs to him) | Ch 12 |
| Wide (outside zone, Twelve) | The backside end chases flat down the line | Backside end | Wide PA Boot / Naked | Ch 10, 29 |
| Wide | Force, flatten, chase: an edge that won't be reached | Play-side edge and force player | Slice, the cutback, or Counter away | Ch 10, 11 |
| Power / Counter | Linebackers key the puller and fly to the hole | Play-side LBs | Power PA; Counter Y; Trap for a penetrator | Ch 11 |
| Any run with jet motion | The backside end chases the jet | Backside end | The reverse (Week 13); the jet itself once he looks back | Ch 33 |
| Any run on long downs | The edges rush upfield, the box goes light | Edge rushers | Draw; Slip screen | Ch 20 |
Read the table by its middle column. Every answer is a defender doing his job. Holloway's alley fill is the right job for a Quarters safety. Ames filling his gap is the right job for a Will. The chasing end is the right job on outside zone, because if he doesn't chase, the cutback kills him. Constraint plays don't punish defenders for mistakes. They punish defenders for being right one snap too long. That's why they work against good defenses.
⚠️ Common mistake: the constraint with no parent. A staff installs a beautiful boot-pass concept and calls it in the first quarter, before outside zone has shown the defense a single snap. The backside end hasn't chased anything yet, so he stays home and the boot runs into him. Result: a scramble out of bounds for a one-yard loss, or worse, a sack. Our own Week 2 version: first-and-10 at our 17, Wide PA Boot Left, and the Stallions' end, Tobias Grange, stayed home. Incomplete. Fix: Teo's "chases twice" rule. We call the boot when the backside end has chased the zone action twice on the drive, and not before. Marisol's boot table makes the point with three numbers: about 10 yards per call when he'd chased twice, about 4.5 when he'd chased once, under 2 when he hadn't.
One picture, five questions
Here is the single most important diagram in our offense. It's not a play. It's a picture: the first look of Anchor A and Anchor D, Eleven personnel, Gun Doubles Right, Harlan attached on the right, Ruffin offset to Vance's left, against Sacramento's two-high nickel "Home" look.
PRE-SNAP: Eleven, Gun Doubles Right -- vs. Nickel "Home" (two-high)
F (Royce) $ (Holloway, 10-11)
C (Brennan) * (Vo) W (Ames) M (Bell) C (Simmons)
E N T E
X (Sutton) H (Powers) O O O O O Y (Harlan) Z (Gaines)
R Q
📊 Diagram (described): The Outlaws line up two-by-two. On the left: Sutton wide and Powers in the slot. On the right: Harlan attached to the right tackle and Gaines wide. Vance is in the shotgun with Ruffin offset to his left. Sacramento shows two deep safeties, Royce on the left and Holloway on the right at ten to eleven yards; Vo, the nickel, over Powers; Ames, the Will, on the left side behind the line; Bell, the Mike, to the right of center; Rollins as the left defensive end, the nose and the 3-technique inside, and Weller outside the right tackle; Brennan and Simmons at the corners.
From that one picture we could run, in the season this book follows, at least seven different plays: Inside, Inside Slant, Inside Glance, Inside Read, Inside Read Bubble, Inside PA, and Duo Pop from its Twelve cousin. Every one of them shares the same first half-second. The linemen step right. Ruffin crosses toward the mesh. Vance extends the ball. And five defenders each have to answer a question before they can know which play it is.
POST-SNAP: the same first second -- every defender's choice has a play
F $ (1) fill or stay?
/
C * (4) sink W (2) fill M
or squeeze? or sit?
E (3) N T E
X H O-->O-->O-->O-->O--> Y ~~arc~~> Z
| \ R==Q mesh, zone right |
v go \ slant (5) Vance: keep, give, or throw | post
📊 Diagram (described): At the snap the whole line steps right on inside zone and Ruffin meets Vance at the mesh. Harlan arc-releases toward the alley. Five defenders carry numbered questions. One: does Holloway, the right safety, fill the alley or stay deep? Two: does Ames, the Will, fill his run gap or sit in the slant window? Three: does Rollins, the backside end, crash the dive or stay home on the quarterback? Four: does Vo, the nickel, sink with Powers or squeeze inside? Five: Vance, at the mesh, decides whether to keep, give, or throw. Sutton runs a clear-out go on the left; Powers runs the slant; Gaines on the right runs the post when play-action is called.
Now the reading key. This is the table I'd hand a new quarterback, and it's the table I'd hand a new defensive coordinator if I wanted to scare him.
| # | Defender | His run job | His other job | If he plays the run → | If he plays the other job → |
|---|---|---|---|---|---|
| 1 | Holloway ($) | Alley fill | Deep quarter over #2 | Inside PA, Z Post behind him | Inside: one fewer man in the box |
| 2 | Ames (W) | Backside A/B gap | The slant window | Inside Slant: throw where he was | Inside: hand off where he was |
| 3 | Rollins (backside E) | Squeeze the dive | Contain the QB | Inside Read: Vance keeps | Inside Read: Ruffin gets the dive |
| 4 | Vo (*) | Squeeze into the box | Cover Powers outside | Bubble/Now: throw it outside, now | Inside: he's not in the box |
| 5 | Vance (Q) | Ride the mesh | Throw or keep | He's the one making the defense wrong |
That table is the threshold concept written in rows. Five defenders, ten possible choices, and every choice has a play attached. The defense's only real escape is to add a sixth defender to the question, and adding a sixth defender empties somewhere else on the field. That's where the passing game comes in.
🛡️ The defense's view: Russ Kimbrough's answer to this picture, which I watched him build over three games and heard him explain in the spring, was not to win any one of those questions. It was to change who answers them, and when: a more patient Will, a rotator with a different job, a big-nickel safety told not to be interested in the run. Section 38.7 shows how each change moved the conflict instead of removing it. A great defense doesn't beat your system by winning one conflict. It beats it by refusing to let one man be the conflict.
🔄 Check your understanding: On the second snap of Game 1 (first-and-10 at the Austin 49, the same Inside Slant call), Ames sat in the slant window instead of filling. Why was that the wrong choice for him, even though it took the slant away?
Answer
Because Ames had two jobs on one side of the field going in opposite directions: a run fit in the backside gap and a pass window behind him. When he sat in the window, he left his run gap, and with Harlan arcing to the alley safety and Weller playing a 5-technique, the box was five blockers on five defenders. Vance read Ames sitting, handed off, and Ruffin gained seven. Sitting took away the slant and gave away the run. There was no right answer for him on that snap; that's the whole design.
38.3 The Passing Core and Its Own Constraints
The run game's constraints are passes. The passing game's constraints are, mostly, runs and screens. That symmetry is not an accident. A defense that sells out to stop the pass has to do it by getting deeper, getting more people into coverage, or getting to the quarterback faster. Each of those has a cost, and the cost is always paid somewhere a pass can't reach.
