76 min read

It was the middle of June, after our spring work was done and a few weeks before training camp. It

Chapter 40 — Watching Football Like a Coordinator: How to See What You've Learned on Every Snap of Every Game

The Hook: The Copy Everyone Else Watched

It was the middle of June, after our spring work was done and a few weeks before training camp. It was Teo Arriaga's last week in our building, and I was carrying two coffees down the hall at ten o'clock at night because some habits outlive the seasons that made them.

Teo had the big screen in the quality control office running something I'd never seen. It was the television broadcast of the divisional playoff game at Sacramento. It wasn't the coaches' film. It was the copy everyone else watched, with the score graphic in the corner, the replays, and the sound on.

"I've never watched it this way," he said. "Not once. I wanted to know what it looked like from the couch."

So we watched it from the couch.

Our first snap of the game was a jet sweep at our own 29. On our film you can see the whole story of that play in the first second: Mekhi Powers starts across the formation at full speed, and the Sentinels' safeties move. Tre Holloway comes down. Ike Royce goes to the middle. Two high safeties before the motion, one high after it. We gained a yard, and it was the most useful yard of the first quarter, because it told us how Russ Kimbrough wanted to handle motion that day.

On the broadcast, the camera followed Powers. Of course it did. He was the man running fast. The safeties were never in the frame. The analyst said it was "a nice little wrinkle to start," and the picture cut to our sideline, where I was looking at my call sheet.

"You can't see it," Teo said. "The roll. The whole reason we called it."

"No."

We skipped ahead to 2:41 of the fourth quarter. Third-and-9 at our 38, down 20–17. You know this snap by now. The broadcast came in tight from behind the line on Darius Vance taking the snap. The safeties were thirteen yards deep and off the edge of the screen. The ball left his hand. The camera whipped downfield, and there was Ike Royce, number 21, in the middle of the field where nobody had been before the snap, catching a pass that was meant for Trey Gaines.

"He just never saw him," the analyst said. "You can't throw that ball."

Then the replay came, from the camera that hangs on wires above the field, behind the quarterback. In slow motion, from up there, it was obvious. The strong safety drops down at the snap, Royce turns and sprints to the middle, and Gaines runs his post straight into him. On the replay everyone in the country could see what Darius had about a second to see in real time.

Teo paused it on Royce's turn and sat back.

"A stadium full of people and however many million at home," he said, "and none of them could see the thing that decided the game until the replay."

"And on the replay," I said, "everybody's a genius."

He laughed, but he didn't let it go. That's why I hired him. "Coach, seriously. If this copy was all I'd ever had — no coaches' film, no call sheet, nobody to ask — could I have learned anything from it? Could I have graded that call?"

I thought about it longer than I usually think about anything.

"Most of it," I said. "Not all of it. And knowing which part is which is most of the job."

This is the last chapter of the book. For thirty-nine chapters you've sat in our staff room. You've learned how we block an area, how we read a defender instead of a receiver, how we pay for play-action with runs, and why a call is a bet and not a promise. Now you're going to take all of it home and turn on a game, any game, on a Sunday this fall. This chapter gives you a repeatable way to watch: what to look at before the snap, what to look at after it, the question to ask about every call, which camera angle can answer which question, how to chart a game yourself, and how to grade a coordinator honestly. That includes the honesty to admit what you can't see from where you're sitting. Then you'll take a ten-snap test, and we'll grade our four anchor plays one last time.

And then Teo leaves, and I'll tell you about that too.

🏃 Fast Track: If you already break down film, skim 40.1–40.3 (the checklist will look familiar) and go to 40.4 ("why this call now"), 40.7 (hindsight and resulting), and the ten-snap self-test in 40.9. The Project Checkpoint's rubric is the deliverable.

🔬 Deep Dive: The pre-snap checklist compresses Chapter 7 and Appendix E. The grading method comes from Chapter 25 and Chapter 28. The charting sheet is in Appendix F.

📺 Film Watcher: This Sunday, pick one offense. Before every snap of its first two drives, say out loud three things: the personnel, the number of high safeties, and the down-and-distance word (short, medium, long, or long-long). Don't watch the ball until you've said all three. By the end of the second drive, you'll notice you're watching a different game than the one the broadcast is showing you.


40.1 One Snap, Four Questions

Every coordinator I've worked with watches film in a loop, whether he'd describe it that way or not. Walt Dupree, my first NFL boss, used to say, "Every play is a question. Know what you're asking." When you watch somebody else's offense, the job turns around. You're trying to figure out what question the play was asking, and whether the defense's answer was the one the offense expected.

That job breaks into four questions, asked in the same order on every snap:

  1. Before the snap: what did they line up in? The situation, the personnel, the formation, the defense's shell and box, the leverage, and what happened when somebody moved.
  2. At the snap: what happened? The scheme, the defender it attacked, the decision the quarterback or back made, and the block or route that decided it.
  3. Why this call, now? Down, distance, field position, score, clock, and the snaps that came before it.
  4. How good was it? The process grade first, and only then the result.
THE COORDINATOR'S VIEWING LOOP (one snap, every snap)

   +---------------------+          +---------------------+
   | 1. BEFORE THE SNAP  |          | 2. AT THE SNAP      |
   | situation, pers.,   |  ----->  | line's first step,  |
   | formation, shell,   |          | the read defender,  |
   | box, leverage,      |          | the decision, the   |
   | motion + reaction   |          | block that decided  |
   +---------------------+          +---------------------+
              ^                                |
              |                                v
   +---------------------+          +---------------------+
   | 4. HOW GOOD WAS IT? |          | 3. WHY THIS CALL,   |
   | process grade first |  <-----  | NOW? down, distance,|
   | result grade second |          | field, score, clock,|
   | name what you can't |          | the last three snaps|
   | see                 |          |                     |
   +---------------------+          +---------------------+
          (the grade feeds your read of the NEXT snap)

📊 Diagram (described): Four boxes arranged in a square with arrows running clockwise. Box 1, top left, is "before the snap": situation, personnel, formation, shell, box, leverage, and motion with the defense's reaction. An arrow runs right to box 2, "at the snap": the line's first step, the read defender, the decision, and the block that decided the play. An arrow runs down to box 3, "why this call, now?": down, distance, field, score, clock, and the last three snaps. An arrow runs left to box 4, "how good was it?": process grade first, result grade second, and a note naming what you can't see. An arrow runs back up from box 4 to box 1, labeled "the grade feeds your read of the next snap."

Notice what's missing from the loop: "Did it work?" It's in there, but it's buried inside question 4, and it comes second. That's on purpose. The broadcast is built around "did it work," because that's what most of the audience wants to know. A coordinator watches the other way around. By the time the result shows up on the screen, he has already decided what the play was for.

Here's the part nobody tells you: the loop is slow at first and fast later. The first game you chart this way will take you three or four hours for one team's offense. By the tenth game, you'll run the first two questions in about the time it takes the broadcast to show the replay. Teo could tag a Sentinels snap on eighteen fields in about the time it took me to drink half a coffee. He wasn't born able to do that. He did it eight thousand times.

💡 Coach's point: Watch the question, not the ball. The ball tells you what happened. The alignment, the first step, and the read defender tell you why, and "why" is the only part you can use next week.


40.2 Before the Snap: The Seven-Item Checklist

You have about ten seconds between the broadcast's last replay and the snap, and less if the offense is going fast. In that window you're running a checklist. Here it is, in the order I run it. If you've read Chapter 7, every item will be familiar. The point now is speed and order.

# Item What you count or name What it tells you
0 Situation Down, distance, field position, score, clock, timeouts Which part of the call sheet the coordinator is in (third-and-long box, red zone, four-minute)
1 Personnel Running backs and tight ends: 11, 12, 21, 13… And the defense's answer: base, nickel, dime What the offense is threatening and what the defense chose to match it with
2 Formation and strength 2x2, 3x1, empty, bunch; where the tight end is; attached or detached; receivers numbered #1, #2, #3 from the outside in on each side Where the offense has put its numbers and its leverage
3 Shell High safeties: two (middle of the field open, MOFO), one (middle of the field closed, MOFC), or zero The coverage families still possible. The shell is not the coverage.
4 Box Defenders inside the tight end or tackles, within about five yards of the ball: light, neutral, or loaded Whether the run has numbers
5 Leverage Corners pressed or off; inside or outside on the slot; safety depth Which routes the defense has given up before the snap
6 Motion and reaction Who moved, and who went with him: a defender traveling (man), defenders sliding (zone), safeties rotating, or nobody The defense's coverage answer, revealed before the snap

I count the shell before the box. That's deliberate: the number of safeties who are deep tells you how many defenders can be in the box. Two high safeties against 11 personnel usually means a six-man box. One high usually means seven, sometimes eight. If you count the box first, you'll miscount it, because you'll miss the safety creeping down from ten yards to eight.

Let's run the checklist on a snap you know: Anchor A, the first snap of the first Sacramento game. (As always, the Outlaws, the Sentinels, and every player and game in this book are fictional composites.)

PRE-SNAP: Game 1, 1st-and-10 at the AUS 32, 0-0
"Eleven, Gun Doubles Right, Inside Slant" vs. nickel

            F                               $       <- (3) SHELL:
                                                       two high
   C              *                                C   (MOFO)
                       (W)      M                   <- (5) corners
                    E       N      T   E               off
   X              H   O   O   O   O   O Y        Z  <- (2) 2x2, Y
                           R  Q                        attached rt
                                                    <- (4) BOX: 6

📊 Diagram (described): Austin is in two-by-two from the shotgun: on the left, Sutton (X) wide and Powers (H) in the slot; on the right, Harlan (Y) attached to the right tackle and Gaines (Z) wide. Ruffin (R) stands to Vance's left. The Sentinels are in nickel with two high safeties, Royce (F) on the left and Holloway ($) on the right. The nickel, Vo (*), is over Powers with outside leverage. The Will linebacker, Ames (W), is in the box on the weak (left) side, with Bell (M) to his right. The front has four linemen. The annotations mark the checklist items: two high safeties, corners off, a two-by-two formation with the tight end attached right, and a box of six (four linemen and two linebackers).

