The whistle blows on Jalen Ruffin at our own 41, and the play clock in the far end zone flips to 40.
In This Chapter
- The Hook: Forty Seconds at the 41
- 25.1 What a Play Call Is
- 25.2 The Forty Seconds
- 25.3 The Inputs: Situation Before Scheme
- 25.4 The Third-and-Seven Decision
- 25.5 A Call Is a Bet
- 25.6 The Other Coordinator
- 25.7 Sequencing: Calls as Investments
- 25.8 The Headset, the Booth, and the Head Coach
- 25.9 Failure, Tilt, and the Next Call
- 25.10 Art and Science
- Spaced Review
- Project Checkpoint: Call Sheet v1 + Play-Caller's Pregame Checklist
- Chapter Summary
- What's Next
Chapter 25 — Play Calling: Art, Science, and the 40-Second Decision Under Pressure
The Hook: Forty Seconds at the 41
The whistle blows on Jalen Ruffin at our own 41, and the play clock in the far end zone flips to 40.
Divisional round in Sacramento. Opening drive of the third quarter. We're down 10–7. We took the second-half kickoff to our 23, and Ruffin on Wide Right and a Devonte Sutton curl got us a first down at the 38. Then Wide Right for two and Inside Slant for one. Jordan Ames sat, so Darius Vance handed it off, and Marcus Bell met Ruffin a yard past the line. Third-and-7. About 12:06 on the game clock.
I don't look at the field. I already know what's on it. I look at the laminated sheet in my left hand, at the box near the bottom that says 3rd LONG, and I wait for one voice.
Mateo "Teo" Arriaga is upstairs in the coaches' booth, and between the whistle and my call he's the only person allowed on my headset. He gives me three things in order: shell, box, one thing.
Teo: "Two high. Six. Nickel's staying on."
The nickel was staying on. Sacramento wasn't subbing a dime. That meant Bell and Ames, the two linebackers they used for their double-A mug, were still on the field. On Teo's page the mug row has three words: Can't call it.
Thirty-six.
Two calls were already in my head before the ball was spotted. One was for "Ruffin gets the first down," and he hadn't. The other was for third-and-long, and I'd been carrying it since Tuesday. It wasn't the call the page pointed at. The page pointed at our quiet-front call, the one we'd repped all week against two high safeties. Twenty-nine straight times since their bye, Sacramento had stayed two-high on third-and-seven-plus against two-by-two.
Thirty-two.
I could still see the other third-and-7. Week 15, this same stadium, our own 38, tied 13–13. Same personnel, same two-by-two, Brooks Harlan flexed off the left tackle into the boundary. They walked both linebackers into the A-gaps. Their end walked out at ten seconds, after my radio was dead. A 300-pound defensive tackle dropped into the window where our hot throw lived. Minus nine. Punt. Gus Ferrante ran that clip nineteen times for the offensive line.
Twenty-nine. I keyed the radio.
"Eleven. Gun Doubles Left, Y Off. Sixty Right. Mills."
It was the Week 15 formation. It was the Week 15 down and distance, give or take three yards. And it carried a protection that had never been on the field in a game: center on the Mike, Ruffin on the other A-gap linebacker, right tackle widens to the corner. Trey Gaines knew that if his corner came, the route became a six-yard hitch.
Twenty-two. Vance repeated it into the huddle once. Aaron Szabo didn't ask for it again.
Fifteen. The radio went dead. From here on, everything I'd say for this snap was already written in the rules.
Eleven. Bell and Ames walked into the A-gaps. Holloway started creeping.
Ten. Grant Weller stood up and walked across to the boundary, outside DeShawn Rollins, exactly the way he had in Week 15.
In Week 15, that walk moved our back. This time nobody moved. Vance pointed at Bell: "Fifty-four's the Mike!" Szabo's head turned to Bell and stayed there. Ruffin set his feet in the left A-gap, across from Ames. Darnell Fitch, our right tackle, eased his outside foot back an inch.
Four. Snap.
I'll tell you what happened. But not yet.
First I want you to make this call yourself, with everything I knew at 12:06 and nothing I learned after. The rest of this chapter is about that decision: the forty seconds, the call sheet, the headset, the other coordinator, the bets. And it's about the hardest lesson in my job, which is that the result of this snap can't tell you whether I called it well.
🏃 Fast Track: If you've called plays, skim 25.1–25.3 for our vocabulary (the call sheet, the forty-second timeline, the quiet window), then read 25.4–25.5 closely. That's the full third-and-seven decision and the threshold concept that changes how you grade every call you make. Sections 25.8–25.9 on the headset, the head coach, and tilt are where experienced coordinators usually find something to argue with.
🔬 Deep Dive: The call-sheet and game-plan templates are in Appendix F. Expected points and success rate, the numbers behind the "out of ten" estimates in 25.5, are in Appendix H, and Chapter 28 argues about them in full. The protection math behind the call is in Chapter 22.
📺 Film Watcher: This Sunday, pick one third-and-long in each half and time it. Note when the huddle breaks, when the defense shows its front, and when anyone walks up. Then ask what the coordinator knew when he sent the call in, which was usually before the front you're looking at existed.
25.1 What a Play Call Is
A play call is a choice from a menu, made under a clock, by one person, with incomplete information. Each piece of that sentence matters.
A choice from a menu. Nobody invents a play on Sunday. In Chapter 1 we built a ladder: philosophy → system → playbook → game plan → call sheet → call. By game day, every rung above the last one is finished. The playbook holds everything we own, several hundred plays if you count formation variations. The game plan holds what we chose for this opponent: 112 calls against Sacramento in the playoff, sorted by situation. The call sheet is that plan printed so it can be read in about four seconds, usually a laminated page or two that the play caller holds on the sideline. The call is one line from it.