How defenses take the passing game away
When our base passes, the quick game and our dropback concepts read by defender, start hurting a defense, the defense has five main answers. Here they are, with what each answer gives up and what in our system cashes it.
| The defense's answer to our passing game | What it gives up | Our constraint | Example from this season |
|---|---|---|---|
| Pressure (five or six rushers) | Coverage defenders; the deep middle in Cover 0 | Hot throws, sight adjusts, screens, the zero bin | Game 1, 3rd-and-8 at the AUS 48: zero bin deep over, 52-yard TD |
| Rush upfield on long downs (ears pinned back) | The space behind the rushers; a light box | Draw, Slip, Middle screen | Week 10, 3rd-and-12 at the POR 28: Middle screen, Ruffin +9 |
| Press and squat on the quick game | The space behind the corner | Double moves: Sluggo, Stop-go; fades | Game 2, Q4: Gaines slant-and-go, 44 yards |
| Man coverage to kill the zone reads | Every defender's eyes; the spare player | Mesh, Shallow, stacks and Lace; the QB scramble | Week 8, 3rd-and-4: Mesh, Powers +11 |
| Drop eight, play everything in front (Tampa 2, Cover 2) | The run game; the hole between the corner and the safety | Smash; runs against a light box; screens | Week 10, Anchor B: Smash, 31-yard TD |
There's a sixth answer, and it's the one that beat us. Disguise the shell, then rotate after the snap. It doesn't give up a spot on the field. It gives up a time: the defense has to move after the ball is snapped, and every rotation leaves a window open for a moment while it closes another. The only constraint for a post-snap answer is a post-snap read. That's 38.8.
Anchor B: a concept that's a system by itself
Smash against Cover 2, the Portland touchdown, is the cleanest example in this book of a single concept working like a little system. It doesn't need a sequence of calls, because it puts the sequence inside one play. Here it is again, the way we drew it for the game: Week 10 at Portland, second quarter, 2nd-and-6 at the Portland 31, tied 7–7, "Eleven, Gun Doubles Right, 60 Left, Smash," against Leon Marchetti's disciplined Cover 2.
PRE-SNAP: Week 10 at Portland, Q2 -- 2nd-and-6, POR 31, tied 7-7
"Eleven, Gun Doubles Right, 60 Left, Smash" vs. Cover 2
F (Price, #30) $
W M
C (Kiel, #23) * C
E T N E
X (Sutton) H (Powers) O O O O O Y (Harlan) Z (Gaines)
Q R
📊 Diagram (described): Outlaws in two-by-two: Sutton wide left, Powers in the left slot, Harlan attached on the right, Gaines wide right, Vance in the gun with Ruffin to his right (the slide is to the left, so the back is on the right). Portland shows two deep safeties, Darnell Price on the left half; a cornerback, Josh Kiel, squatting at about five yards outside Sutton; a nickel over Powers; two linebackers; a four-man front.
POST-SNAP: Kiel has the flat -- hitch or corner? He can't take both
~18 yds corner <.............................
^ \ .
| \ F (Price) held on the hash
~12 yds | H stem ^ by Harlan's seam
| / \ ^
~6 yds X hitch <-- C (Kiel) squats \ | Y seam
^ \ |
| W M \ | Z dig <--+
X H Q Y Z
📊 Diagram (described): On the left, Sutton runs a hitch and stops at six yards; Powers stems straight up to about twelve yards and breaks to the corner of the end zone at about eighteen. Kiel, the Cover 2 corner with the flat, squats on the hitch. Price, the left-half safety, is held toward the hash by Harlan's seam up the middle. That leaves a hole between Kiel and Price on the left sideline, and the ball goes over Kiel to Powers on the corner route for a 31-yard touchdown. On the right, Gaines runs a dig at twelve yards behind the play.
Kiel squatted on the hitch, because squatting on #1 is his flat rule. Vance threw the corner over him and under Price. Powers, 31 yards, touchdown, 14–7. In the third quarter, on the next Smash call, Kiel sank under the corner route. Hitch to Sutton, plus eight. The corner must choose; both choices lose. That's the RPO idea, Anchor A's idea, played out in a pass concept: one defender, two jobs, read him and throw where he isn't.
But here's what makes it part of a system and not just a good play. Portland's answer to Smash wasn't to coach Kiel better. Teo's chart had Kiel squatting on 17 of 21 smash and hitch looks; that was who he was. Their answer was structural. On third-and-medium, Marchetti went to Tampa 2 on 11 of 16 snaps, running his Mike down the middle to take away the seam that held Price. On third-and-long he lived in Cover 2 and Tampa 2 and dared us to throw short of the sticks. So the rest of our Portland plan was the punishment for those answers: Levels with a vertical (our rule is "no levels without a vertical"), Drive to make the hook players choose, and, at 4:58 of the fourth quarter on 3rd-and-12 at the Portland 28, up 24–13, "Eleven, Gun Trips Right, Middle." Teo's chart had Portland in Cover 2 or Tampa 2 on 14 of its last 17 third-and-10-plus snaps. Ruffin gained nine to the Portland 19, and Evan Dorsey kicked the 36-yard field goal that made it 27–13.
Look at what each of those calls was punishing. A Mike who runs deep down the middle is no longer in the hook zone, so the Drive dig at 12 yards (2nd-and-8 at the Portland 42 in the third quarter, Gaines, plus 15) lived in the grass he left. And on third-and-12 the Middle screen went to the one place the whole defense, Mike included, had just run away from. A good answer to an adjustment isn't "Smash again." It's the play that lives where the adjusting defender used to stand.
🎬 Film room: When you study a concept like Smash on film, don't stop at the snap where it worked. Go find the next time the same offense faced the same defense in the same situation. Did the defense change the corner, the safety, or the coverage? Did the offense have a call ready for the change? A coordinator's system is visible in the second and third calls against an adjustment, not in the first one.
🔄 Check your understanding: Why is Smash a weak call against Cover 3, even though it's a "system in one play" against Cover 2?
Answer
Because Smash attacks the Cover 2 corner, who has the flat and has to choose between the hitch and the corner route. In Cover 3 the corner bails to the deep outside third, so he's already standing where the corner route is going, and the flat belongs to a different defender. The one-defender, two-jobs conflict that makes Smash work doesn't exist. Concepts are coverage-specific: you choose them by the defender they attack, not by how good they looked last week.
38.4 The Floor Under Everything: Protection, and the Ball as a Blocker
Everything above assumes your quarterback is standing when the defense makes its choice. Pressure is the one defensive answer that attacks the whole system at once. It doesn't take away a play; it takes away time, and every play needs some. A six-man pressure shortens every question on the field: the RPO read has to happen faster, the concept has less time to develop, and the play-action fake turns into a liability, because the quarterback has his back to the rush.
That's why one of the five Outlaws non-negotiables is a protection rule, not a play: every protection has an answer for one more rusher than it blocks. It's the floor of the system. If that floor has a hole, it doesn't matter how good the web above it is.
Protection is a menu of trades
Every protection we own is a trade between blockers and routes. More blockers buy more time and fewer places to throw. Fewer blockers buy more places to throw and require the ball to come out on time. A system needs the whole menu, because the defense chooses how many people to send, and you want the right trade ready for each choice.
Our menu runs from 50 (five blockers, empty, the ball as the sixth) through 60 Left/Right (our base half-slide: one rusher can always be left over, and he's the hot), 70 (the tight end stays) and 80 Max (eight blockers, two or three routes, the zero bin's deep over), to Boot/Naked (move the launch point, only after the end has chased twice) and the play-action protections that take a run's name. Each extra blocker is a route you didn't run, and each route you didn't run is a defender who gets to help somewhere else. Chapter 22 owns the details.