Run the items:

  • Situation: first-and-10 at our 32, opening drive, 0–0. This is the script. The call is probably a question, not an answer.
  • Personnel: 11 (one back, one tight end, three receivers). Sacramento answered with nickel, five defensive backs.
  • Formation: Doubles, the tight end attached on the right. Strength right.
  • Shell: two high. Middle of the field open. Light-box territory.
  • Box: four linemen and two linebackers: six. The nickel is outside, over the slot, so he isn't in the box. Count our blockers: five linemen and Harlan, six. Six on six, and one of their six has two jobs.
  • Leverage: the nickel is over Powers with outside leverage. That leaves the inside window, the slant window, to the Will linebacker.
  • Motion: none on this snap. (Note that, too. "No motion" is information on a team that motions on more than half its snaps.)

That last line of the box count is the whole snap. One of their six has two jobs. Ames is responsible for his run gap and for the slant window. You could see the conflict before the ball was snapped. You didn't need to know our call. You only needed the count.

🔄 Check your understanding: On the broadcast, the camera often starts tight on the quarterback and the line. Which two checklist items are you most likely to miss on a broadcast copy, and what can you do about it?

Answer

The shell (the safeties are often out of the frame at the snap) and leverage on the outside receivers (the corners are at the edges of the screen or off it). What you can do: look during the pre-snap wide shot, which most broadcasts show for a few seconds after the huddle breaks, and write "?" in the shell column when you didn't see it. Then check the replays. The overhead or end-zone replay often shows where the safeties ended up, and that tells you what the shell turned into, even if it can't tell you what it looked like before.

The motion item, done properly

Motion is the only item on the checklist that the offense creates for you. Every time a receiver moves before the snap, the defense has to answer in front of you. In Chapter 27 we called this motion for information. Here is the short table I'd tape to the remote:

What the defense does when he moves What it usually means Watch for
One defender runs with him across the formation Man coverage The offense's man-beater: a rub, a crosser, a double move
Defenders bump over a gap or a man; nobody crosses Zone, matching by a rule The offense attacking the side that got lighter
A safety drops down and another goes to the middle Rotation to one high The run going away from the rotation, or a throw behind the rotated safety
Nothing moves at all A disguise, a check, or a team that won't tell you The "both" call: a concept with a man-beater on one side and a zone-beater on the other

The caution on that table is the one Russ taught us in Week 15. A defense that knows you motion for information can lie with its reaction. In the first quarter of that game, our motion on third down showed us a zone slide, and then they played man. The table tells you what a reaction usually means. It can't tell you what Russ meant.

⚠️ Common mistake: reading the formation and skipping the reaction. A viewer sees a receiver motion across and says "jet motion" and stops there. The motion is the offense's question. The defense's reaction is the answer, and the answer is the useful part. The fix: every time anyone moves, take your eyes off the man in motion and put them on the safeties and the defender across from where he's going. Write down who moved with him.


40.3 At the Snap: Four Things in Two Seconds

Pre-snap you have ten seconds. After the snap, most of what matters happens in the first two. You won't see all of it the first time. That's why you'll watch the play more than once (40.5). But you should know what you're looking for, in order:

  1. The line's first step. It tells you the play family before the ball is handed off or thrown.
  2. The read defender. Find the man the play is built to put in conflict, or the man left unblocked on purpose.
  3. The decision. The quarterback's handoff, pull, or throw, or the back's cut.
  4. The block, or the route, that decided it.

The first step tells you the family

What the five linemen do on the first step The family What to look at next
All five step the same direction, flat or slightly upfield, helmets level Zone (inside or outside) The back's aiming point and the first down lineman past the center
Play-side linemen block down; a guard or tackle pulls Gap (power, counter, trap) The pullers: who kicks out, who wraps up through the hole
Linemen set back, hands up, weight back ("high hats") Dropback pass The protection: who's blocking whom, and who's free
Linemen fire out like a run, but don't go past the line Play-action The quarterback's turn after the fake, and the defender who bit
Linemen set, then let rushers through and release Screen The linemen getting out in front of the catch
Linemen fire out like a run and keep going (watch the one-yard rule) RPO or run The quarterback's eyes at the mesh: is he reading somebody?

That last row is worth a word. On an RPO (a run-pass option, where the quarterback hands off or throws based on one defender's reaction), the linemen block a run. In the NFL they can't be more than a yard downfield when the pass is thrown, which is why you'll see linemen on an RPO "hit, don't travel," as Gus Ferrante teaches it. College and high-school codes have allowed more distance; check your current code. If you see a lineman three yards downfield and a pass completed past the line, either it's a flag or the pass was thrown behind the line.

🧩 Productive struggle: Before you read on, try this. The line's first step looks like zone to the right. The back takes the handoff aiming at the right guard. The tight end on the right doesn't block the end in front of him; he releases past him to the second level. The end is unblocked and crashes down the line. The back is tackled for one yard by that end. Is this a busted assignment or a design? What would you need to see to tell the difference?

Answer

You can't tell from the run alone. An unblocked end could be a missed block, or he could be the read defender on a zone read, in which case the quarterback was supposed to keep the ball when the end crashed. Look at the quarterback. If he rides the back, eyes on the end, and gives it anyway while the end crashes, the decision was wrong (or he was told to hand it off). If he never looked at the end, it may be a bust, or a designed give with the end left for a block that didn't come. If the tight end's release took him to the alley safety, that's a design called "Arc" in our language (the tight end arcs around the read end to block the safety). The lesson: an unblocked defender is a question, not an answer. Find the man who was supposed to account for him, and the man who was supposed to read him.

Find the read defender

Every play we've drawn in this book stresses a specific defender. In zone runs, it's whoever the back reads. In option and RPOs, it's the man the quarterback reads. In pass concepts, it's the defender the concept puts in a high-low or inside-outside bind, the one who can't cover both routes. In play-action, it's the defender whose run key the fake is aimed at.

Here's the post-snap picture of Anchor A, snap 1, drawn as a viewing order: where your eyes go first, second, and third.

POST-SNAP (Game 1, snap 1): three looks, in this order

   ^ go            ^  <-- (3) THE BALL: Powers's slant, +17
   |              /:
   |             / :
   X            H  :      (W)   <-- (2) THE KEY: Ames comes
                   :       |         downhill on the fake
                   :       v
                   :  E   N   T   E
                   :  ^   ^\  ^  /^
                   :  O   O   O   O   O   Y--> $  (arc)
                   :.......Q    R===>
                    (pulls)     fake  <-- (1) THE LINE: all five
                                          step right = zone

📊 Diagram (described): After the snap, the whole Austin line steps to the right: inside zone. Harlan, the tight end, arc-releases toward Holloway, the safety on the right. Ruffin takes the fake going right. The Will linebacker, Ames, comes downhill toward the run. Vance pulls the ball out of Ruffin's belly, and a dotted line shows the throw going up and to the left to Powers, who has run a slant from the left slot into the space Ames just left. Sutton runs a go route on the far left to clear the corner. The labels number the viewer's three looks: first the line (all stepping right, so zone), second Ames (the key, coming downhill), third the ball (the slant, 17 yards).

That's the sequence for any RPO. First the line, which tells you it's a run play. Then the defender whose run fit and pass job overlap. Then the ball. If you watch the ball first, you see "slant completion, 17 yards." If you watch in order, you see "the Will linebacker had two jobs and chose the run, so the throw went to the job he gave up." Only the second one teaches you anything.

The decision and the block that decided it

The third look is the decision. The quarterback's head tells you a lot. On a quick throw he looks at one defender and the ball comes out in about a second. On a dropback his eyes move: middle, then a side, then the checkdown. Our quarterback's ball comes out at about 2.9 seconds against a four-man rush, and his hot throws at about 1.0 to 1.3. Count it yourself ("one-thousand-one…"). When the ball is still in his hand past three seconds against four rushers, something went wrong before the throw. Either the coverage won, the quarterback missed the open man, or the protection broke and moved him.

The fourth look is the key block: the one block, or the one route, the play can't survive without. On inside zone it's usually the double-team that has to climb to the Mike linebacker. On power it's the kick-out. On a quick game snap it's the receiver's release against press. On the third-and-7 protection in Game 3 it was our right tackle widening to the corner blitzer. When a play fails, start at the key block. Most failures die there, and the broadcast will almost always show you the wrong guy.

🔄 Check your understanding: On the broadcast, a back takes a handoff and gains three yards. On the replay you see the right guard and right tackle block down to their left while the left guard runs behind them to the right. Name the family, and name the block you'd check first if the run had been stopped for no gain.

Answer

A gap scheme: down blocks plus a pulling guard, so power or counter (one puller who wraps through the hole suggests power; a second puller would suggest counter). The block to check first is the one the play can't survive without: the kick-out of the end man on the line, or, if the puller is the wrapper, whether he found the linebacker. On a gap run, the front side is a promise, and the pullers are the ones who keep it.