Under a clock. Forty seconds from the whistle to the snap on most plays (25 seconds after certain administrative stoppages), and the radio to the quarterback's helmet cuts off at 15. The real window for choosing is about six seconds. We'll take it apart in 25.2.
By one person. On our staff that's me. Wade Harmon, our head coach, owns fourth downs and timeouts. Teo owns the booth's report. Gus owns the line's word, Rome Pettaway owns the quarterback's head, and Marisol Vega owns the numbers. The call is mine. A staff that votes on third down is a staff that snaps the ball late.
With incomplete information. This is the piece people outside the job forget. When I call the play, the defense usually hasn't lined up. I know their personnel, because they've subbed or they haven't. I know their tendencies, because Teo has tagged every snap for weeks. I know what happened on the last play. I do not know the front, the shell, or the pressure, because in most cases the defense hasn't shown them yet, and when it does they may be lies. Most of the time the call goes in before the picture you see on television exists.
That last fact shapes everything else in this chapter. If you call plays as though you can see the defense, you'll call the right play against the wrong defense all day. You have to call plays that are good against the defenses you might get, weighted by how likely each one is. We'll put a name to that idea in 25.5.
The call sheet
Here's the shape of our call sheet. The real one is two sides of an 11-by-17 card, laminated, with the plays in 8-point type.
CALL SHEET (front side) -- Sacramento, divisional round
+-----------------+-----------------+------------------+---------------+
| OPENERS 1-15 | BASE RUNS | RPO / KILL | SHOTS |
| (script, with | by front: | Inside Slant | free down / |
| if A / if B) | Over / Under / | Inside Glance | sudden change |
| | Bear | Wide Now | confirm x2 |
+-----------------+-----------------+------------------+---------------+
| EARLY-DOWN PASS by shell: | PLAY-ACTION | SCREENS |
| two-high / one-high / press | (married to the | RB / H / Y |
| | run above it) | |
+--------------+--------------+-------------------+-------------------+
| 3rd SHORT | 3rd MEDIUM | 3rd LONG | 3rd LONG-LONG |
| 1-3 | 4-6 | 7-10 | 11+ |
| | vs. 3x1 | QUIET front | draw / slip / |
| | vs. 2x2 | MUG | one dagger |
+--------------+--------------+-------------------+-------------------+
| ZERO BIN (Blue, top left) | SERIES CARDS: run card | Quarters card |
+----------------------------+------------------------+---------------+
📊 Diagram (described): The front of our call sheet is a grid of boxes. The top row holds the 15-play opening script, base runs sorted by the defensive front, RPO and Kill calls, and the shot plays with their trigger rules. The second row holds early-down passes sorted by the coverage shell (two-high, one-high, press), play-action calls each listed under the run it imitates, and screens. The third row is third down, split by distance into short, medium, long, and long-long; medium is split by formation, and long is split into "quiet front" and "mug." The bottom row holds the zero bin (the plays that beat Cover 0) and the series cards that chain calls together.
The back side holds the red zone (fringe, high red, low red, goal line), two-minute, four-minute, backed up, two-point, and a strip down the right edge we call the result column, where I make one pencil mark per series. More on that mark in 25.9.
Three rules shape the sheet more than any box does.
- Two answers per likely look. A box with one call is a box the defense solves on the first try. (Chapter 24 built the boxes; the sheet just has to show both answers side by side.)
- Sorted by what I'll know when I call it. Down and distance, field zone, and personnel come first, because I know those at 34 seconds. Coverage comes second, because often I'm guessing it. A sheet sorted only by coverage asks me to know something I usually don't know yet.
- Words, not numbers, except on the wristband. The sheet says "Mills." Vance's wristband says "Blue 23." The sheet is for my eyes; the band is for his.
💡 Coach's point: The call sheet isn't a menu of your favorite plays. It's a list of answers you already agreed on, arranged in the order you'll need them. The sheet makes you fast enough to think during the six seconds that matter.
🔄 Check your understanding: Why does our call sheet sort third down by distance and formation first, and coverage second?
Answer
Because at the moment of the call, about 34 to 28 seconds on the play clock, the caller knows down, distance, field position, score, and the defense's personnel for certain, but usually doesn't know the coverage. The defense often hasn't aligned yet, and its pre-snap shell may be a disguise. A sheet sorted by things you know lets you find the box instantly; inside the box, the calls are chosen to be good against the coverages you might get.
25.2 The Forty Seconds
Every play caller has a routine for the forty seconds whether he knows it or not. Ours is written down, because a routine you can't say out loud is one you'll abandon the first time you're angry.
THE 40 SECONDS (Austin's normal snap; play clock in seconds)
40 34 28 20 15 5 0
|-------|-------|-----------|-------|----------------|-----|
result call & into the huddle Mike ID, motion, snap
+ Teo's person- helmet break walk-ups, checks target
read nel ("Blue 23") (radio (rules only) :06-:03
cuts at 15)
^^^^^^^^^^^^^^^ ^^^^^^^^^^^^^^^^^^^^^^
my window Vance's window
📊 Diagram (described): A horizontal timeline of the play clock from 40 down to 0. From 40 to 34 the result comes in and Teo gives his booth read. From 34 to 28 I choose the call and the personnel. From 28 to 20 the call goes into Vance's helmet, either in words or as a wristband number. From 20 to 15 the huddle breaks, and the radio cuts off at 15. From 15 to about 5 Vance identifies the Mike linebacker and the offense runs its motion and handles any walk-ups, using rules only. The snap comes with about 6 to 3 seconds left. My window is the first twelve seconds; Vance's window is the last fifteen.