The arithmetic is the same at every level. Count the possible rushers. Count your blockers. If the rushers win the count, decide before the snap who's free, and make sure the ball beats him there. That's why in our offense the quarterback is part of the protection: when rushers outnumber blockers, the hot throw or the sight adjust is the extra blocker.
Anchor C: the floor, patched
Here's what happens when the floor has a hole, and what a system does about it.
In Game 2, Week 15 at Sacramento, on 3rd-and-7 at our 38 with 6:51 left in the third quarter and the score tied 13–13, the Sentinels showed us a double-A mug from a one-high look. Our rule then was "six-man, back on the Mike," with the hot to Powers in the field slot. At the snap, the corner came, a 3-technique dropped into Powers's window, the end walked late enough to move our back, and Bell ran free through the A-gap. Darius looked hot, found Rashad Coleman standing in the window, and was sacked for nine. Russ told me later they call that pressure Mug Cat Fire. I didn't know the name in December. I knew the picture.
Wednesday's QB room after that game had a line on the whiteboard I still think about: 6 blockers + 1 ball = 7. Russ made it 4 + 0. He'd beaten the arithmetic by taking away the ball as a blocker. The new rule came out of the Tuesday meeting after Week 16, and it was Teo's proposal: center on the Mike, back on the other A-gap backer, right tackle widens to the corner, and Gaines sight-adjusts to a six-yard hitch if his corner blitzes.
Then came Anchor C: the divisional round, Austin's opening drive of the third quarter, 3rd-and-7 at our 41, Sentinels up 10–7, about 11:30 left in the quarter. "Eleven, Gun Doubles Left, Y Off, 60 Right, Mills." Same formation as the Game 2 sack: Harlan a step off the line outside the left tackle, Gaines on the line wide right.
PRE-SNAP: Divisional round, Q3 -- 3rd-and-7, AUS 41, SAC 10-7
"Eleven, Gun Doubles Left, Y Off, 60 Right, Mills"
F (Royce) $ (Holloway, creeping)
C (Brennan) * (Vo) C (Simmons)
S E N W M T
X Y LT LG C RG RT H Z
R Q
mug rule: C on M (Bell) | R on W (Ames) | RT widens to the corner
📊 Diagram (described): Sutton wide left, Harlan a step off the line about four yards outside the left tackle, Powers in the right slot, Gaines on the line wide right, Ruffin on Vance's left. Sacramento shows Royce deep and Holloway creeping down on the right; Weller walked out to the boundary beside Rollins; the nose on the left guard; Ames and Bell in the two A-gaps; Coleman outside the right guard; Vo over the right side; Simmons over Gaines and Brennan over Sutton. The note gives the new rule: center on Bell, Ruffin on Ames, right tackle widens to the corner.
POST-SNAP: the patched floor -- five rush, five blocked, ball is #6
[DEEP 1/3] C [DEEP 1/3] F [DEEP 1/3] $
S drops T drops (into H's window)
E N W M C ==> blitz
X X X X X
LT LG R C RG (helps) RT (widened)
Z hitch at 6
Q ...................................> +10
📊 Diagram (described): Five Sentinels rush: Rollins, the nose, Ames, Bell, and Simmons off the right edge. Weller and Coleman drop, exactly as in Week 15. This time each rusher has a blocker: the left tackle on Rollins, the left guard on the nose, Ruffin stepping into the left A-gap on Ames, the center on Bell, and the right tackle, widened at the snap, on Simmons. The right guard helps inside. Gaines sees his corner leave, stops at six yards, and the ball arrives at about 1.1 seconds; he turns it into a ten-yard gain and a first down at the Sacramento 49.
Zero free rushers. Gaines, plus ten, first down. The mug corner fire was the Sentinels' only call of that pressure all game.
Now look at Anchor C the way this chapter wants you to. It isn't "the protection call that worked." It's the system growing an arrow. In Week 15, the defense found an answer (a pressure that removes the hot) that nothing on our map punished. The arrow was empty, and it cost us nine yards and a punt in a game we lost by three. In the five weeks between that sack and the divisional round, we wrote a play at the end of that arrow and repped it four ways on Thursday. When the answer came back, the play was there.
💡 Coach's point: Your protection rules are part of your play-calling system, not a separate department. Every pressure the defense owns is a defensive answer, and every one needs an arrow: a blocker, a hot, a sight adjust, or a screen.
🔄 Check your understanding: In the Game 2 sack, Austin had six blockers and a hot throw. Why did the arithmetic fail, and which part of the Game 3 rule fixed it?
Answer
The arithmetic assumed the hot throw would be available as a seventh "blocker." The pressure dropped a 3-technique into the hot window, which took the ball away as a blocker, and the late walk moved the back off the man he was supposed to block, so Bell came free. Six blockers plus a ball turned into fewer real blockers and no ball. The Game 3 rule gave every interior rusher a blocker by assignment, not by movement: center on the Mike, back on the other A-gap linebacker, right tackle widened to the corner. With all five rushers blocked, the dropping lineman in the hot window didn't matter, and Gaines's sight adjust beat the vacated corner.
38.5 Tempo: Not a Play, a Multiplier
You'll notice tempo isn't a row in the run tables or the pass tables. That's on purpose. Tempo doesn't create a conflict. It changes how long the defense has to fix one.
Think about what a defensive answer actually is. When Holloway starts filling, someone on Russ's staff has to see it, say it, and decide whether to change it. When our RPO is beating the Will, the fix is a call: a different fit, a rotation, a sit. When our Eleven runs are beating his nickel, the fix is a substitution. Every one of those fixes is a decision that takes time: a substitution, a sideline conversation, a new call relayed through the green-dot linebacker.
Tempo taxes the time. Chapter 32 named the four taxes: personnel (the defense can't substitute if you don't), the call (it has less time to choose and disguise), the legs (tired defenders fit late), and the disguise (a defense that gets lined up late can't fake much). So tempo is a multiplier on whatever is already winning the conversation. If your base play is ahead, tempo keeps the defense from finding its answer. If your constraint just hit, tempo keeps the defense from correcting for it before the base play comes back.
Here's Anchor D as a conversation, with the tempo marked.
THE FOUR-PLAY SEQUENCE AS A CONVERSATION (Game 1, Q4, SAC 27-20)
snap down & spot we showed Holloway learned result
---- ------------ -------------- ------------------- ----------
1 1-10 AUS 49 Inside right "it's a run: fill" Ruffin +5
2 2-5 SAC 46 Inside right "fill harder" Ruffin +7
3 1-10 SAC 39 Inside PA, bites on the fake; Gaines +35
Z Post post behind him
...... a 15-yard gain: FAST (no huddle, no substitution) ......