📋 Try it — Find the conflict: Week 10 at Portland, the second quarter, 2nd-and-6 at the Portland 31, tied 7–7. You see this pre-snap: Austin in 11 personnel, Doubles, the tight end attached on the right. Portland is in its two-high shell, with the corners rolled up near the line at about five yards with outside leverage, and the safeties at about twelve yards on the hashes. On the left, Sutton is wide and Powers is in the slot. Name the defender most likely to be put in conflict by a two-receiver concept on the left, the two routes that would do it, and what you'd watch at the snap to know which route gets the ball.

Worked answer

The conflict defender is the left cornerback, Josh Kiel (#23). A rolled-up corner with outside leverage in a two-high shell is a Cover 2 flat defender: his job is the short outside zone, with a safety over the top. Put a short route under him (a hitch by the outside receiver at about six yards) and a deep outside route over him (a corner route by the slot, stemming to about twelve and breaking to the pylon) and he's in a high-low bind. If he sits on the hitch, the corner route is open behind him and in front of the half-field safety. If he sinks under the corner route, the hitch is open in front of him.

At the snap, watch the corner's feet, not the receivers. If his weight goes forward to the hitch, the ball goes over him. If he backpedals and turns, it goes under. And glance at the left safety: if something in the middle (a seam by the tight end) holds him on the hash, the corner route has room to land.

This is Anchor B. Kiel squatted on the hitch; Powers caught the corner for a 31-yard touchdown. You could have known the conflict before the ball was snapped.


40.4 "Why This Call, Now?"

This is the question that separates a coordinator's eyes from an analyst's. An analyst asks, "Was that a good play?" A coordinator asks, "Why did he call that play, here, now, after those three?"

A play call is a sentence in a conversation. It means something because of what came before it, and it changes what can come after it. You can't grade it by itself any more than you can grade one line of an argument. These are the five context questions I ask of every call I watch:

Question What you're looking at What it might explain
1. What does the down and distance demand? Short (1–3), medium (4–6), long (7–10), long-long (11+) A third-and-long screen isn't always a give-up. On 3rd-and-12 in field-goal range, it can be a field-goal-improving play
2. Where is the ball? Own territory, midfield, "fringe" (their 30–21), red zone, goal line Shots live in a zone (ours: our 40 to their 35). Inside the 10, the field shrinks and the timing changes
3. What do the score and clock want? Lead or trail, by how much; time and timeouts Four-minute offense runs to keep the clock moving. Two-minute offense throws to the sideline to stop it
4. What did the last three snaps buy? Same formation? Same run? How did the key defender play it? Sequences: a run that made a defender aggressive, followed by the play-action aimed at him
5. Who is being attacked? A young corner, a slow linebacker, a safety with two jobs Matchup calls: the same concept thrown at the same defender again and again

Question 4 is the one fans skip and coordinators live on. I've told you since Chapter 1 that I don't call plays. I call the next play's reason to work. When you watch someone else, look for that reason.

Anchor D through question 4

Here's the four-play sequence from Game 1, watched the way you'd watch it from the couch, with only question 4 in mind. Sentinels lead 27–20 in the fourth quarter. Same formation every snap: 11 personnel, Doubles Right, Harlan attached right, Ruffin offset left. Same Sentinels look every snap: two-high Quarters, Holloway the strong safety on the right at ten or eleven yards.

Snap Call What Holloway did What the snap "bought" for the next one
1st-and-10, AUS 49 Inside zone right Filled downhill into the alley Receipt one: he'll come down on this look
2nd-and-5, SAC 46 Inside zone right Filled harder and earlier Receipt two: he's sure now
1st-and-10, SAC 39 "Inside PA, Gaines Post" Bit on the fake Gaines 35 yards behind him to the SAC 4. And now he doesn't trust the run
1st-and-goal, SAC 4 Inside zone right One step, then stopped Ruffin's 4-yard TD, 27–27

Watched one play at a time, that's two ordinary runs, one great play-action pass, and a short touchdown run. Watched as a conversation, it's one argument with four sentences. That's our "two receipts" rule: we call the play-action only after two same-look runs the defense has paid for. (For the whole sequence, frame by frame, see Chapter 19.)

The fourth snap is the one you'll miss if you stop thinking at the big play. Why run inside zone on first-and-goal right after a 35-yard play-action? Because the play-action changed Holloway. He'd been wrong three times in a row, twice by being right and once by being fooled. A safety who's just been burned for believing the run arrives a half-step late to the next one. The touchdown run was the play-action's second payment.

🛡️ The defense's view: When Russ watched that drive in the spring, he wasn't upset with Holloway for biting. He told me Holloway's alley fill was right on the first two snaps, and his job made the third one possible. A strong safety in Quarters who has the alley and a deep quarter has two jobs, and play-action is built for men with two jobs. Russ's point about play-action, the one he made to me about a different snap that spring, was this: "You can't lie to a man with one job." So when you watch a defense get beaten by a sequence, look for its answer the next time the offense shows the same look. The good ones don't try to stop the play. They change the man's job.

🎙️ On the headset: Here's what I was actually thinking before snap 3, for comparison with what you could see from the couch. "Two receipts. Holloway triggered at the mesh on the first one, and Teo says he left on the guard's first step on the second one. Royce is on the far hash with Sutton. Ball's at their 39, so a post lands in the end zone or close to it. Inside PA, Gaines Post. Seven-man protection, because if Rollins comes free on a seven-step fake, none of this matters." From the couch you could see every one of those facts except Teo's note about the guard's first step. You could have called that play.

🔄 Check your understanding: A team runs inside zone on first down from the same formation three straight series, gaining 4, 3, and 5 yards, and the strong safety comes downhill on all three. On the fourth series it lines up in that formation on first down and calls a quick screen to the other side. Give one reason this could be a good call and one reason it might be a missed chance.

Answer

Possibly good: if the defense had started loading the box to the run side, a screen away from the run attacks the other half of the field, which has fewer defenders. It still uses the run's credibility, because it comes from the same picture. Possibly a missed chance: the safety had paid three receipts. This was the moment for play-action behind him, and a screen doesn't cash the receipts. To grade it fairly you'd want the safety's alignment on this snap. If he was already cheating down before the snap, the play-action may have been less attractive than it looks.


40.5 Three Angles: Broadcast, All-22, and End Zone

You've heard me and Teo talk about "the copy" all book long. There are really three, and each one can answer some questions and not others. Matching the question to the angle is half of watching well.

Angle What it is What it's good for What it hides
Broadcast The television feed: a high sideline camera that follows the ball, plus replays from many angles Situation, personnel (if they show the substitution), the result, and replays that sometimes show the whole field The safeties at the snap, the corners' leverage, the backside of the play, anything away from the ball
All-22 (sideline coaches' film) A high, wide sideline shot that keeps all twenty-two players in the frame Coverage: safety depth and rotation, where corners bail, every route, where the ball went and why The front's techniques and first steps (players overlap), the quarterback's eyes
End zone A camera behind the play, high, looking straight down the field The line: splits, techniques, first steps, double-teams, gap fits, linebacker depth; the quarterback's drop and eyes; run lanes Depth: you can't tell how deep a safety is, or how far a route went
WHAT EACH ANGLE CAN SEE (top view; offense moving up)

           F . . . . . . . . . . . . . . . $    <- All-22 only
                                                   (broadcast: on
   C  . . . . . . . . . . . . . . . . . . . . C    replays, if
                                                   you're lucky)
          +-----------------------------+
          |  *     W      M             |   <- the broadcast
          |     E   T   N   E           |      frame at the
   X      |  H   O   O   O   O   O   Y  |   Z  snap: the box
          |          Q  R               |      and the slots
          +-----------------------------+
                                            <- X and Z: often
                 [ END ZONE CAMERA ]           out of frame
            behind the offense, high: the
            line's steps, the gaps, the QB's
            drop and his eyes

📊 Diagram (described): A top view of an offense and defense with a rectangle drawn around the middle of the picture: the box, the two slot players, the line, and the backfield. That rectangle is roughly what a broadcast shows at the snap. Outside the rectangle, the two outside receivers, the two cornerbacks, and the two deep safeties are marked as visible only on the All-22, or on a broadcast replay if you're lucky. Below the offense, a note marks the end-zone camera, high behind the offense, which shows the line's steps, the gaps, and the quarterback's drop and eyes.

If you can only get the broadcast, you can still do most of the work. You just do it differently. Use the wide shot before the snap, which most broadcasts hold for a few seconds when the offense breaks the huddle, to get personnel, formation, and the shell. Use the replays to get what happened away from the ball. The overhead replay from the cable camera behind the offense is often the best look at the safeties a broadcast will give you. And write "?" in any column you didn't actually see.

In recent seasons the league's own streaming subscription has offered coaches' film, both the sideline All-22 and the end-zone angle, to the public. What's included, and what it costs, has changed from year to year, so check what's available before you build your study around it. If you can get it, it changes everything. You'll stop guessing at the coverage.

The four freeze-frames

On any snap you care about, don't just watch it four times. Stop it four times, at the same four moments, and ask one question at each:

Freeze at… Look at… The question
1. The snap Safeties, corners, box What did they show? (shell, leverage, box)
2. The quarterback's back foot hitting (or the mesh on a run) The safeties again, and the read defender What did it turn into? (the shell after rotation, the key's first step)
3. The ball leaving (the throw or the handoff) The quarterback's target and the defender nearest it What did he decide, and was it the right defender to decide on?
4. The first contact or the catch The key block or route Who won the matchup the play depended on?

Frame 2 is the one almost nobody uses, and it's the one that would have shown you the whole Game 3 interception. At the snap, two high safeties at thirteen yards. At Vance's back foot, Holloway already coming down and Royce already turning to the middle. Two frames, about a second apart, and the whole story is between them.