Read the timeline as two windows. The first twelve seconds belong to me. The last fifteen belong to Vance and the rules. In between, the call travels through the radio and the huddle, and nothing can be added to it.
40–34: the result and the read. The ball is spotted, the chains move or they don't, and Teo speaks. That's the quiet window: between the whistle and the call, only Teo talks on my headset, and he says three things in this order: shell, box, one thing. "Two high, six, Will at five." "One high, seven, Vo's walking." The shell is what the safeties showed on the last snap. The box is how many defenders were within about five yards of the ball. The one thing is whatever Teo thinks I need for the next call, and choosing it is the hardest job in the booth. Wade and Marisol have their own channel. Wade comes onto mine only to give me a decision, never to make conversation.
34–28: the call. If I've done my job, this takes two seconds, because I decided before the ball was spotted. I keep two calls ready after every snap, one for "we gained enough" and one for "we didn't." The whistle tells me which one to use. Teo's one thing can change it. Personnel goes out with the call. If we're changing personnel, the signal goes to the sideline at the same moment.
28–20: into the helmet. The call goes into Vance's helmet in our grammar: personnel, formation, motion, protection, concept, tags. "More than ten words, call the band." If the call runs longer than ten words, I say the wristband number instead ("Blue 23") and Vance reads it off his wrist. Long calls are how radios cut off mid-sentence. Ours did once, in Week 2 at St. Louis, and it cost us a timeout.
20–15: the huddle. Vance says the call once, formation through tags, plus the snap count. Only Szabo may ask for a repeat. The huddle breaks at about 15, which is when my radio goes dead.
15–5: the rules. Here the play stops being mine. Vance and Szabo identify the Mike. Motion goes. The defense walks up, bails, or spins. Everything that happens now is handled by rules we installed during the week: the protection card, the hot rules, the sight adjustments, the Kill and Check words. In Chapter 22 I put it this way, and I'll repeat it because it's the most important sentence on this page: your protection rules are your voice after fifteen seconds.
The snap. Our target is :06 to :03. Snapping late doesn't waste the clock for its own sake. It makes the defense show its hand before we commit, and it gives Vance time to apply the rules.
🔍 Why this works: The defense has its own forty seconds, and its call usually goes in after ours. A defensive coordinator sees our personnel before he calls his, so he has a small information edge on the call. We take it back after the call. Formation, motion, and the late snap make the defense declare before we have to. A routine that decides early, gets the call in early, and snaps late isn't about speed. It moves our decision to where we're blind and holds our execution until they're not.
Two timelines, one down and distance
Here are the two third-and-sevens side by side, as close as Teo's booth clock and the film let us reconstruct them.
| Play clock | Week 15, 3rd-and-7, AUS 38, 13–13 | Divisional, 3rd-and-7, AUS 41, SAC 10–7 |
|---|---|---|
| 40 | Whistle on Ruffin's +3 | Whistle on Ruffin's +1 |
| ~37 | Teo: shell, box, one thing | Teo: "Two high. Six. Nickel's staying on." |
| ~29–27 | Call in at 27 | Call in at 29 |
| ~15 | Radio cuts | Radio cuts |
| ~11 | Mug shows; Vance: "Fifty-four's the Mike!" | Mug shows; Vance: "Fifty-four's the Mike!" |
| ~10 | Weller walks to the boundary; Vance moves Ruffin out to him | Weller walks to the boundary; nobody moves |
| Snap | About 3 seconds left | About 4 seconds left |
| After the snap | Coleman drops into the hot window; Bell free; sack at 1.9 s, −9 | Coleman drops into the hot window; nobody free; Gaines hitch, +10 |
Most of the two rows match. The defense did the same thing in both games, and the call reached Vance on time in both. The difference sits in the row at ten seconds. In Week 15 our rules let a late walk move a blocker. In the playoff, our rules let a late walk move only the throw. Nothing I said on the radio could have made that difference at ten seconds. It had to be built into the rules on Tuesday.
🎙️ On the headset: Week 15, at ten seconds, I wanted to key the radio and say "Keep Jalen on the Mike." I'd have been talking to a dead radio. That's the moment I understood that my real call on third-and-long isn't the concept I say at 29. It's the rule sheet I hand Vance on Saturday, which is still working when I can't talk to him anymore.
⚠️ Common mistake: deciding at the spot. A young caller waits for the ball to be spotted, looks at the down and distance, and then starts reading his sheet. He's using the seconds he needs to get the call in to choose it. The call arrives at 22, the huddle breaks at 12, and the offense snaps at :01 with no time to apply a rule, or takes a delay-of-game penalty. The fix: two calls ready before the whistle, one for each likely result. The spot only tells you which one.
🔄 Check your understanding: In the two-timeline table, which row explains the different results, and why couldn't the play caller change it during the game?
Answer
The row at about ten seconds, when Weller walked out to the boundary. In Week 15 our rule told Vance to move Ruffin to the walker, which left Bell free. In the playoff the rule kept every blocker on his man; the walker was an extra, not a reassignment, and the quick throw was Gaines's sight-adjust hitch behind the blitzing corner. The caller couldn't change it because the radio to the quarterback cuts off at 15 seconds on the play clock. Anything after 15 has to be handled by rules installed before the game.