4 1-G SAC 4 Inside right "one step, then Ruffin TD
stop" 27-27
📊 Diagram (described): A four-row table of the tying drive's last four snaps. Snap one, first and ten at the Austin 49: inside zone right, Holloway learns it's a run and fills, Ruffin gains five. Snap two, second and five at the Sacramento 46: inside zone again, he fills harder, Ruffin gains seven. Snap three, first and ten at the Sacramento 39: play-action off the same look, Holloway bites and the post behind him gains 35 to the four. A dotted line marks that the gain triggered fast tempo. Snap four, first and goal at the four: inside zone, Holloway takes one step and stops, Ruffin scores, and the game is tied 27–27.
After the 35-yard post, we went fast. A gain of fifteen or more is the first trigger on our fast-mode card, and that snap is the example Chapter 32 uses for it. Think about what Russ needed in that moment: a sideline conversation with his safety about the play-action, maybe a different call for the goal-line snap. Fast mode taxed exactly that. Holloway had to defend the next snap with the lesson from the last one still fresh and nobody's correction in his ear. He split the difference, one step and stop, which is the worst place to be against a run.
🔍 Why this works: Tempo is valuable in proportion to the conflict you've already created. Going fast with a play that doesn't put anybody in conflict just runs a plain play sooner. Going fast right after a constraint has hit freezes the defense in its most confused state: the defender has just learned his answer was wrong and hasn't been told what to do instead. That's why our fast calls are almost all post-snap read plays. A post-snap read carries its own constraint inside it, so it doesn't need a sideline decision either.
That last point is why the Fast row on our call sheet looks the way it does. Every call on it has to "decide by itself": Inside Slant with a Now tag, Inside Read Bubble, Wide Now, Inside PA (only after an Inside run has hit on the series), Slants, Stick. No Twelve, no shifts, no long motions, nothing that needs a Kill or a Check. Tempo freezes our adjustments too. Motion that asks a question needs time for the answer, and we're a motion team, not a tempo team. So we go fast as a change-up, and only with plays that carry their own answers.
Tempo multiplies your mistakes, too
The Week 12 game against St. Louis taught me the other half. Up 20–16 with 10:40 left, Vance hit Powers on Inside Glance for 19 to the Stallions' 17. Big gain, so: "Fast, again." Three snaps later we kicked a 32-yard field goal, up 23–16, at about 9:50. Wade's point on the sideline was simple: we were the team that wanted the game to be shorter, and fast mode had just given St. Louis about forty seconds back. Our fast card now has a veto at the top: no Fast with a fourth-quarter lead.
And there's tempo the other way: the Slow gear of the four-minute offense, where we huddle long and snap with two or three seconds left. That's a constraint too, but on the clock rather than on a defender. With a lead late, the defense's answer to everything is "get the ball back." Slow is the play that punishes that answer by spending its time.
⚠️ Common mistake: tempo as identity without a system underneath. A staff decides to "play fast" every snap, installs a huge no-huddle menu, and calls plays at race tempo that have no conflict built in. Result: three-and-outs in 50 seconds, a tired defense of their own, and a defense on the other side that only has to be sound, not clever. Fix: tempo goes on top of a constraint system, not in place of one. Ask the fast test of every call: does it decide by itself, from a picture you've already shown, with an answer for one more rusher?
🔄 Check your understanding: Why did the Outlaws go fast after the 35-yard post on the tying drive but deliberately not go fast after the 19-yard Glance in Week 12?
Answer
In Game 1, Austin was trailing 27–20 with time to spare, and the 35-yard gain had just put Holloway in his most confused state; going fast kept Sacramento from correcting him before the base run came back. In Week 12, Austin was ahead in the fourth quarter. The clock was worth more than the defense's confusion, and fast mode gave St. Louis about forty seconds back. Same trigger, different situation. The score and the clock outrank a tired defense; that's why the veto exists.
38.6 The System Map: Every Answer Has a Price
Now put it together. This is the page Teo needed for Tab 2 of his binder, and it's the page you'll build in this chapter's Project Checkpoint. It answers one question per row: when the defense does this, which part of our system makes it pay?
The defensive-answer table
Teo's scouting page for every opponent starts from these five big answers, the ones every defense in the league uses in some form. Under them go the specific answers that live in our own tape; four of ours are in the table too.
| The defense's answer | What it's trying to stop | What it gives up | The part of our system that punishes it | Example |
|---|---|---|---|---|
| Load the box (eight, a safety walked down) | The run | A deep defender; the space the eighth man left | RPOs (Glance, Slant), play-action, the quick game: "eight, take the throw" | Week 12, Q4: Inside Glance, Powers +19 |
| Rotate a safety down (pre-snap single-high) | The run and the short middle | The deep half he left | Play-action shots behind him; the boot is a shot against single-high | Week 12: Wide PA Boot, Harlan +18 |
| Pressure (five or more) | The pass; the quarterback's time | Coverage players; often the deep middle | Hots, sight adjusts, quick game, screens, the zero bin | Week 8: "Blue five" Slants vs. zero, Powers +14 |
| Two-high, light box (six) | Explosive passes | A run defender | The run: Inside, Duo, Inside Read; draws; "the run for credibility" | Game 1, Q2: Inside, Ruffin 45-yard TD |
| Man coverage | The zone reads and hi-lo concepts | Every defender's eyes; the spare player | Mesh, Shallow, stacks and Lace, double moves, the scramble | Week 8: Mesh, Powers +11 |
| The Will fills the zone flow | Inside zone's backside | The slant window | Inside Slant | Game 1, snap 1: Powers +17 |
| The backside end chases | Wide's cutback | Containment on the backside | Wide PA Boot, Naked, Slice; the reverse off jet | Week 13: reverse, Gaines +31 |
| The corner's eyes on #2 (Quarters) | The vertical by #2 | His man's out-cut | The bender | Game 3, Q2: Powers +14 |
| Disguise, then rotate after the snap | Pre-snap reads | Nothing on the field; a moment in time | All Go (reads after the snap); the middle-read tag | Game 3, 2:41: the empty arrow |
Two things about this table. First, the "Example" column isn't decoration. Every row has a snap. If you can't find a snap where your punishment actually punished, the row is a hope, not a rule. Second, the last row is the one that cost us a season, and it's the one where the "gives up" column says nothing on the field. That's what makes post-snap disguise the hardest answer in football: the only thing it gives up is time, so the only thing that punishes it is a read that happens in time.
The map itself
The table is how you study it. The map is how you see it. Ours hangs on the quality control office wall. Here's the spring version, compressed to fit this page.
THE OUTLAWS SYSTEM MAP -- spring version (each row: answer -> price)
BASE PLAY THE DEFENSE'S ANSWER PUNISHED BY
------------ -------------------------- --------------------------
INSIDE (11) alley safety fills -------> Inside PA, Z Post
Will fills ---------------> Inside Slant
8th man in the box -------> Inside Glance / quick
backside end crashes -----> Inside Read (keep)
WIDE (12) backside end chases ------> Wide PA Boot / Naked
force, flatten, chase ----> Slice / Counter
POWER, CTR LBs key the puller -------> Power PA / Counter Y
JET end chases the jet -------> Reverse (once)
QUICK GAME press, squat -------------> Sluggo / Stop-go
DROPBACK 5-6 rushers --------------> hots / screens / zero bin
rush upfield, light box --> Draw / Slip / Middle
man everywhere -----------> Mesh / Lace / scramble
Cover 2 corner squats ----> Smash
Quarters eyes on #2 ------> Bender / Scissors
rotate AFTER the snap ----> MIDDLE-READ TAG (new)
======================================================================
TEMPO Fast freezes an answer that's losing; Slow spends the clock
UPKEEP every arrow gets a "since" date; zeros get killed
======================================================================
When a price gets paid, the BASE PLAY is cheap again. Start over.