🎬 Film room: On the All-22, freeze-frame 2 is easier to find than you'd think: watch the quarterback's feet and stop the moment his last step of the drop lands. On a three-step drop that's roughly a second after the snap; on a five-step from the gun, a little longer. Count the safeties in the middle third of the field in that frame. That number is the coverage. The number at the snap was the disguise.

🔄 Check your understanding: You want to know three things about a snap: (a) whether the 3-technique slanted inside at the snap, (b) whether the field safety rotated down, and (c) whether the quarterback looked at the read defender before he handed off. Which angle answers each?

Answer

(a) The end zone: first steps and techniques. (b) The All-22: safety depth and movement (freeze-frame 2). (c) The end zone, which shows the quarterback's helmet and eyes at the mesh. The broadcast might catch (c) on a close replay, and almost never (b) at the moment it happens.

🏫 Any level: High-school and small-college film is usually shot from the press box, a sideline view that's closer to the broadcast than to an NFL All-22. Most programs also shoot an end-zone angle from a lift or a scaffold, and they trade film through an online film-exchange service (Hudl is the one most coaches will name). Two differences matter for watching. First, the hashes. In high school they sit 53 feet 4 inches from each sideline, much wider than the NFL's 70 feet 9 inches, so the wide side of the field ("the field") and the short side ("the boundary") matter more. Chart the hash on every snap; at that level formation strength often follows it. Second, the quarterback. At lower levels many reads are simplified or made by the sideline. If the quarterback never seems to look at the read defender, he may not be reading anyone. The coach may be calling it.


40.6 Charting a Game Yourself

Watching with a method is good. Writing it down is better, because writing turns "I think they like play-action on second-and-short" into a number you can check. That's what Teo did for a living. You can do a smaller version on a Sunday night.

The twelve-field chart

Teo's hiring test, which I'll tell you more about at the end of this chapter, asks a candidate to tag thirty defensive snaps on twelve fields in two hours. Here are twelve offensive fields built the same way. This is the chart I'd give a reader who wants to chart one offense for one game:

# Field Example entry
1 Quarter and clock Q2 9:14
2 Down and distance 2-8
3 Field position (and hash) STL 44
4 Score (offense first) 7–7
5 Personnel (offense / defense) 11 / nickel
6 Formation and strength Doubles R, Y attached
7 Motion and the reaction Z zip; nobody traveled
8 Box count 7
9 Shell: before → after 1H → 1H (Cover 3)
10 Family and concept Dropback, Smash L
11 Key defender and what he did Left corner bailed to the deep third
12 Result (yards; success Y/N) Hitch +7; Y (2nd down needs 60% of 8 = 4.8)

Field 12 uses success rate from Chapter 28: a play succeeds if it gains 40 percent of the yards to go on first down, 60 percent on second, and all of it on third or fourth. People get the second-down line wrong all the time. Write the rule at the top of your sheet.

The example row is a real snap from Week 12 against St. Louis: "Eleven, Gun Doubles Right, Zip, 60 Left, Smash, X Fade, Alert Post." Smash against Cover 3 isn't the matchup Smash was built for. The corner bails to his deep third instead of squatting, so the corner route is covered. But the hitch is open underneath him, and that's what Vance threw. Seven yards, and Moose Odom converted the third-and-1 that followed. You'll meet it again in the self-test.

Three passes, not one

Teo makes three passes through every game he tags. You should too, because trying to see everything in one pass is how you end up seeing nothing.

  1. Pass one, the situation and the alignment. Fields 1–9. Don't watch the play. Stop at the snap every time.
  2. Pass two, the play. Fields 10–12. Watch the snap through the whistle, freeze-frames 2 through 4.
  3. Pass three, the grade. Only after the first two are done for the whole game. That's where you write "why this call now," the process grade, and the result grade (40.7).

On a broadcast copy, budget about three hours for one team's offense in one game, closer to two once you're practiced. Teo's eighteen-field defensive tagging of a whole opponent game took him five to seven hours. That's the difference between a hobby and a job, and it's why he worked a week ahead.

From chart to tendency

Once you've charted a few games, you'll want to say things like "they run play-action on 2nd-and-short from this formation." Before you say it, apply three rules we live by:

  • Sample size: Marisol's threshold for calling something a tell is about 70 percent on at least ten snaps. Below ten snaps it's a hunch. Write it as a hunch.
  • Since dates: every tendency gets the date it started being true. A team that changed its coordinator, its quarterback, or its left tackle three weeks ago is a different team.
  • Kill the zeros: look for combinations that never happen, a formation with zero runs, a personnel group with zero play-action. A zero is often louder than a 70 percent.

📈 By the numbers: Small samples lie in both directions. If a play really works about 60 percent of the time, ten calls of it can easily come out at four successes or eight, by luck alone. That's why we never trust a ten-snap number by itself, and why the "since" date matters as much as the percentage. The two-high streak that ended our season was twenty-nine snaps long. It was real evidence. It still had a "since" date, and Russ knew the date better than we did.

🔄 Check your understanding: You've charted four games of a team and found that on 2nd-and-1 to 3 it threw play-action on 6 of 9 snaps. Is that a tell? What would you write on your page?

Answer

Not yet, by our standard. It's 67 percent on nine snaps: under ten snaps and just under 70 percent. Write it as a lean with its sample and its since date: "2nd-and-1–3: PA 6 of 9 (since Week 3) — lean, check." Then watch the next game for it. If you see three more and all three are play-action, it's 9 of 12 (75 percent on twelve). Now it's a tell.


40.7 Hindsight, Resulting, and the Honest Grade

Here's the hardest thing in this chapter, and it's not about football.

Resulting means judging a decision by how it turned out. The call worked, so it was a good call. The call failed, so it was a bad call. Poker players have used the word for years, and it describes almost every sports broadcast you've ever watched. Hindsight bias is its partner: once you know how something turned out, it seems as if it was obvious all along. "You can't throw that ball." "Why would you run there?" Of course you can't. You know where the safety went. Now.

A play call is a bet on a distribution, not a promise of an outcome. That was the threshold idea of Chapter 25, and this is where you use it on somebody else.

The 34-second fence

On our sideline, the call is decided by the time the play clock reads 34. So when I grade a call, mine or anyone's, I build a fence at 34 seconds: before I look at the result, I write down what the caller knew at that moment and what the call was trying to do. From the couch, you do it by pausing at the snap.

At the pause, write three things:

  1. The purpose. What is this call for? (Convert a third-and-7, set up a later play, run clock, take a shot.)
  2. The likely looks. Given what this defense has done today in this situation, what are the two most likely coverages or pressures?
  3. The answer to each. Does the call have a good answer against both? Which one beats it?

Then press play. Now the result can't rewrite your reasoning, because your reasoning is already on paper.

WHAT WAS KNOWABLE, AND WHEN (Game 3, the Mills call)

 TUE 12:40 a.m.   "Mills" written in the 3rd LONG vs. 2x2 box.
                  One answer, no branch.
 TUE 7:00 a.m.    Rome's margin note: "Post vs. two-high is
                  pre-snap. Want a post-snap check?"
                  Lyle: "Rule four covers."
 THU              Scout team: two high on every third-and-long
                  rep. The rotation is never repped.
 SUN, 2:41 Q4     3rd-and-9, AUS 38, down 20-17. Two high at
                  13 yds. The streak: 29 for 29 since the bye.
 ---------------- the 34-second fence --------------------------
 SNAP + ~1.2 s    Holloway down; Royce turning to the middle.
 SNAP + ~2.4 s    Ball out to the post. Royce intercepts.
 REPLAY           Everyone can see the spin.

📊 Diagram (described): A timeline of the Game 3 interception, read top to bottom. On the Tuesday of game week, Mills is written in the third-and-long box with no second answer; a few hours later, the quarterbacks coach asks in writing for a post-snap check, and the coordinator answers that rule four covers it. On Thursday, the scout team shows two high safeties on every third-and-long rep and never rotates. On Sunday, at 2:41 of the fourth quarter, the call is made on third-and-9 against a two-high look with a 29-for-29 streak behind it. A dashed line marks the 34-second fence: the moment of the call. Below the line are the things nobody could know at the call: the safety turning at about 1.2 seconds, the throw at about 2.4, and the replay that made it look obvious.

Everything above the fence was knowable at the call, and some of it was knowable on Tuesday. That's the honest grade of that snap. The interception wasn't bad luck; the plan had a hole in it, and Rome had pointed at the hole in writing. But notice what the grade is not. It is not "you can't throw that ball." Darius threw a good ball to the right spot against the picture the whole week told him he'd get. The mistake was ours, upstream, days before the throw.

Two grades, always

Marisol grades every call twice, once by result and once by process. Her process questions, adapted for someone watching from outside, are these:

Process question From the couch, ask…
Was it from the right box? Does the call fit the down, distance, field, score, and clock?
Was the information in the box right? Does it fit what this defense has shown today in this situation?
Was there a better call in the box? Name one alternative and say what it would have done against the likely looks
Did we execute what we'd repped? Was the failure (or success) the call, a rule, or a player? (On our sheet that's C, R, or X)
PROCESS vs. RESULT: where every snap you grade belongs

                       RESULT GOOD               RESULT BAD
                 +------------------------+------------------------+
  PROCESS GOOD   | EARNED                 | UNLUCKY / THEY MADE A  |
                 | keep it; expect the    | PLAY -- keep it; don't |
                 | defense to answer it   | let the scoreboard     |
                 | (Anchor C)             | change your mind       |
                 +------------------------+------------------------+
  PROCESS BAD    | LUCKY                  | DESERVED               |
                 | fix it anyway; it      | fix the part that      |
                 | won't work twice       | failed: call, rule, or |
                 | (the gift you didn't   | player (C / R / X)     |
                 | earn)                  | (the Mills plan)       |
                 +------------------------+------------------------+

📊 Diagram (described): A two-by-two grid. Columns are "result good" and "result bad"; rows are "process good" and "process bad." Good process with a good result is "earned": keep it, and expect the defense to answer it, with Anchor C as the example. Good process with a bad result is "unlucky, or they made a play": keep it, and don't let the scoreboard change your mind. Bad process with a good result is "lucky": fix it anyway, because it won't work twice. Bad process with a bad result is "deserved": fix the part that failed, the call, the rule, or the player, with the Mills plan as the example.