25.3 The Inputs: Situation Before Scheme
Before I can think about Russ Kimbrough, I need to know where I am. Every call starts with the same six facts, in the same order. I don't skip any of them, because the one you skip turns out to be the one that decided the play.
| Input | The question it answers | How it changes the call |
|---|---|---|
| Down and distance | Which box? | Third-and-7 and second-and-7 aren't the same down. On third down, short of the sticks means the drive is over. |
| Field position | What does failure cost, and what does success buy? | At our own 41, a failure means a punt. At their 35, it's Evan Dorsey's range. At their 8, it's points or nothing. |
| Score | How much variance do I want? | Behind late: more variance. Ahead late: less. Early: play the down, not the scoreboard. |
| Time | How many more chances will we get? | 11:30 in the third leaves maybe six or seven possessions. 2:41 in the fourth leaves one. |
| Timeouts | Can the clock hurt me? | Matters most in two-minute and four-minute. On third-and-7 in the third quarter, almost not at all. |
| Who decides fourth down? | What happens after a failure? | If Wade will punt on fourth-and-7 from our 41, then my "failure" outcome is a punt, not a turnover on downs. |
That last row isn't in most textbooks, and it should be. On our staff, Wade owns fourth downs with input from Marisol and me. He's made calls I wouldn't have. In Game 1 against Sacramento, on fourth-and-goal at their 3 with a 7–0 lead, Marisol's chart said "go, close," I'd have gone, and Wade kicked the field goal. We won that game 34–27. Because Wade owns fourth down, I have to predict him on third down. If I know he'll punt on fourth-and-3 from our 45, my third-and-7 call can't treat a four-yard gain as a setup for fourth down. It's a punt with four more yards on it.
Field zones change the menu
We talk about the field with the same words Teo's scouting report uses: backed up (our goal line to our 10), coming out (our 11 to 24), open field (our 25 to their 31), the fringe (their 30 to 21), the high red (20 to 11), the low red (10 to 4), and the goal line (3 and in). The call sheet is organized the same way. A third-and-7 in the open field and a third-and-7 in the high red are different calls, because in the red zone the end line acts as an extra defender and a sack can cost you the field goal.
📈 By the numbers: Expected points (EP) is the average number of points a team scores from a given down, distance, and field position, based on many seasons of real play-by-play. Expected points added (EPA) is the change in expected points caused by one play. Public models (nflverse and others) differ in the details, but they agree on the shape of our playoff third-and-7 at our own 41. Converting is worth roughly two points of expected value more than punting. A sack makes it a little worse than a plain punt, because the punt comes from farther back. An interception is worse by several points. Those gaps, not the averages, are what you're choosing between. Read them as approximate, with a model's error bars. They're a scale for the stakes, not a forecast.
🔄 Check your understanding: Why does the play caller need to know the head coach's fourth-down tendency when he calls third down?
Answer
Because the fourth-down decision determines what a third-down failure costs. If the head coach will go for it on fourth-and-short, a third-down call that gains five of seven yards has real value; it sets up a manageable fourth down. If he'll punt, the same five yards just move the punt a little. The caller is choosing among outcomes, and the head coach's fourth-down rule decides what some of those outcomes are worth.
25.4 The Third-and-Seven Decision
This is the decision I promised you in the hook. I'm going to give you everything I had at 12:06 of the third quarter: every input, every number Teo and Marisol handed me, everything I knew about the other sideline. Then you're going to make the call before I tell you mine.
The situation
- Game: Divisional playoff at Sacramento. A loss ends our season.
- Down, distance, field position: 3rd-and-7 at our own 41. The line to gain is our 48.
- Score and time: Sentinels 10, Outlaws 7. About 11:30 left in the third quarter, opening drive of the second half. Each team has all three timeouts.
- What failure costs: Fourth-and-7 at our 41 is a punt. Wade would punt it, Marisol's chart would punt it, and so would I. A sack like Week 15's turns it into a punt from around our 30.
- What success buys: A first down near midfield, with Dorsey's range line (about their 35, a kick in the low fifties) another 15 yards or so away.
- Personnel: We're in Eleven (one back, one tight end, three receivers): Vance, Ruffin, Gaines, Powers, Harlan, Sutton, and our five linemen. Sacramento is in nickel and didn't sub on third down. Bell and Ames, the mug linebackers, are both on the field.
What Teo's page said
Teo's divisional "Kimbrough by situation" page had eight rows. Four of them were about this snap. (These are the words as they stood on the page in January. In the spring, after Russ walked me through his own names, Teo penciled them over ours.)
| Row | Situation | Teo's line |
|---|---|---|
| 5 | 3rd-and-7+ vs. 2x2 | "Stayed two-high, every snap since their bye." |
| 6 | Mug | "Can't call it." |
| 7 | Pressure | "Low and flat." |
| 8 | Tells | "None found… since Week 10." |
Row 5 is the streak: twenty-nine snaps in a row. Row 6 is the honest shrug: since their bye, they'd shown the mug on long downs often enough to scare us and called the pressure rarely enough that Teo couldn't find a pattern. Row 7 says their overall pressure rate was low and steady. Row 8 says Russ had cleaned up every tell we used to have on him.
Read rows 5 and 6 together, because that's the trap. Row 5 counted snaps where Sacramento showed a quiet front and two high safeties on third-and-long. Our protection page put the same idea in Vance's words: "Quiet front, two-high on third-and-long: it's two-high." Row 6 was a different picture. When both linebackers walked into the A-gaps, the streak didn't apply.
What the film and the numbers said
- Week 15, same down and distance. The sack. Our old rule ("six-man, back on the Mike") plus a late walk plus a dropping 3-technique equaled Bell free. You know the clip by now.
- Vance against pressure. In the four-game sample Sacramento charted on us before Week 6, Vance was 28 of 36 for 9.4 yards per attempt, 4 touchdowns, and no interceptions against five or more rushers. Against four rushers with post-snap rotation he was 31 of 52, 6.1 yards per attempt, 1 touchdown, 3 interceptions. Russ knew both numbers as well as we did. That's why he pressured us on purpose and rotated on us constantly.