📊 Diagram (described): A three-column map. The left column lists our base plays: Inside from Eleven, Wide from Twelve, Power and Counter, the jet, the quick game, and the dropback game. The middle column lists each defensive answer to those plays, such as the alley safety filling, the Will filling, an eighth man in the box, the backside end chasing, press coverage, five or six rushers, man coverage, the Cover 2 corner squatting, Quarters corners watching the number two receiver, and rotation after the snap. Arrows run from each answer to the play that punishes it, such as Inside PA, Inside Slant, the boot, the reverse, double moves, screens and the zero bin, Mesh, Smash, the bender, and, last, the new middle-read tag. Two bands across the bottom add tempo as a multiplier and self-scouting as upkeep. The final line says that when a punishment lands, the base play becomes cheap again.
My rule for the map, and the only rule: every arrow needs a play at both ends. An answer with no play after it is a hole. A play with no answer before it is decoration. Both get circled in red.
📋 Try it — Find the conflict: The St. Louis Stallions (DC Frank Ostrowski) play single-high Cover 3 with an eight-man box. Against our Eleven, Gun Doubles Right, the strong safety walks down to about seven yards outside Harlan, the free safety, Wes Carrow, sits in the deep middle at about 13, and the Will is stacked over the guard. On inside zone, that walked-down safety fills the cutback from outside in, and he's been right all game. Name (a) the conflict defender, (b) his two jobs, and (c) the play in our system that punishes him, with where the ball goes.
Answer
(a) The walked-down strong safety, the eighth man in the box. (b) His run job is to fill the cutback or the alley on inside zone; his pass job in Cover 3 is the curl/flat or the seam area on his side as the eighth man rotates back into coverage. He can't do both on the same snap at full speed. (c) The RPO Inside Glance: the line blocks inside zone, Vance reads the safety at the mesh, and if he triggers on the run, Powers's skinny post goes into the space he left, under Carrow. That's the Week 12 snap: first-and-10 at the Stallions' 36 with 10:40 left, the safety filled the cutback, and Powers caught the Glance for 19. If the safety stays home, the box is back to seven and Ruffin gets the handoff: "seven or fewer, run it." A strong answer also names the follow-up: once the safety stops filling, the plain run is cheap again.
38.7 The Four Anchors as One System
This book has leaned on four plays more than any others. You've seen each one in the chapter where it lives. Here they are side by side, because seen together they're one idea at four different speeds.
| Anchor | Game and situation | The conflict defender | His two jobs | The time scale | What made him wrong |
|---|---|---|---|---|---|
| A: "The RPO the linebacker can't win" | Game 1, Austin's first two snaps (1st-and-10 at the AUS 32, then AUS 49), 0–0 | Jordan Ames (#45), the Will | Backside run fit / slant window | Half a second (at the mesh) | Filled → slant, Powers +17; sat → handoff, Ruffin +7 |
| B: "Smash vs. Cover 2" | Week 10 at Portland, Q2, 2nd-and-6 at the POR 31, 7–7 | Josh Kiel (#23), the Cover 2 corner | The flat (hitch) / the hole under the safety (corner) | About two seconds (the drop and the break) | Squatted → corner, Powers 31-yard TD; later sank → hitch, Sutton +8 |
| D: "The four-play sequence" | Game 1, Q4, SAC 27–20; AUS 49 to the SAC 4 | Tre Holloway (#31), the field safety | The alley / the deep quarter behind him | Four snaps (about three minutes) | Filled twice → post behind him, +35; hesitated → run, TD |
| C: "The third-and-seven decision" | Game 3, Q3, 3rd-and-7 at the AUS 41, SAC 10–7 | The pressure itself (the dropper in the hot window) | Rush / drop into the hot window | About five weeks (Week 15 to the divisional round) | Every rusher blocked; Gaines sight hitch, +10 |
Read the "time scale" column and the whole book comes into focus. The RPO creates its conflict in half a second, at the mesh. Smash creates it in about two seconds, in coverage. The play-action sequence creates it over four snaps, by teaching a defender a lesson and then cashing it. And the protection answer was built across a month of the season, by taking a defensive answer that had beaten us and writing a play at the end of its arrow. Same idea, four clocks: make one defender, or one defensive rule, responsible for two things, and go wherever he isn't.
How the anchors talked to each other
Here's what the season taught me that no single anchor could. The anchors weren't independent. Each one changed what the defense did to the others.
After Anchor A, Russ's staff fixed Ames. By the divisional round his eyes were better and his trigger was slower, and our Inside Slant on the Q3 opening drive gained a yard. But a patient Will is a Will who's late to his run fit, which is part of why Inside could still gain yards on that defense. And because we'd shown the slant, Russ had to spend a player or a rotation on the slant window more often, which is part of why his rotation existed at all.
After Anchor D, Russ changed his rotator's job. A safety who'd been burned once by play-action off the run didn't fill the same way in Week 15, and his big-nickel safety against our Twelve had one job: don't be interested in the run. That made our 12-personnel runs worse in Game 2 (Marisol's Week 15 number: 3 of 8 successful), and it's why Twelve became a boot package for the divisional round. The defense took away one conflict, and we moved the conflict to the backside end.
Anchor C existed only because of the Game 2 sack, which existed because Russ had built a pressure that took our hot throw away. We answered it. He called it once in the divisional round, and it paid us ten yards.
And then, with 2:41 left in the divisional round, Russ played the answer we hadn't priced. Two-high on third-and-long had stayed two-high 29 straight times since Sacramento's bye. On the 30th, at 3rd-and-9 at our 38, down 20–17, the strong safety rotated down at the snap and Royce spun to the deep middle. We'd called Mills, and Mills against two-high is a good call. Darius read the shell he'd been shown, threw the post at about 2.4 seconds, and Royce caught it.
There's a conversation I had with Wade about the map in the spring that I want you to hear, because it's the clearest statement of the threshold concept I know.
Wade: You've got an arrow for everything I'd do. What if I don't do anything? What if I just play Home, four-man rush, Quarters, every snap, and tackle well?
Lyle: Then we run Inside until you stop us. And Jalen gets five a carry against six in the box, and we're in second-and-five all day, and Darius throws Stick on third-and-two.
Wade: So the map only matters when I answer.
Lyle: The map only matters when you're good. A bad defense loses to the base plays. A good one answers, and every answer costs something. The map is the bill.
Wade: And Russ found the one answer that didn't have a bill.
Lyle: He found the one where the bill came due after the snap. We were paying before it.
What the exchange reveals: a constraint system doesn't need the defense to make mistakes. It needs the defense to make choices. A defense that chooses nothing gets beaten by your base plays; a defense that chooses something gets beaten by the play behind that choice. The only escape is a choice your system can't see in time. That's what the last section of the map is for.
🔄 Check your understanding: Anchor A and Anchor D both beat Sacramento in Game 1. Why did each of them make the other one harder to stop in later games?