The broadcast lives in two boxes: top left and bottom right. Every call that works was smart, and every call that fails was dumb. The two boxes that teach you the most are the other two. "Unlucky" is where good coordinators get fired by people who result. "Lucky" is where bad habits hide until the playoffs.

🔍 Why this works: Separating the grades isn't about being fair to coordinators. It's about learning. A result grade tells you what happened once. A process grade tells you what will happen the next time the same situation comes up, because the process is the part that repeats. If you grade only results, you'll copy lucky calls and abandon good ones, and you'll learn the wrong lesson from almost every close game you watch. That is exactly the mistake Curtis Bale made about his own offense: his first season's explosive plays told him his process was right, and nothing in his grading could tell him it wasn't until the defenses changed and the results did too.

⚠️ Common mistake: grading the call by the replay. The replay shows you the open man the quarterback didn't see, from an angle the quarterback never had, at a speed he didn't get. The fix: grade the decision from the freeze-frame at the moment of the decision (freeze-frame 3), and grade the call from the pause at the snap. Use the replay only to find out what happened, never to decide what should have happened.

📋 Try it — Whose fault?: Game 3, second quarter, 3rd-and-8 at our 31. Austin has detached Harlan from the line, so it's 11 personnel with no tight end attached. Pre-snap, the Sentinels show two high. Late in the play clock, Holloway walks down toward the line on the side of Powers's slot. Vance reads the late walk as pressure and throws hot, immediately, to Powers on a quick out: our rule says the hot throw goes to the side the late defender walked from. The ball is on time and on target. Holloway, who walked down and then didn't rush, drives on it and breaks it up. Punt. Whose fault?

Worked answer

Nobody's on our side, and you have to be able to say that. The rule was right (hot to the side the late defender came from), Vance followed it, and the ball was on time. Rome's note on that snap was "Right rule, right ball. They made a play." This one goes in the "unlucky / they made a play" box: good process, bad result. Credit Holloway and Russ. The walk-down was a lie built to beat our rule, and a good defense will win some of those.

What you'd note for later is not a fix but a question: is our hot rule becoming readable? If the same walk-and-sit beats the same hot throw again, the rule needs a branch. That's exactly how the Week 15 sack turned into a protection change. One snap is information; two of the same is a pattern. And notice what you couldn't tell from outside: whether Vance had a better option. On the broadcast you'd never know what the backside routes were. Write "?" and move on.


40.8 The Four Anchors, Graded One Last Time

Four plays have walked through this whole book. You've seen each one diagrammed, argued about, and retold from both sidelines. Here they are one last time, graded the way I'd want you to grade any coordinator: process first, then result, then how long it stayed good, which we call its shelf life.

Anchor A — "The RPO the linebacker can't win"

Game 1, the first two snaps: "Eleven, Gun Doubles Right, Inside Slant." Ames triggered on snap 1: Powers +17 on the slant. Ames sat on snap 2: Ruffin +7 into the box he'd left. The drive went seven plays, 68 yards, and Vance kept a zone read for the touchdown. 7–0.

Process: A. The conflict was visible pre-snap (six on six, one of their six with two jobs), the read key was a young linebacker who'd triggered on 14 of 18 zone flows on Teo's tape, and the call couldn't lose on either answer.

Result: A. Two snaps, 24 yards, a touchdown drive.

Shelf life: short, and I didn't manage it well. Inside Slant averaged a little over 6 yards a call before Game 2. In Game 2 it was four calls for about 13 yards. By Game 3, Ames was a much better player. He'd stopped triggering on the first flow. I kept the call in the openers longer than its evidence said I should. That's my recurring mistake, in its mildest form.

Final grade: A for the call, B for my management of it. A conflict has a half-life. The better the defender, the shorter it is.

Anchor B — "Smash vs. Cover 2"

Week 10 at Portland, 2nd-and-6 at the Portland 31, tied 7–7: "Eleven, Gun Doubles Right, 60 Left, Smash." Josh Kiel squatted on Sutton's hitch; Powers caught the corner route over him for a 31-yard touchdown. On the next Smash in the third quarter, Kiel sank, and Sutton's hitch went for 8.

PRE-SNAP (Anchor B): "Eleven, Gun Doubles Right, 60 Left, Smash"
Week 10 at Portland: 2nd-and-6 at the POR 31, tied 7-7

            S                                  S   <- two high,
                                                      on the hashes
  (C)                                             C  <- corners
             *        W        M                     rolled up, out-
                  E      T  T      E                 side leverage
   X        H     O    O    O    O    O   Y        Z
                            Q  R

📊 Diagram (described): Austin is in two-by-two: Sutton (X) wide left and Powers (H) in the left slot; Harlan (Y) attached right and Gaines (Z) wide right. Ruffin is to Vance's right. Portland shows a two-high shell with both safeties on the hashes, the corners rolled up near the line with outside leverage, a nickel over Powers, two linebackers, and four down linemen. The left corner, Kiel, is marked in parentheses as the defender the play reads: he is the Cover 2 flat defender on the two-receiver side where Smash is called.

POST-SNAP (Anchor B): Kiel sits on the hitch; the ball goes over him

  ^ corner route (~18)    S  <-- Price: held on the hash by
   \  .                         Harlan's seam
    \   .                                         ^ seam
     |    .    <-- the throw: over Kiel,          |
     |      .      under Price, 31-yd TD          |
     |        .                                   |
  ^  |          .
  |  |            .
  6  |              .
 (C) <-- Kiel squats on the hitch
  |  |                .
   X H        O    O    O    O    O   Y        Z
                           Q

📊 Diagram (described): After the snap, Sutton runs a six-yard hitch on the far left and Kiel, the left corner, comes forward to sit on it. Powers, from the left slot, stems straight upfield to about twelve yards and breaks toward the left pylon, his corner route ending at about eighteen yards. A dotted line shows Vance's throw going over Kiel and in front of the left safety, Price, who is held near the hash because Harlan's seam up the middle threatens the area between the safeties. The result is a 31-yard touchdown.

Process: A. Smash was the right concept for the right coverage. Teo's Portland chart had Kiel squatting on 17 of 21 smash and hitch looks, and the concept was built to put exactly that corner in a high-low bind. It also had a correct answer to the other choice, which is why the third-quarter hitch was a completion and not a sack.

Result: A. A touchdown on the first call and a completion on the second.

Shelf life: long, against the right coverage. Smash beats Cover 2 because Cover 2's corner owns the flat. Against Cover 3 the corner bails to the deep third, the corner route is covered, and the concept degrades to "take the hitch." That's what it did against St. Louis in Week 12: a 7-yard hitch, not a touchdown. A concept that degrades gracefully against the wrong coverage is still a good call.

Final grade: A. The cleanest call in the book. But Leon Marchetti's discipline is what made it predictable. A less disciplined corner, one who doesn't squat as reliably, makes the read harder.

Anchor C — "The third-and-seven decision"

Game 3, the opening drive of the third quarter: 3rd-and-7 at our 41, down 10–7, about 11:30 left in the third. The Week 15 sack on the same down and distance was the history. The call was "Eleven, Gun Doubles Left, Y Off, 60 Right, Mills," the same formation picture as the sack but a six-man protection rebuilt for the mug, with Gaines's sight-adjust to a hitch if the corner blitzed. Russ called the mug corner fire zone, its only call of the game. Zero free rushers, Gaines hitch +10, first down. Field goal at the end of the drive, 10–10.

ANCHOR C (Game 3, 3rd-and-7, AUS 41): count the rushers first

PRE-SNAP: one high (Royce in the middle); Bell and Ames walked
up in the A gaps; Simmons pressed on Gaines

                        F
                                              $    <- climbs to
   C                                               C  the field 1/3
                 E    T  W   M  T    E                at the snap
   X               O   O   O   O   O               Z
               Y       R   Q                 H

POST-SNAP, who takes whom:
   CENTER  --> Bell (M)      the new rule: center takes the Mike
   RUFFIN  --> Ames (W)      the back takes the other A-gap backer
   RT      --> Simmons (C)   widens to the corner blitzer
   LT, LG, RG                big on big
   GAINES  --> hitch at 6    sight adjust: corner blitz = ball now
   Free rushers: 0.  Gaines +10, first down at the SAC 49.

📊 Diagram (described): Before the snap, Austin has Sutton (X) wide left with Harlan (Y) flexed a step off the line just outside the left tackle; Powers (H) in the right slot and Gaines (Z) wide right. Ruffin is on Vance's left. The Sentinels show one high safety, Royce, in the middle; Holloway is at about eight yards on the right, ready to climb to the deep third on the field side. The Will, Ames, and the Mike, Bell, stand in the two A gaps on either side of the center, and Simmons, the right corner, is pressed on Gaines. Below the picture, a list shows the protection after the snap: the center blocks Bell, Ruffin blocks Ames, the right tackle widens to Simmons as he blitzes, and the other three linemen block the men in front of them. Gaines converts his route to a six-yard hitch because the corner blitzed. No rusher is free, and Gaines gains ten yards for a first down.