- Matchups. Sutton against Cole Brennan, their second-year boundary corner, was our best one-on-one, but only without a safety over the top. Over two seasons Sutton was 2 of 7 against Cover 2 or with a safety over him. Two-high on third-and-long meant a safety over him on every snap. Gaines against Malik Simmons, their field corner and corner blitzer, was a fair fight that we could win quickly.
- Rollins. DeShawn Rollins, their premier edge rusher, against Ben Achterberg. On third-and-7-plus, Ben's first kick in his pass set went about a yard deeper, which is what Rollins's inside spin feeds on. We'd already taken a Rollins sack on a third down in the first quarter.
- Our own protection page, "Sacramento: front honesty," on the card Vance carried: 1. Count the safeties before the front. 2. Mug: assume one drops. Block the Mike with the center, the other with the back, and make the corner your hot. 3. Standing front: big on big, hot off the widest. 4. Quiet front, two-high on third-and-long: it's two-high. 5. Late walk: throw hot to the side he walked from.
- The plan's two boxes. Our third-long box, built on Tuesday, was split in two: quiet front → Mills (the post-dig) with a checkdown to Ruffin, and mug → the six-man protection built for the mug, with Gaines's sight-adjust hitch. We'd repped Mills against two-high all Thursday and the mug protection against every version of the mug the scout team could show.
And one thing I knew that no page could tell me: Russ Kimbrough is the smartest defensive coach I face. He doesn't call pressures for fun. He calls them when the picture tells him they'll land.
Here's the picture I could create. I chose the formation. The defense chose everything else.
PRE-SNAP: Week 15, Q3, 6:51 -- 3rd-and-7, AUS 38, 13-13
(ball on the left hash: boundary left, field right)
F (Royce)
$ (Holloway, ~8)
C (Brennan) * (Vo) C (Simmons)
S E N W M T
X Y LT LG C RG RT H Z
R Q
old rules: slide to the crowd (left) | R on the Mike | a late walk
moves R | A-gap pressure -> hot to H in the field slot
📊 Diagram (described): Our Week 15 formation: Sutton wide left, Harlan flexed about four yards outside the left tackle into the boundary, Powers in the right slot, Gaines wide right, Ruffin on Vance's left. Sacramento has one safety, Royce, deep in the middle and Holloway down at about eight yards on the right. On the line, left to right: Weller, who has just walked out to stand outside Rollins; Rollins; the nose head-up on the left guard; Ames in the left A-gap; Bell in the right A-gap; Coleman outside the right guard. Vo is over the right side, Simmons is off Gaines, and Brennan is on Sutton. The notes list our old rules: slide toward the crowd, back on the Mike, a late walk moves the back, and the hot is Powers in the field slot.
POST-SNAP: Week 15 -- five rush, two "rushers" drop, Bell free
[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
R->(air) LT LG C | RG (air) RT (late)
v
Q on the hot at 0.8 s; sacked at 1.9
📊 Diagram (described): At the snap five Sentinels rush: Rollins, the nose, Ames, Bell, and Simmons off the right edge. Weller and Coleman, both counted as rushers, drop: Weller to the boundary flat and Coleman back into the short zone under Powers, the hot window. Our left tackle, left guard, and center have slid left onto Rollins, the nose, and Ames. Ruffin, moved out to Weller, blocks nobody; the right guard, whose man dropped, blocks nobody; the right tackle picks up Simmons late. Bell runs free through the right A-gap. Behind the rush, Brennan, Royce, and Holloway play three deep. Vance is on his hot at 0.8 seconds, sees Coleman in the window, and is sacked at 1.9.
That's what the formation had produced the last time. Now it's your turn.
🧩 Productive struggle: You're the play caller. It's 12:06 of the third quarter, 3rd-and-7 at your own 41, down 10–7 in a playoff game. You have Eleven personnel on the field, and Sacramento is in nickel with both mug linebackers in. You have everything above: Teo's four rows, the Week 15 clip, Vance's pressure numbers, the matchups, the front-honesty page, and the two-part third-long box. Before you read on, write down three things:
- Your call, in one line: formation, protection, concept.
- The defense you think is most likely, and what your call does against it.
- The defense you're most afraid of, and what your call does against that.
Then grade yourself on only one question: does your call have an answer for both? If it does, you've already learned most of this chapter. If it only beats the defense you expect, keep reading.
Four defensible calls
There was no single right answer on that sideline. I can give you at least four calls I'd defend in front of any staff in the league. Each one bets on a different defense, and each one is worth knowing.
Call A: Trust the page. "Eleven, Gun Doubles Left, 60 Right, Mills" from our standard Mills look, Harlan attached on the left. This is what the Tuesday plan pointed at. Gaines runs a post at about 12, Powers a dig at 12, Sutton a go, Harlan to the flat, Ruffin stays in. Against a quiet front and two-high Quarters, it puts the field safety in a conflict: if he jumps the dig, the post is behind him; if he stays on the post, the dig is in front of him. The mug rule rides along with every third-and-long call, so even if the mug shows, the protection has an answer. - Bets on: Row 5. Twenty-nine straight two-highs. - Strength: The best concept we had against the most likely coverage, repped all week. - Weakness: It asks the post to be open in a stadium where the middle of the field had a habit of closing late. And it gives Russ a picture he hasn't seen, so if he has a pressure in his pocket, we can't predict which one he'll call.
Call B: Respect the sack. "Eleven, Gun Doubles Left, 80 Max, Gaines dig, Sutton go" — eight in protection counting Harlan's hold, two receivers out. Nobody runs free against any five- or six-man pressure. - Bets on: Pressure. - Strength: Almost no sack risk and no hot throw to cover. - Weakness: Against a quiet front and two-high, which the page says is the likely look, it's two receivers against seven droppers on third-and-7. It also tells Russ we're scared of a pressure he called once in a month. A quarterback who's 28 of 36 against the blitz shouldn't be protected out of his best weapon.