Answer
Both attack the same picture, Eleven, Gun Doubles Right, with an inside-zone first second, but they stress different defenders: Anchor A the Will (Ames) at the mesh, Anchor D the field safety (Holloway) over four snaps. When Russ fixed Ames by making him patient and adding rotation toward the slant window, he spent safety help on the RPO, which complicates what Holloway can do against the run and play-action. When he fixed Holloway by making him less interested in the run, the box got lighter, which helps the plain run and makes the Will's fit harder. Each fix shifted weight onto another defender. That's what a system does: the defense can move the conflict, but it can't remove it.
38.8 Closing the Empty Arrow: The Middle-Read Tag
🧩 Productive struggle: You own Mills, and against two-high it's the right concept: the safety over #2 has the dig and the post, and he can't take both. Your problem is the snap where the shell lies. Design a fix with three constraints. (1) It can't take the post away against a defense that really does stay two-high. (2) It can't add time to the drop; the ball still comes out around 2.4 to 2.9 seconds. (3) It can't depend on the quarterback feeling doubt, because the snap that kills you is the one where the whole week has made him sure. Write your fix in two sentences before you read on.
If your answer moved the decision from before the snap to after it, and gave the quarterback a second throw that's ready on the same rhythm, you've built what we built. Here it is as we installed it this spring. Earlier chapters described the idea. This is the mechanism.
The tag, defined
The middle-read tag is a rule attached to a pass concept that turns its middle-of-the-field throw into a post-snap decision. It's printed on every Blue-panel third-and-long call (7 yards or more) that has a throw in the deep middle, and on those calls it can't be taken off. On any other call, I add it with one word at the end of the call: "Eyes." ("Eleven, Gun Doubles Left, 60 Right, Mills, Eyes.") We didn't use "Middle" because Middle is already our screen, and a receiver who hears "Middle" in the huddle should think of exactly one thing.
It works in four beats.
| Beat | When | What Vance does | What it replaces |
|---|---|---|---|
| 1. Name both throws | In the huddle, before the snap | Says to himself, "Open, post. Closed, dig." Every tagged concept has its two throws printed on the band | Picking the throw from the pre-snap shell |
| 2. Eyes to the grass | First step of the drop | Looks at the empty space between where the two safeties stood, not at either safety | Reading the shell, then going to a receiver |
| 3. Declare | Third step (the top of the gun drop) | "Open" if nobody has arrived in that grass at dig depth or deeper; "closed" if a safety has arrived or is arriving | A pre-snap "two-high" that nobody re-checks |
| 4. Throw on rhythm | Same time as always | Open: the concept's normal read. Closed: never the post; the closed throw; if it's covered, rule four | Rule four as the only safety net |
Rome Pettaway teaches it the way he taught the middle read from the start: two looks with one pair of eyes. The first look is at grass, not people. The second is at the defender the concept was built to stress. Notice what beat 4 says on the "open" side: the tag doesn't replace the concept's read. It decides which read is live. If the middle is open, Mills is Mills: read the safety over #2; if he jumps the dig, the post; if he sits on the post, the dig.
The same snap, with the tag
Here's the divisional-round snap redrawn with next season's call.
PRE-SNAP: 3rd-and-9, AUS 38, 2:41, SAC 20-17 -- redrawn with the tag
"Eleven, Gun Doubles Left, 60 Right, Mills, Eyes"
$ (Holloway, 13) F (Royce, 13)
C (Brennan) W M * C
E N T E
X (Sutton) Y (Harlan) O O O O O H (Powers) Z (Gaines)
R Q
on the band: OPEN = post (then read #2's safety) | CLOSED = dig
📊 Diagram (described): Harlan is attached on the left beside the left tackle with Sutton wide left; Powers is in the right slot and Gaines wide right; Ruffin is on Vance's left and stays in to block. Sacramento shows two safeties at thirteen yards, Holloway on the left and Royce on the right, with Vo over Powers and Ames and Bell behind a four-man front. The note says the band carries both throws: if the middle is open, the post (then the usual read of the safety over the number two receiver); if closed, the dig.
POST-SNAP: middle CLOSED. Step 1: eyes to the grass. Step 3: "closed."
^ X go F (Royce, spun to the middle)
| ^
| \ Z post -- NEVER into him
| \
| <------------------------- H dig at 12 -- YES, under him
| | \
| $ (Holloway, rotated down) | \
| <-- Y flat | \
X Y O O O O O H Z
R Q throw: the dig at about 2.4 s
📊 Diagram (described): At the snap Holloway rotates down on the left toward the curl and flat, and Royce spins from the right into the deep middle. Sutton runs the go up the left sideline; Harlan runs to the left flat; Gaines's post heads toward the middle, straight into Royce; Powers runs up to twelve yards and breaks left on the dig, underneath Royce. Vance's eyes went to the empty middle on his first step and found Royce arriving by his third, so the call is "closed": no post. The dig comes out on the same rhythm, at about 2.4 seconds.
On that snap, the dig would have been a first down. Vance knew it before I did. "Rule four, Coach," he said on the sideline. The free square says when the middle is in doubt, throw the ball you know. The tag's whole purpose is to stop asking him to be in doubt. It asks him to look, every time, on a schedule.
One drop, two answers
The design rule that goes with the tag is the part most staffs will skip, and it's the part that matters most: a concept can carry the tag only if its closed throw lands on the same rhythm as its open throw. If the closed throw needs another hitch step, the quarterback will hold the ball while he waits for it, and the pressure Russ didn't send will get home anyway.
| Concept | Open throw (then the concept's read) | Closed throw | If that's covered |
|---|---|---|---|
| Mills (post-dig) | Post; read the safety over #2 | Dig at 12, under the middle safety | Flat or the back (rule four) |
| Dagger (seam + deep dig) | Seam between the safeties | Dig at 15–16, under the middle safety | The back |
| Y Cross | X post | The cross, under the middle safety | H flat |
| All Go | Seams away from the middle safety's hips | The seam on the other side | The back |
All Go is on that list because it was the model. It has always asked the middle question after the snap; that's how it went for 19 yards to Harlan against the rotation on the first snap of our next-to-last drive in the divisional round, earlier in the fourth quarter. Mills asked the question before the snap. A concept with no closed throw on rhythm doesn't get the tag. It leaves the third-and-long box and goes to early downs, where an incompletion is cheap.
Two practice rules come with it. In every Thursday third-and-long period, at least one rep in three rotates after the snap, and the quarterback doesn't know which one; Darius's line in the spring install was "Show me the spin on Thursday, or don't call it on Sunday." And in the film room, Rome freezes every third-and-long rep at the quarterback's third step and makes him say "open" or "closed" before the tape rolls on.
Teo spent most of the spring trying to break the tag. His best attempt was a scout safety who waited until after the quarterback's third step to rotate. Here's the exchange that settled it.
Teo: He rotates late. Step three, Darius sees open. He throws the post. Late safety's there.
Rome: From thirteen yards deep and a hash away, starting after step three? Show me the tape.
Teo: It's a little late.
Lyle: Then it's a defense that has to disguise less and run farther, on the biggest third down of the game. Good. That's a price. The tag doesn't have to make the rotation impossible. It has to make it expensive.