Process: A. This is the one I'd hang on the wall. The protection rule was rebuilt after the Week 15 sack, repped four ways on the divisional-week Thursday, and called into exactly the look it was built for. It had an answer to the mug and an answer to two-high. Marisol's grade for it was "earned."

Result: A. First down. The drive stalled at the 24 and we kicked the field goal.

Shelf life: one game, and that's fine. Russ called the mug once all day and never came back to it. A good answer to a pressure usually ends the pressure. That's a success, even though it means you only get to use the answer once.

Final grade: A. The best call I made all season. It got us three points. If you want to understand why coordinators sound strange when they talk about their best calls, start there.

Anchor D — "The four-play sequence"

Game 1, the fourth quarter, down 27–20. Four snaps from the same look: inside zone +5, inside zone +7, "Inside PA, Gaines Post" for 35 to the 4, inside zone for the touchdown. 27–27 with 7:31 left. (We won it later on the zero-bin deep over, 34–27.)

Process: A. Two receipts, then the play-action, then the run that the play-action made easier. Same formation, same backfield, same first step by every lineman: illusion and sequence in four snaps. The protection was seven men because the play needed time, and it got time.

Result: A. A tying touchdown drive.

Shelf life: that afternoon. Here's the honest part. Holloway's aggressiveness on that drive came partly from the 45-yard touchdown run in the second quarter, when he took a bad angle and had been coming downhill harder ever since. We didn't create that. We noticed it, and Teo charted it ("four for four, down"), and we used it. A sequence is only as good as the defender's memory. You can design the sequence, but you can't design the bad angle.

Final grade: A. But know what you were given. The sequence was ours; the bad angle was a gift.

And the snap that isn't an anchor

You can't grade our season honestly without the snap that ended it. Game 3, 3rd-and-9 at our 38, 2:41 left, down 20–17: Mills. Two high at thirteen yards before the snap; Holloway down and Royce to the middle after it. Interception.

The call: B. It was a defensible concept for the look the evidence predicted. The plan: D. One answer, a blank contingency row, no rotation on Thursday, and a margin note from Rome that I answered with three words instead of a branch. The quarterback's decision: the throw to the pre-snap answer. Darius owned that half on the sideline with three words of his own. The result: F. Season over.

Snap Process Result Shelf life What it taught
A — Inside Slant (Game 1) A A Short: the Will got better A conflict has a half-life
B — Smash vs. Cover 2 (Week 10) A A Long vs. Cover 2; degrades to the hitch vs. Cover 3 Match the concept to the coverage
C — 3rd-and-7, the mug (Game 3) A A One game: Russ stopped calling it Fix the rule, not the play
D — the four plays (Game 1) A A That afternoon Sequence is coordination; know what was a gift
Mills (Game 3) Call B, plan D F None Never let a tendency outlive its evidence

Look at the table one more time. Four A's and a D, and the D is the one everyone remembers. That's coaching. It's also why this chapter exists. If you graded our season by the last snap, you'd learn nothing true about the other four, and nothing useful about the fifth.

🔄 Check your understanding: Anchors C and Mills were the same concept, called by the same coordinator against the same defense, about ninety minutes apart. Give the one sentence that explains why one gets an A for process and the other a D.

Answer

Anchor C's call had a practiced answer for both of the defense's likely looks (the mug and two-high), while the Mills call had an answer for only one (two-high holding), even though the defense had a second look available and our own quarterbacks coach had asked for a branch in writing. Same routes, different distributions, different preparation.


40.9 The Ten-Snap Self-Test

Here's your final exam. Ten snaps, described the way you'd see them on film. Nine come from our season. One is a hypothetical Sunday snap, because the point of this book is that you'll be watching other people's offenses from now on.

For each snap, answer three things before you open the answer:

  • (a) The scheme: the family and the concept.
  • (b) The key: the conflict defender, the read defender, or the matchup the call attacks.
  • (c) The grade: process and result separately, and one thing you can't know from the film.

Score one point for each part you get substantially right. That's 30 points. 27 or more: you're watching like a coordinator. 21–26: you're watching like a quality control coach in his first season, which is a compliment. 15–20: keep charting; go back to 40.3 and 40.7. Below 15: reread 40.2 through 40.4 and take the test again next week. It isn't going anywhere.

Snap 1. Game 1, second quarter, Austin up 10–7, 2nd-and-6 at the Sacramento 45. Austin is in 11 personnel, Doubles Right, Harlan attached right, Ruffin offset left. Sacramento is in nickel with two high safeties. At the snap, all five Austin linemen step right; Ruffin aims at the right guard. The end on the tight end's side squeezes hard inside. The Will linebacker fills the wrong gap. Ruffin bounces the run outside. The strong safety, Holloway, takes a flat angle, and Ruffin runs past him for a 45-yard touchdown.

Answer

(a) Inside zone to the tight end side. (b) No single conflict defender. This is a plain run, and the back reads the blocks. Three defenders made it explosive: the end who over-squeezed, the Will who filled the wrong gap, and the safety's flat angle. (c) Process: B+, a sound early-down run from our base look, the run that buys credibility. Result: A+. But be honest about where the yards came from. Most of the 45 were the defense's mistakes, not the call's genius. Can't know from the film: whether the Will's gap was his assignment error or a stunt he was told to run. And the part you'd only learn later is that this snap changed Holloway. He came downhill harder for the rest of the day, which is what made Anchor D possible two quarters later.

Snap 2. Week 8 vs. Birmingham, third quarter, 6:40, Forge leads 17–14. 3rd-and-6 at the Birmingham 37. Austin is in 11 personnel. The call came in over the radio as "Blue 38." With under ten seconds on the play clock, the Forge walks both safeties down. There's no one deep, and every receiver is pressed or has a defender over him.

PRE-SNAP (Snap 2): the Forge walk down inside :10
(alignment simplified)

         --- nobody deep ---

   C           *     $    M     W    F               C
                   E    T    N    T    E
   X           H     O    O    O    O    O   Y        Z
                               Q  R

📊 Diagram (described): Austin is in a 2x2 formation with one back. The Birmingham defense has no deep safety: both safeties have walked down to the line of scrimmage area with the linebackers, making a crowd of defenders near the line. Both corners are pressed on the outside receivers, and a nickel is over the slot. A note at the top says "nobody deep." The diagram simplifies the exact alignment; what matters is the empty middle of the field.

Answer

(a) Austin's zero bin. Vance identified the Mike ("Fifty's the Mike!"), then changed the call to "Blue five!", our Slants. Powers on the slant, +14. (b) The shell is the whole story: Cover 0, no deep safety, man coverage across the board, more rushers than we can block. The slant beats press man inside, with nobody in the middle to help, and the ball comes out in about a second, before the extra rusher arrives. (c) Process: A. The answer was built Tuesday and given to the quarterback, who can go to the zero bin without permission. Result: A. Four plays later, Ruffin scored, 21–17. Can't know from the broadcast: whether the change came from the sideline or the quarterback. You'd only see Vance yelling. In our system it was his call to make, and he made it.

Snap 3. Week 10 at Portland, fourth quarter, 4:58, Austin up 24–13. 3rd-and-12 at the Portland 28. Austin: "Eleven, Gun Trips Right, Middle." Portland's defense on third-and-long has been Tampa 2 or Cover 2 on 14 of its last 17 snaps of third-and-10-plus. At the snap the Mike linebacker turns and runs to the deep middle. Ruffin catches the screen in the space he left and gains 9, to the Portland 19. Austin kicks a 36-yard field goal: 27–13.

Answer

(a) A middle screen to the running back. (b) The Tampa 2 Mike. His job on third-and-long is to run to the deep middle, and the screen goes exactly where he left. (c) By success rate it's a failure: 9 yards on 3rd-and-12. By process it's an A. Think about the situation. Up 11 with under five minutes left, already in range, the job was to make the kick shorter and give the defense no sack and no turnover. It turned a 45-yard try into a 36-yarder and made it a 14-point game. This is the opposite of what I once called a "give-up" screen, "a punt with extra steps." This was a field goal with extra yards. Can't know: how much Dorsey's leg and the wind figured into it. Kick range is a sideline fact, not a film fact.

Snap 4. Week 12 vs. St. Louis, second quarter, 9:14, tied 7–7. 2nd-and-8 at the St. Louis 44. "Eleven, Gun Doubles Right, Zip, 60 Left, Smash, X Fade, Alert Post." Gaines zips in toward the formation and nobody follows him. The Stallions show one high safety. At the snap, the corner on Gaines's side bails deep. Vance throws Gaines's hitch: +7.

Answer

(a) Smash to the motion side, with the backside X on a fade. (b) The corner on the Smash side. Against Cover 3 he isn't a flat defender, so there's no high-low conflict to win. He bails to the deep third, so the corner route is covered and the hitch is the throw. The Zip motion told us "zone" before the snap (nobody traveled). (c) Process: B+. Smash isn't the ideal Cover 3 concept, but it had a built-in answer for Cover 3 and the motion confirmed the coverage. Result: success (+7 on 2nd-and-8 clears the 4.8-yard line), and Odom converted the 3rd-and-1 that followed. Can't know: whether the "Alert Post" was alive, meaning whether Vance looked at it first and passed, or never had the look.

Snap 5. Game 3, first quarter. 3rd-and-3 at the Sacramento 38. Austin is in 2x2. Sacramento shows one high safety, with the field corner, Simmons, pressed on Gaines. Austin throws a slant to Gaines. Simmons's press slows Gaines's release; the slant isn't open on time. Vance holds the ball. At about 2.9 seconds, Rollins beats Achterberg and sacks Vance for −7. Punt.