Call C: Play for the punt. Draw or Slip screen, the same calls our long-long box held. - Bets on: Nothing in particular. It aims for a field-position result. - Strength: The lowest-variance call on the sheet. Against a mug, a draw goes right at linebackers who are rushing past the ball. Against two-high soft coverage, a Slip screen can get seven if Ruffin makes one man miss. If it fails, we punt from the 41 or better, with no sack and no interception. - Weakness: It converts maybe twice in ten. In a playoff game where you're down three and the other team's defense is better than your offense on paper, punting a drive away because you were afraid is how you lose 20–10.
Call D: The one I called. "Eleven, Gun Doubles Left, Y Off, 60 Right, Mills." It's the Week 15 formation: two-by-two, Harlan flexed a step off the line about four yards outside the left tackle so Sutton stays an eligible end, and Gaines on the line wide right to give us seven. The protection is a 60 Right half-slide: the center, right guard, and right tackle slide right, toward Coleman, the 3-technique, and Ruffin sets on the man side, the left. And the mug rule is live: center on the Mike, back on the other A-gap linebacker, right tackle widens to the corner. The concept is still Mills, so against a quiet front it's Call A with Harlan detached. If Simmons blitzes, Gaines's post becomes a six-yard hitch. - Bets on: Both. And it bets that showing Russ the Week 15 picture makes the mug more likely, which I wanted. - Strength: Against a quiet front it's our best two-high concept. Against the mug it's the protection we repped all week, and Gaines's hitch is one of the easiest completions in football, because the man who covers it is rushing the quarterback. - Weakness: If Russ had a third answer, a pressure built to beat our new rule, I'd have invited it. In Chapter 22 we drew that answer: drop both mugged linebackers and rush from the edges. Call D has the fattest tail of the four. When it fails, it fails with a sack.
And one call I wouldn't defend, even though fans wanted it: Call E, "Sutton back-shoulder." Our best one-on-one, isolated into the boundary. Against two-high, with a safety over the top on every snap, he'd been 2 of 7 over two seasons. That isn't a call. It's a highlight you're hoping for. It's the call Curtis Bale makes, and I'll come back to Curtis.
🔄 Check your understanding: Calls A and D use the same concept. What's the actual difference between them, and why does a formation change count as part of the bet?
Answer
The difference is the formation and the explicit 60 Right slide. Call D shows Sacramento the exact Week 15 picture: two-by-two with Harlan flexed into the boundary. That picture changes which defense is likely. Russ had beaten that formation with the mug before, so showing it raises the chance he spends his mug on it. The formation is part of the bet because the offense's picture is one of the inputs to the defense's call. You can't choose the defense, but you can change the odds of what you get.
What happened
At :04 Szabo snapped it. Sacramento ran the same pressure it had run in Week 15. I learned in the spring that Russ calls it Mug Cat Fire. On our sideline we called it "the double-A mug with the corner fire," because I didn't know his names during the season. It was the only time they called it all game.
PRE-SNAP: Divisional, Q3, ~11:30 -- 3rd-and-7, AUS 41, SAC 10-7
"Eleven, Gun Doubles Left, Y Off, 60 Right, Mills"
F (Royce, deep middle)
$ (Holloway, ~8)
C (Brennan) * (Vo) C (Simmons)
S E N W M T
X Y LT LG C RG RT H Z
R Q
60 Right: C, RG, RT slide right; R on the man side (left)
mug rule: C on M (Bell) | R on W (Ames) | RT widens to the corner
late walk (S): nobody moves; he's an extra, not a reassignment
📊 Diagram (described): The same formation as Week 15: 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 showed two high earlier in the play clock, but by the snap it is one high: Royce alone in the deep middle and Holloway down at about eight yards on the right, ready to climb to the deep third. Ames and Bell are in the A-gaps; Weller has walked out to the boundary outside Rollins; the nose is on the left guard and Coleman outside the right guard; Vo is over the right side and Simmons over Gaines. The notes show the call: a slide to the right, Ruffin on the left, and the mug rule that puts the center on Bell, Ruffin on Ames, and the right tackle on the corner. When Weller walks, nobody moves.
POST-SNAP: same pressure, new rules -- zero free rushers
[DEEP 1/3] C [DEEP 1/3] F [DEEP 1/3] $
S drops (curl/flat) T drops (H's window again)
* (middle hook)
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 drops to the boundary curl and flat, Vo to the middle hook, and Coleman drops too, exactly as in Week 15. This time every rusher has a blocker: the left tackle on Rollins, the left guard on the nose, Ruffin stepping up into the left A-gap on Ames, the center on Bell, and the right tackle, who widened at the snap, on Simmons. The right guard's man dropped, so he helps inside. Coleman sits in the window under Powers again, but the ball isn't going there. Gaines sees Simmons leave, stops at six yards, and turns; Vance throws it at about 1.1 seconds, and Gaines turns it into a ten-yard gain and a first down at the Sacramento 49.
Zero free rushers. Gaines, hitch, ten yards, first down at the Sacramento 49.
The drive went on. On 3rd-and-2 at their 29, Gaines went in zoom motion across the formation and caught a quick out for five, first down at the 24. Three plays later it stalled, and Dorsey kicked the field goal: 10–10.
Look at everything that didn't change between the two snaps. Same formation. Same down and distance, near enough. Same defensive call. Same Weller walk at ten seconds. Same Coleman drop into the same window. Vance made the same Mike call. What changed was the rule sheet, and a play caller who decided to show Russ the exact picture he'd beaten, because the rule sheet was finally ready for it.