What that reveals: the tag isn't a cure. It's a new arrow, and Russ will draw one back. That's the threshold concept again. You don't finish a system. You keep answering.
🛡️ The defense's view: Here's how I'd expect Russ to see the tag, based on everything he's ever told me. His January plan leaned on our post-dig deciding before the snap: "The coverage is a contract, and I get to read the contract before I sign it," I told him, and he answered, "That's why we change it after you sign." The tag is our way of not signing until after he's changed it. His likely responses are a later rotation, which shortens his own disguise, or a robber sitting on the dig, which takes a man out of his rush or his deep help. Either one shows up on our QC page by the third game, and either one has its own arrow on our map. That's the conversation.
📋 Try it — Make the call (hypothetical, next season): Sacramento again. Q3, tied 10–10, 3rd-and-8 at the Sacramento 44. The booth read: "Two high. Six. Nickel." The QC page: two-high on 3rd-and-7+ has stayed two-high on 11 of 11 since their Week 4, and the page now writes that streak with a question mark. Wade's fourth-down letter for this spot, set before the series: "G" (go for it on fourth-and-4 or less). Your call, and why?
Answer
A strong answer: "Eleven, Gun Doubles Left, 60 Right, Mills" from the Blue panel, which carries the middle-read tag automatically (or Dagger, which carries it too). Reasoning: (1) two-high is still the most likely look, and against it Mills's post-or-dig read is the best play on the sheet; (2) the question mark means the rotation is possible, and the tag's closed throw (the dig at 12) converts against it; (3) Wade's "G" means a four- or five-yard gain isn't a failure: fourth-and-short in four-down territory is a fine result, so the call doesn't need to force the post. A weaker answer picks a different concept "because they'll rotate." That's trusting the opposite tendency instead of building the branch. The point is to be right after the snap, not to guess better before it.
38.9 Keeping the System Honest: Self-Scout, Tendency-Breaking, and What to Cut
A system of constraints has a weakness a list of plays doesn't: it has tendencies. If your boot always follows two Wide runs, a patient defensive coordinator will see the second Wide run and tell his end to stay home. Your system is a set of promises, and a defense can read promises.
That's why self-scouting is system maintenance, not a separate chore. Marisol's rules are the ones I'd hand any staff. Kill the zeros: every personnel group and formation we use regularly has to show every major play family at least a few times, because a formation that never passes is a run tell. Her tell threshold: a look that predicts one play family about 70 percent of the time or more, on ten or more snaps. Her Monday-after-Game-2 finding was exactly that kind of tell: Ruffin offset toward Harlan in two-by-two on early downs meant counter or counter play-action on 14 of 19. That's a constraint system leaking its own sequence.
Teo's version, since the divisional loss, lives on his page: every count carries a "since" date, and perfect streaks get a question mark, not a bold. The red line on my whiteboard is the same rule in seven words: "Never let a tendency outlive its evidence." It applies to theirs and to ours.
What to cut
Every spring, before we add anything, we cut. Our install rule, the one I took from watching Curtis Bale's second season, is "Nothing new goes into the plan unless something comes out." Here are the five tests we use, and every one of them is a question about the map, not about a play's average.
| Test | The question | Fails when | Example of the failure |
|---|---|---|---|
| Orphan | Can you name the base play it protects and the answer it punishes? | It has no parent and no answer | Bale's 38 formations that each carried one play |
| Dead parent | Are we still calling the play it imitates? | The constraint outlived its base play | A boot called without the Wide runs that pay for it |
| Reps | Can we rep it enough to run it in a game? | It never gets enough reps to be safe ("Rep it or cut it") | The reverse needed twelve clean reps in a row before it went on the sheet |
| One sentence | Can I explain it to Darius in one sentence? | The explanation needs a whiteboard | Any tag that "eats the concept" |
| Same price | Does another play already punish this answer? | Two plays charge the same price, and one is worse | Two different double moves vs. press when the band has room for one |
Notice the difference between the playbook and the call sheet. You cut from the call sheet every week; it's a game plan, and most of your plays shouldn't be on it against most opponents. You cut from the playbook in the spring, and only when a play fails these tests across a whole season.
⚠️ Common mistake: cutting the parent because its average is low. Marisol's early-season ledger had plain Inside from Gun Doubles at about zero expected points per play over 58 calls. The children from the same picture were strongly positive: the RPOs at about +0.19, play-action at about +0.41, the family as a whole at about +0.12. Cut the plain runs and the children lose the thing they imitate. Fix: grade the family, then check the specific defender who makes the family work, like Teo's Holloway card (15 of 21 triggers on film). If you can't show that defender's feet, the average wins.
🔄 Check your understanding: A staff notices that its play-action pass from Twelve has averaged 14 yards a call, and that its Twelve runs have averaged 3. A young coach suggests calling more play-action and fewer runs from Twelve. Which test in the table does that idea threaten, and why?
Answer
The dead-parent test. The play-action's value comes from defenders who believe the Twelve run is coming. If the staff calls fewer runs, the safeties and linebackers have less reason to fit the run hard, and the play-action stops fooling anyone. Its 14-yard average is partly borrowed from the runs. The fix is to keep enough runs from the same picture to keep the defense honest, and to watch the specific defenders the play-action targets, not just the yardage.
38.10 The Minimum Viable Offense, at Every Level
If an offense is a system of constraints, the smallest possible offense isn't "your ten best plays." It's one base play and a price for each answer the defenses you'll face actually give. Everything else is depth.
A MINIMUM VIABLE OFFENSE (high school): one engine, five prices
INSIDE ZONE (or POWER)
|
+-------------+--------------+--------------+--------------+
| | | | |
safety fills end crashes 8 in the box rushers pin corner
| | | ears back squats
v v v | |
PA off it zone read bubble/now v v
(seam/flat) (QB keeps) (count it) one screen hitch -> fade
📊 Diagram (described): One base run at the top, inside zone or power, with five lines coming down from it, each labeled with a defensive answer and ending in the play that punishes it. A safety who fills is punished by play-action off the same run. An end who crashes is punished by the zone read. An eighth man in the box is punished by a bubble or now screen. Rushers who pin their ears back are punished by one screen. A corner who squats on short throws is punished by turning the hitch into a fade.