Answer

(a) Quick game: a slant. (b) Simmons, and the release. Against press on 3rd-and-short, the slant lives or dies at the line. (c) Grade it in pieces. The call: C+. On a short third down against press, the box had one answer, and one answer lets the defense solve it on the first try. The execution: X. A quick-game ball held to 2.9 seconds becomes a dropback the protection wasn't built for. Rome's rule now is that at 1.5 seconds on a quick play, the ball goes to the checkdown, the dirt, or the third row. Credit Simmons, who won the rep. The fix came two quarters later: on 3rd-and-2 at the SAC 29, against the same look, Gaines came in zoom motion, started his route already moving, and caught a quick out for 5. Can't know: whether the protection had a different rule for that call. And don't blame Achterberg for losing at 2.9 on a play designed to be over at 1.3.

Snap 6. Game 3, second quarter, 2nd-and-7 at the Austin 36. "Eleven, Gun Trey Right, 60 Left, Bender." Sacramento is in its two-high Quarters look. Gaines (#1) runs a hitch at 6. Powers (#2) stems straight upfield past nine yards like a seam, then breaks out at twelve. Vance pumps toward the hitch. Simmons, the corner, comes forward toward the pump. Powers is open behind him: +14.

Answer

(a) The bender, our answer to Quarters: a vertical stem by #2 that becomes an out at 12, paired with a short route by #1. (b) The Quarters corner. His rule is that when #1 is short, his eyes belong to #2 until #2 declares. The pump asks him to break that rule, and the stem "lies" about where

2 is going. (c) Process: A, a planned test of one defender's discipline, scripted as the

thirteenth call with a branch (if the eyes leave, the bender; if they stay, the hitch). Result: A. The drive ended in Sutton's back-shoulder touchdown, 7–3. Can't know: whether Simmons was coached to jump or just did. That's the difference between a tendency and a mistake, and you'll find out the next time.

Snap 7. Same game, third quarter. 3rd-and-6 at the Sacramento 44. The same call, the bender. This time Simmons's eyes stay on Powers through the pump. Vance throws the out anyway. Simmons breaks it up. Punt.

Answer

(a) The bender again. (b) The same corner, and this time he did his job. Russ told me in the spring: "Your bender was the right answer, so we stopped letting the corner look at the quarterback." (c) Process: C for the call. I called a first-half winner after halftime without asking whether it had survived their halftime. The decision: −. The branch said hitch, and the out went anyway. Result: F. Notice the pattern: the throw went to the pre-snap answer against a post-snap change. That's the same habit that showed up at 2:41 of the fourth quarter. I heard Rome say "Eyes stayed, ball went" on the headset and didn't connect it to the third-and-long box. Can't know from outside: that there was a branch. From the couch, this looks like a good play the defense beat.

Snap 8. Game 3, fourth quarter, 12:48, Sacramento up 17–10. 1st-and-goal at the Sacramento 9. The drive has been mostly outside zone to the left from 11 personnel, and the defensive end on the right, Rollins, has chased flat down the line after every one. Now Austin comes out in 12 personnel under center: "Twelve, Under Ace Left, Wide PA Boot Right." Vance fakes outside zone left, keeps it, and rolls right. Harlan, who started on the left, crosses the formation; Brandt is in the right flat. Vance hits Harlan for a 9-yard touchdown. 17–17.

Answer

(a) A play-action bootleg off outside zone. The quarterback boots away from the run fake, and the boot-side guard pulls to protect the edge. (b) The backside end, Rollins. A boot only works if the end on the boot side chases the fake. Teo had him chasing on every outside zone of the drive. (c) Process: A. Our divisional-week plan was to run outside zone from Eleven and save Twelve for the boot, and we waited for field position where the crosser was a touchdown throw. Five runs paid for one boot; the boot is a payment for runs you've already made. Result: A. Can't know: that we'd planned it that way since Tuesday. From outside, it looks like a well-timed call. It was really five receipts and a plan.

Snap 9. Game 2, fourth quarter, about 11:00, Sacramento up 23–13. 1st-and-10 at the Austin 41. "Eleven, Gun Trips Right, 60 Right, Z Sluggo, H Seam, Y Drag." The booth has counted Simmons pressing on six of the last seven first downs, and he presses Gaines again. Gaines fakes the slant; Simmons jumps it; Gaines goes up the sideline. Powers's seam occupies Royce. Ruffin is on Rollins's side to chip. Gaines catches it at the Sacramento 25 and is run down at the 15: 44 yards.

Answer

(a) A double move (our "Sluggo," the slant-and-go) to the outside receiver in Trips, with the slot on a seam. (b) Simmons, the pressed corner. His press is a bet that the quick throw comes fast, and the double move bets the other way. Royce is the second key: the seam keeps him from helping over the top. (c) Process: A. The booth gave us a count (6 of 7), the situation needed yards in chunks, and the chip on Rollins bought the extra second a double move needs. Result: A. Three plays later, Ruffin scored from the 2, 23–20. Can't know: the booth count. A fan could have kept it, though. It's exactly the kind of thing the twelve-field chart catches.

Snap 10 (hypothetical). A Sunday this fall, two teams you'll watch on television. The visiting offense has 3rd-and-4 at the home team's 35, down 3 in the third quarter. It's in 3x1, tight end as the #3 receiver to the trips side. The defense shows two high safeties. At the snap, one safety drops down toward the trips side and the other goes to the middle. The quarterback throws a quick out to the slot receiver at about five yards, but the slot has kept running straight up the field. The ball sails out of bounds. The analyst: "The receiver ran the wrong route." Punt.

Answer

(a) Probably a quick concept with an option route for the slot, a route that changes by coverage. (b) The rotating safety. Many option-route rules turn on whether the middle of the field stays open or closes. (c) The honest grade is "?". Under one common rule, the slot breaks out against a closed middle and stays vertical against an open one. If the quarterback and the receiver read the rotation at different moments, one saw two-high and the other saw one-high. Either one could be "right" depending on the rule and its timing, and you don't know the rule. That's the point of the snap. What you can grade: the call fit the situation (a quick answer on 3rd-and-4), and the disconnect came after the snap. What you'd need to finish the grade: the route rule, and what the receiver saw. Write "QB/WR read disagreement on rotation — rule unknown" and move on. A coordinator would never accept "the receiver ran the wrong route" without the rule in front of him.

🔄 Check your understanding: Three of the ten snaps (3, 7, and 10) are traps for anyone who grades by result. What's the trap in each?

Answer

Snap 3 failed by success rate but was an excellent call for the clock and score. Snap 7 looks like "good play, defense won" from outside, but there was a branch and it wasn't taken, and the call itself was a first-half answer used without checking the halftime adjustment. Snap 10 invites a confident verdict ("wrong route") that the film can't support without the route rule.


40.10 The Honest Limits of Watching from the Outside

I promised Teo an honest answer to his question on the couch, so here it is. From the outside, with good film and a good method, you can see most of what happened and a lot of why. You can't see all of it, and the parts you can't see are often the parts that decided the play.

What you can't see Why it matters What to write instead
The call itself You don't know the concept name, the tags, or which route was the alert "Probably smash; backside unclear"
The rules Option routes, protection rules, hot rules, sight adjustments, check-with-me menus "Rule unknown," not "wrong route"
Health A tackle playing on one ankle, a quarterback who's been hit eleven times "Why did they stop calling X?" as a question
The week What was practiced, what wasn't, which rotation the scout team never showed "Looked unrehearsed?"
Communication The radio cut at 15 seconds, the long call, the missed signal "Late snap; possible communication issue"
The defense's call Was that a coverage bust or a disguise that worked? "Busted or designed? Watch next week."
The plan's other branch Was there a check that wasn't taken? Note the snap; look for the same picture again

Here's the rule I've followed for twenty-two years when I watch someone else's tape: I write "probably" in front of every assignment. "The tackle probably had the end." "The slot probably had an option." When I find out later, from a coach, from a scouting meeting, from a spring dinner with the man who called the defense, I cross out "probably" or I cross out the sentence. More of the sentences get crossed out than you'd think.

Film from the outside is a hypothesis generator. It's very good at telling you what to ask, and only fair at telling you the answer. That isn't a reason to stop watching. It's a reason to watch with a pencil instead of a verdict.

It's also a reason for a little respect. Every coordinator you'll watch this fall, including the ones whose calls you'll hate, spent twenty-plus hours that week on a plan you'll see one-sixtieth of on any given snap. They know things about their tackle's ankle and their quarterback's confidence that you never will. When you're right about their call, you're right from a distance. When you're wrong, you're usually wrong about something they knew. Grade them like you'd want your own Mills call graded: on what was knowable at 34 seconds, and with the replay turned off.

💡 Coach's point: The best film watchers I know are the quickest to write "?" Confidence about the parts you can see, humility about the parts you can't, and a clear line between them. That's the whole skill.


Spaced Review

Answer each question before you open it.

1. (Chapter 7) "The shell is not the coverage." What does that sentence mean, and which of this chapter's freeze-frames is built around it?

Answer

A two-high look before the snap (middle of the field open) can rotate to one high (middle closed) after it, and the reverse. The pre-snap shell is what the defense shows; the coverage is what it plays. Freeze-frame 2, at the quarterback's back foot, exists to catch the difference. That's where Royce's spin lived in Game 3.

2. (Chapter 25) A play call is a bet on a distribution. How does the 34-second fence apply that idea to someone else's call?