🛡️ The defense's view: Put yourself in Russ's chair at the same moment. He's ahead 10–7, on his own field, in a playoff game. His page on us says Austin's third-and-long protection is the one he beat in Week 15. He has one pressure he trusts against that exact picture, and Austin just showed him that exact picture. A great defensive coordinator calls that pressure, and I'd have called it too. Our job wasn't to stop him from calling it. It was to make sure it found nobody free when he did. I never asked him why he'd waited until the third quarter to spend it. There are some things you leave on the other side of the table.
And if he'd stayed quiet?
The mug was never the likeliest defense. Most of the time, row 5 says, Sacramento shows a quiet front and stays two-high. So you have to grade Call D against that branch too.
POST-SNAP: the same call vs. a quiet front and two-high Quarters
(the branch Thursday's practice was built on)
F (Royce) Z post ^ $ (Holloway): THE READ
| carries X \ jumps the dig -> post
| \ sits on post -> dig
^ X go W M \
| \___________________________
| <---------------------------- H dig at 12|
| Y--> flat | |
X Y LT LG C RG RT H Z
R Q
📊 Diagram (described): Against two-high Quarters with four rushers, the same call becomes the plan's quiet-front answer. Sutton runs a go on the left, and Royce, the left safety, carries him. Harlan runs to the left flat. Powers runs a dig at twelve yards from the right slot, breaking left under the linebackers. Gaines runs a post from the right, breaking toward the middle at about twelve. Holloway, the right safety, is the read: if he drives on Powers's dig, Vance throws the post behind him; if he stays on the post, Vance throws the dig in front of him. Ruffin stays in to block.
On that branch, Call D is Call A with Harlan detached, and the same concept we'd practiced all Thursday. It isn't better than Call A on that branch. It's about the same. Call D only wins its argument with Call A on the mug branch, where the protection and the sight-adjust turn Russ's best pressure into one of our easiest throws.
💡 Coach's point: A good third-down call doesn't need to be the best call against every defense. It needs to be acceptable against every defense you might reasonably get, and very good against at least one. Call D was about as good as Call A against the quiet front and much better against the mug. Call B was very good against pressure and bad against the quiet front. Call C was mediocre against everything. That's how you rank them.
25.5 A Call Is a Bet
🚪 Threshold concept: A play call is a bet on a distribution, not a promise of an outcome. Judge the call by the decision process, not by the single result.
This is the idea that changes how you watch football, how you coach, and how you sleep on Sunday night. It's also the hardest idea in this book to actually believe, as opposed to agreeing with it.
Out of ten
On Thursday of divisional week, Marisol and I did something we do for every box that worries us. We wrote it out of ten. For each call in the box, we asked: if we ran this snap ten times against what Sacramento was likely to show, what would happen? We started with how often each defense would come. Teo's streak said two-high was the likeliest look by a lot. The mug row said "can't call it," so we gave it two or three times in ten for an ordinary picture, and, by my guess, three or four in ten if we showed Russ the Week 15 picture.
These are estimates. They're built from our charting, Sacramento's charting, Vance's numbers, and judgment, and they have honest error bars of a snap or two either way. Marisol wrote "± 1–2" at the top of the card in red, and she was right to. The point isn't the decimals. It's the shape.
| Call | Convert | Punt (no sack) | Sack | Turnover |
|---|---|---|---|---|
| A — Mills, Harlan attached, trust the page | about 5 | about 3½ | about 1 | about ½ |
| B — 80 Max, two routes | about 3 | about 6 | about ½ | about ½ |
| C — Draw or Slip screen | about 2 | about 8 | about 0 | about 0 |
| D — Week 15 picture, 60 Right, mug rule, Mills | about 5½ | about 2½ | about 1½ | about ½ |
OUT OF TEN (Thursday estimates, +/- 1-2 snaps) # = one snap in ten
convert punt sack turnover
A Mills ##### ###+ # +
B 80 Max ### ###### + +
C Draw/Slip ## ########
D Our call #####+ ##+ #+ +
+ = about half a snap. Convert is worth ~2 points over a punt.
📊 Diagram (described): A bar chart of four calls, each split into four outcomes out of ten snaps. Call A converts about five times, punts about three and a half, takes about one sack, and turns it over about half a time. Call B converts about three, punts about six, with half a snap each for a sack and a turnover. Call C converts about two and punts about eight, with essentially no sacks or turnovers. Call D, the call I made, converts about five and a half, punts about two and a half, takes about one and a half sacks, and turns it over about half a time. A note says a conversion is worth roughly two points of expected value more than a punt.
Now read it the way a play caller has to read it.
Call C is the safest call and probably the worst one. It almost never loses the game in the third quarter, and it almost never wins anything either. Two conversions in ten is a punt with a small lottery ticket attached.
Call B buys safety with conversions. It cuts the sack and turnover risk but gives up two conversions in ten to do it. Since a conversion is worth about two points and a sack only a few tenths more than a punt, that's a bad trade.
Calls A and D are close, and the difference is in the tails. D converts a little more often, and it gets sacked a little more often too, because it invites pressure on purpose. If you believe our protection was ready and Vance was the quarterback his numbers said he was, D is the better bet. If you think the rule had a hole Russ could find in one week, A is the better bet. That's a judgment, and staffs can argue about it honestly. In fact we did, on Thursday. Gus wanted A, and his margin note said: "Don't ask him to call it."
So I made a bet. I called D because I thought our best third-and-seven in the building that day was Russ's best pressure, hitting a protection we'd rehearsed for four weeks, with the quarterback in the league I'd most trust against the blitz. That's the decision. The ten-yard hitch was the result.
Before and after
Here's what this threshold concept changes.
Before: A good call is one that works. The hitch to Gaines was a great call because it got ten yards. The Week 15 sack was a bad call because it lost nine. You judge the play caller by what happened.