That's ten or twelve plays with formations and both directions, and it's a real offense. It's close to what I built in my first high school season in northeast Ohio, after starting four-wide and losing 14–7 and 21–6. In Week 5 we rebuilt around two-tight-end wing Power and one play-action pass to our tight end, Danny Szymanski, off the same look, plus a couple of companions. We won 20–13 that night and five of our last six. It wasn't a better playbook. It was a smaller system that our players could execute and that had an answer for the one thing every defense in that league did to us: bring the safety down.
| Level | Engine | The minimum prices | Roughly how many plays | What changes |
|---|---|---|---|---|
| High school | One run family (inside zone or power) | PA off the run, a read or an RPO, one screen, one quick concept with a double-move tag, one boot or counter | 12–20 with formations | Two-way players and short practices; wide hashes (the boundary is short); check your code's ineligible-downfield limit (often 2 yards) |
| College | Two run families (zone plus a gap run) | The HS set plus a hi-lo concept, a man-beater, a five-man protection with hots, and the QB run if you have one | 30–45 | Wide hashes again; looser ineligible rule historically (up to 3 yards; check the current code); tempo is a bigger weapon |
| Pro | Inside and outside zone, plus gap runs | Everything on the map, plus a weekly zero bin, a shots box, and the middle-read tag on third-and-long | A 60–70 play core, game-planned down each week | One-yard ineligible rule; narrow hashes; defenses that answer within a quarter |
| Film room | Whatever the team runs most on 1st-and-10 | Find its PA, its RPO, its screen, its pressure answer | Chart 20 snaps of a game | You're building their map from the outside; mark every arrow you can't see as "?" |
🏫 Any level: The level changes the prices more than the engine. The wider high school and college hashes make the field side a natural home for bubbles and boots, and the boundary side a natural home for the run. The looser ineligible-downfield limits historically allowed in college and high school make some RPOs legal there that would be flagged in the NFL, so check your current code before you install any RPO whose pass goes past the line. And at lower levels your engine is chosen by your linemen: if they can't move laterally, don't build on outside zone.
🩺 Player safety: A minimum offense built on a running quarterback puts all of its weight on one player's body. Our budget in Austin is about three designed quarterback carries a game, up to five in the playoffs with Wade's sign-off, and quarterback power only inside the 10 or on short yardage. At lower levels the quarterback may also play defense; Teo's high school quarterback in El Paso was also the starting free safety, and Teo's carry budget began only after an injury. Set the budget before the season, not after the hit.
Spaced Review
Answer each before you read the reminder underneath it.
1. From Chapter 9: What's the Outlaws' box-count rule for inside zone, and what does the "eight" half of it tell you about where the conflict moves?
Reminder: "Seven or fewer, run it. Eight, take the throw." When the defense puts an eighth man in the box, that man has a run job and a coverage job, and he's now the conflict defender. The RPO or the quick throw goes where he isn't. That rule is a whole row of this chapter's map.
2. From Chapter 22: In protection arithmetic, what counts as a blocker that isn't a person?
Reminder: The ball. When rushers outnumber blockers, the hot throw or the sight adjustment is the extra blocker. Game 2's sack happened because the pressure dropped a defender into the hot window and took the ball away as a blocker; Game 3's rule blocked every rusher instead.
3. From Chapter 25: Why was Mills on 3rd-and-9 in the divisional round a defensible call that still deserves a process grade below an A?
Reminder: A call is a bet on a distribution. Against the likeliest look, two-high, Mills was the best bet on the sheet. But its distribution had a terrible tail, an interception against rotation, and the plan had no branch to shrink that tail. The process failure was the empty arrow, not the concept.
If any of the three felt shaky, reread the section in the chapter it came from before you build this chapter's map. The map is made of exactly these pieces.
Project Checkpoint: System Map
Last chapter you built your Coordinator Interview Binder: ten tabs, under 30 pages, each opening with a one-sentence rule. Tab 2 was waiting for this.
This chapter's deliverable: the System Map, your constraint matrix. It shows how your run, pass, screen, play-action, RPO, and tempo pieces answer each other, and where they don't.
Format. Two pages. - Page 1, the map. One diagram, in the style of 38.6: base plays on the left, the defensive answers you expect in the middle, and an arrow from each answer to the play that punishes it. Add a band for tempo (when you go fast, when you never do) and a band for upkeep (how you self-scout). - Page 2, the matrix. A table with one row per defensive answer and these columns: the answer · the defender who carries it · what it gives up · our punishing play · the tell that says call it now · the snap that proves it (from your film, a scrimmage, or a game) · since (the date you last saw the answer). - At the bottom: an Empty Arrows list. Every answer you know a defense could give that has no play after it. Be honest. This is the most valuable line on the page.
Worked example (one matrix row, pro level):
| Answer | Carrier | Gives up | Our play | Tell | Proof | Since |
|---|---|---|---|---|---|---|
| Field safety fills the alley vs. Inside | Holloway ($) | His deep quarter | Inside PA, Z Post | Two receipts: two same-look Inside runs he's filled on this game | Game 1, Q4: +5, +7, then +35 | Week 6 (changed by Week 15) |
Choose your level: - High school: One engine and five prices (38.10's tree). One page, both sides. Your Empty Arrows list should include whatever your best opponent does that you can't rep in your practice time. - College: Two engines. Add a hash column (field vs. boundary) to the matrix, because the same answer often gets a different price on each side of a college field. - Pro: The full map, plus a version against the Sacramento Sentinels using Appendix G. Your Empty Arrows list must include post-snap rotation, and your answer to it. - Film room: Build the map of a real NFL offense from the outside, from three games of public film. Mark every arrow you think exists but can't prove with a "?". Then watch its next game and see which question marks it answers.
Next chapter: the capstone. You'll take every deliverable you've built, this map included, and assemble the complete playbook and a full game plan against the Sentinels.
Chapter Summary
The threshold concept. An offense is a system of constraints. Every play is the answer to the defense's answer to another play. You don't judge a play alone; you judge it by its family and by the answer it punishes.
The three words. A base play is cheap, sound, and repped. A defensive answer gives one defender a second job or adds a defender. A constraint play looks like the base play and goes where that answer emptied, so the base play becomes cheap again.
The map, in one table
| The defense's answer | What punishes it in the Outlaws' system |
|---|---|
| Alley safety fills | Inside PA, Z Post |
| Will fills | Inside Slant |
| Eighth man in the box | Inside Glance, the quick game ("eight, take the throw") |
| Backside end crashes / chases | Inside Read; Wide PA Boot, Naked, the reverse |
| Linebackers key the puller | Power PA, Counter Y |
| Pressure (five or six) | Hots, sight adjusts, screens, the zero bin |
| Rush upfield, light box | Draw, Slip, Middle screen |
| Man coverage | Mesh, Shallow, stacks and Lace, the scramble |
| Cover 2 corner squats | Smash |
| Quarters corner's eyes on #2 | The bender |
| Rotation after the snap | All Go; the middle-read tag |
| Tempo | Multiplies whichever side is winning; Fast after a constraint hits, never with a fourth-quarter lead |
The middle-read tag. Printed on every Blue third-and-long call with a deep-middle throw; "Eyes" adds it anywhere else. Name both throws; eyes to the grass on step 1; declare on step 3; open → the concept's read, closed → never the post, the closed throw on the same rhythm, then rule four.
The decision framework, for any play you're thinking about adding or calling 1. Which base play does it protect? 2. Which defensive answer does it punish, and who carries that answer? 3. Is that answer actually being given? (What's the snap, and what's the "since" date?) 4. Does it look like its parent for as long as it has to? 5. If the defense answers it, what's my next arrow, and is there one? 6. Can I explain it in one sentence and rep it enough to run it?
If you can't answer 2 and 5, you don't have a play yet. You have half a sentence.
What's Next
You have every piece now, and a map of how they fit. Chapter 39 asks you to build the whole thing: the complete playbook, the weekly game-plan process, and a full plan against the Sacramento Sentinels. You'll also sit in on the staff meeting where Teo Arriaga, who started this book asking how to tag a coverage, stands up and defends his own third-down and red-zone package against the people who taught him.