Answer

If a call is a bet, you grade the bet by what was knowable when it was placed: the situation, the likely looks, and the call's answer to each. The fence makes you write those down before the result can rewrite your judgment, so you grade the distribution, not the one outcome that came up.

3. (Chapters 19 and 38) What is the "two receipts" rule, and how does it show that an offense is a system of constraints?

Answer

Call the play-action only after two same-look runs that the defense has "paid" for by playing the run aggressively. The play-action exists because of the runs, and the next run works better because of the play-action. Each play is an answer to the defense's answer to another play. That's the constraint idea from Chapter 38, and it's why you can't grade one snap of a sequence by itself.


Project Checkpoint: Film Grading Rubric

Where you are. In Chapter 39 you assembled your Complete Playbook and Sentinels Game Plan: 60–70 plays, the weekly process, a script, a call sheet, and situational boxes. That's the offense you'd install. This last deliverable proves you can see an offense, yours or anyone's, on film.

The deliverable: a Film Grading Rubric plus a ten-snap self-test you run on a real game. Format: one page for the rubric, one page for the ten graded snaps.

The rubric. Five categories, each scored 0 (missed), 1 (partly), or 2 (complete), for 10 points a snap and 100 points for ten snaps:

Category 2 points means you…
1. Before the snap Named the situation, personnel on both sides, formation and strength, shell, box, leverage, and the motion reaction (or wrote "?" where you couldn't see)
2. The scheme and the key Named the family and concept, and the conflict or read defender
3. The decision and the winner Said what the quarterback or back decided, whether it matched the key, and which block or route decided the play
4. Why this call, now Tied the call to down, distance, field, score, clock, and the snaps before it
5. The honest grade Gave a process grade before the result grade, used C / R / X for failures, and named one thing you couldn't know

A filled-in row (from the Week 12 Smash, snap 4 above):

Before: Q2 9:14, 2-8 at STL 44, 7–7; 11 vs. nickel; Doubles R, Y attached; Z Zip, nobody traveled; 1H; box 7; corner leverage "?" (cut away before the snap). Scheme/key: Smash to the Z side, X fade backside; Smash-side corner, who bailed to his deep third. Decision/winner: threw the hitch under the bailing corner; Gaines won the hitch; the corner route was covered. Why now: 2nd-and-8, early down, middle of the field; motion confirmed zone. Grade: process B+ (Smash vs. Cover 3 is second-best, but it had the hitch answer); result success; can't know if the Alert Post was alive. Score: 2/2/2/2/2 = 10.

Choose your level:

  • High school: Run the test on your own team's game film, with your offensive staff grading the same ten snaps independently. Compare the category-5 grades. Where your staff disagrees is where your teaching isn't clear yet.
  • College: Run it on an opponent's film and add a sixth category, field or boundary, because the wide hash changes formation strength and play direction at your level.
  • Pro: Run it on All-22 of a team you play twice a year, and have your defensive staff grade category 2 and category 4 blind. If they can name your conflict defender from your call, so can the other team.
  • Film room: Use a broadcast copy of any game this fall. Pick ten snaps from one offense, including at least three third downs and one red-zone snap. Score yourself honestly, then score again after you get the coaches' film if you can. The gap between your two scores is a measure of what the broadcast hides.

What comes next. Nothing in this book. The next deliverable is the next game you watch.


Chapter Summary

The loop. Every snap, every time: before (situation, personnel, formation, shell, box, leverage, motion and reaction) → at the snap (the line's first step, the read defender, the decision, the key block) → why this call, now (down, distance, field, score, clock, last three snaps) → the grade (process first, result second, name what you can't see).

Tool Use it for Remember
Pre-snap checklist The defense's structure and the conflict, before the ball moves Count the shell before the box
Line's first step The play family in the first half-second Zone steps together; gap blocks down and pulls; pass sets back
Four freeze-frames Snap, back foot, ball out, first contact Frame 2 shows the rotation the snap hid
Three angles Broadcast for situation and result; All-22 for coverage; end zone for the line and the QB's eyes Write "?" for what your angle can't show
Twelve-field chart Turning watching into evidence Three passes; tendencies need 10+ snaps, about 70%, and a "since" date
34-second fence Grading the call, not the outcome Write the purpose and likely looks before you watch the result
Process/result grid Sorting earned, unlucky, lucky, deserved The two off-diagonal boxes teach the most
C / R / X Locating a failure: call, rule, or execution One letter per failed snap
"Probably" Every assignment you infer from outside Film is a hypothesis generator

The anchors, graded last: A (Inside Slant): A call, B management, short shelf life. B (Smash vs. Cover 2): A. C (3rd-and-7, the mug): A, the season's best call, three points. D (the four plays): A, with an honest note about the gift. Mills: call B, plan D, result F.


Closing: The Light in the QC Office

Teo's last night in the building was a Thursday in June. The FCS head coach had called him in April; he'd told the man he wanted to finish the spring with us first, and he did. He spent May playing Russ in what we called the Answer Game, answering every new call the way he thought Russ would. Only after the last spring practice did he say yes. I didn't plan for his last night to be a Thursday, but Thursday is third-down day, so it fit.

When I came down the hall at ten with the two coffees, the quality control office was cleaner than I'd ever seen it. The cut-up sheets were gone. The cords were wrapped. The whiteboard was wiped except for one line in the corner, in Teo's handwriting: shell, box, one thing. The diner placemat with LINEAGE on it was still pinned above the desk, where it had been since before he got here. So was the System Map, our whole offense on one sheet, with its one rule printed across the top: every arrow needs a play at both ends.

"You're leaving that," I said.

"It's not mine. It belongs to the office." He took his coffee. "The next guy should have to ask you what it means."

On the desk were two folders. The first was the QC hiring test, finished: tag thirty snaps on twelve fields in two hours; find two tells, each with its sample size and a "since" date; write one row of a "by situation" page in plain English a quarterback could read on a Saturday; defend it at the whiteboard against a defensive coach for fifteen minutes. Wade had read it and asked only one question, whether he could be the defensive coach for part four. Two candidates have taken it as I write this. One of them found a tell Teo hadn't. Teo was so happy about that you'd have thought he'd found it himself.

The second folder was the "Kimbrough by situation" template, blank. Eight rows, each with a "since" column and a column headed "?". On the inside cover he'd stuck a note for whoever sits in his chair next:

Every row gets a since date. Every streak gets a question mark. When a defense this good stops giving you tells, it's because it wants you to stop looking.

The last sentence was mine, from the film session four days after the divisional game. I hadn't known he'd written it down. He writes everything down. That's the job, and it turns out it was also the lesson.

We talked for a while about nothing: his apartment, the drive to the new school, whether FCS quarterbacks will understand "rule four" or whether he'll have to write his own list. (He'll write his own list. I told him to number it.) Then he asked me the question I think he'd been holding since the night we watched the broadcast copy.

"Coach. If you'd had the middle-read tag in January, does Darius throw the dig?"

"Probably."

He laughed, because he knew exactly what I meant by "probably." I can't know, and neither can you. The tag is built now. It's printed on every Blue-panel third-and-7-plus call that has a deep-middle throw in it, and the huddle word "Eyes" adds it to any other call. It has four beats: name both throws before the snap; eyes to the empty middle on step one; declare it open or closed on step three; if it's open, run the concept's read, and if it's closed, never the post: the dig, then the flat or the back. The branch is written into the call so the quarterback never has to doubt himself into it. Darius helped write it. The scout team rotates on one third-and-long rep in three now, and nobody tells the quarterback which one. "Never let a tendency outlive its evidence" is still on my whiteboard in red. I'll find out in the fall whether any of that is enough, and so will Russ, who is certainly building something to beat it right now. That's the conversation. It doesn't end; it just changes seasons.

I'm still here. Wade's still here; he was retained the week after the loss, and my napkin is still on his desk. Marisol has already sent me the first draft of next season's self-scout, with sample sizes and caveats. Rome has a new margin, and I've promised to answer whatever he writes in it with more than three words. Gus still grumbles about RPO linemen. Darius threw in the indoor facility three mornings a week this spring, and every third-and-long rep he finished the same way: eyes to the middle, then the ball he knew. Somewhere in the second week, he stopped saying "Rule four, Coach" after the checkdowns. He doesn't need to say it anymore. It's in the call.

Teo left a little after eleven. At the door he turned around, and I could see him decide not to make a speech, which I appreciated, because I'd have laughed before he finished it.

"Shell, box, one thing," he said.

"Shell, box, one thing."

Then he was gone, and the light in the QC office was off for the first time in two years.


That's the book. Here's what you can do now that you couldn't when you opened it.

When the offense breaks the huddle, you'll count. Personnel, formation, strength, the shell, the box, the leverage, and when someone moves, the answer. You'll see the conflict before the snap, the linebacker with two jobs or the corner with two receivers, and you'll know where the ball should go if he chooses wrong. At the snap you'll watch the line's first step and know the family before the ball moves. You'll find the defender the play is built to fool. You'll ask why this call, now, after those three, and more often than not you'll have an answer. And when the play fails, and the stadium groans, and the analyst says you can't throw that ball, you'll do the thing almost nobody watching will do. You'll freeze it at the snap, write down what was knowable, and grade the bet before you grade the outcome.

You'll be wrong sometimes. Write "probably." Cross it out when you learn more.

The offense has one structural advantage: it knows where the ball is going, and the defense doesn't. Everything in this book was a way to make that advantage bigger: conflict, sequence, illusion, and eleven people moving on the same count. Now you get a smaller version of the same advantage. You know what to look for, and almost everyone around you is watching the ball.

Every play is a question. Go find out what it's asking.