After: A good call is one whose range of outcomes, given what the caller knew when he made it, was better than the alternatives. The result of one snap is a single draw from that range. The hitch to Gaines was one draw from a call that converted about half the time. If Russ had stayed quiet and Holloway had made a great play on the dig, the call would have been just as good, and we'd have punted. The Week 15 sack was one draw from a set of rules that had been sound on the Week 6 evidence and had become readable by Week 15. Most of the failure was in the rules, not the call.
Once this clicks, three things change.
- You stop learning the wrong lessons from single results. A caller who judges by results will abandon a good call after one bad draw and repeat a bad call after one lucky one. Over a season, that's the difference between a caller who improves and one who just reacts.
- You start asking better questions after a failure. Not "why did that fail?" but "was the information right, was the box right, was the alternative better, and was the execution what we'd repped?" Those are four different questions with four different fixes.
- You can live with the job. Every play caller in the NFL will call hundreds of good plays that fail this season. If you take each failure as a verdict, you won't last.
Poker players have a word for judging a decision by its result: resulting. Annie Duke's book Thinking in Bets brought it to a general audience, and once you have the word you'll notice resulting in every postgame show you watch.
Process and outcome
Put the two ideas in a grid and every call you've ever made fits in one of the four boxes.
| Good result | Bad result | |
|---|---|---|
| Good process | Earned. Anchor C: the right bet, and the draw went our way. Repeat the process. | Bad beat. The call was right and the draw wasn't. Keep the process; check the execution. |
| Bad process | Dumb luck. The defense busted, the throw was wrong, and it worked anyway. Don't repeat it. | Earned, unfortunately. The call was wrong and it lost. Fix the process. |
The dangerous box is the bottom left. Dumb luck feels exactly like being good, and a staff that grades by results will put the lucky call back on next week's sheet.
The most famous example in recent football history sits right at the center of this grid: the Seattle Seahawks' second-and-goal pass from the 1 at the end of Super Bowl XLIX, intercepted at the goal line. We took that one apart in Chapter 14. Whatever you believe about that call, notice how much of the public argument was about the result and how little was about the clock, the timeouts, and the personnel, the things the decision was actually made from.
🔍 Why this works: Football outcomes are noisy, and that's why the process view is right. A perfectly designed third-down call against the coverage it beats still fails some of the time: a receiver slips, a defender makes a great play, a ball tips. A badly designed call against the wrong coverage still succeeds some of the time, because the defense busts. If your sample is one snap, the noise is as big as the signal. You can't learn from the result alone; you can learn from the reasoning, because the reasoning is the part you control and the part you'll use again.
Grading our own
On Monday, Marisol grades every call twice: once by result (success rate and EPA, the numbers in Chapter 28) and once by process. The process grade asks four questions: Was it from the right box? Was the information in the box right? Was there a better call in the box? Did we execute what we'd repped? Here are three Sacramento third downs graded both ways.
| Snap | Result grade | Process grade | Where the fix lived |
|---|---|---|---|
| Week 15, 3rd-and-7, AUS 38 (sack, −9) | Fail | Call: fine. Rules: failed, readable after four games of film. | The protection card |
| Divisional, 3rd-and-7, AUS 41 (hitch, +10) | Success | Earned: right box, right information, protection and sight-adjust executed. | Nothing to fix; keep it, and expect Russ to answer it |
| Divisional, 3rd-and-9, AUS 38, 2:41 Q4 (Mills, intercepted) | Fail, badly | Call: defensible. Plan: one answer and a blank contingency row. QB: read the shell before the snap, didn't recheck the middle after it. | The plan, then the quarterback's post-snap read |
The third row hurts. It's the same concept as the second row, ninety minutes later, from almost the same picture. Both were Mills. At 11:30 of the third quarter, Russ picked the branch we'd practiced. At 2:41 of the fourth, he picked the branch we hadn't: two high before the snap, the strong safety rotating down and Ike Royce spinning to the middle after it. He'd stayed two-high twenty-nine straight times, and the thirtieth was the interception. Chapter 24 owns the plan's part in it, and Chapter 26 tells that last drive snap by snap. Here I only want you to see the pattern: the same call can be a great bet at one moment and a flawed one at another, because the distribution under it has changed even when the routes haven't.
⚠️ Common mistake: grading the call by the replay. The broadcast shows the end-zone angle of the interception and the booth analyst says, "You can't throw that ball." True, and it tells you nothing about the call. The fix: before you grade any call, write down what the caller knew at 34 seconds and what the alternatives were. If you can't, you're grading the result.
🔄 Check your understanding: Using the process-and-outcome grid, where does the Week 15 sack belong, and what's the evidence?
Answer
Mostly in "good process, bad result" for the call, and in "bad process" for the rules. The concept and the decision to throw on third-and-7 were reasonable. But the protection rules were written from Week 6 evidence and had become readable: Russ's staff had found that our hot throws went to the field slot window 8 of 8 times when the A-gaps showed pressure, and they knew a late walk would move our back. The fix belonged to the rules, which is exactly where we put it.
📋 Try it — Make the call: Next season, Week 6 at home against Sacramento. Second quarter, 8:20 left, tied 7–7. 3rd-and-6 at the Sacramento 44. You're in Eleven, and Sacramento is in nickel. Your booth reads "Two high, six, Vo's been inside on the last two." Your updated page says: on 3rd-and-4–6 against 2x2, Quarters with the corners off about 7 times in 10, the nickel blitz about 2 in 10. Your call sheet's third-medium box against 2x2 holds Flood, Stick, and "Quick Outs." Wade will punt on fourth-and-anything here. Make the call, then write it out of ten against both likely defenses. (This is a hypothetical game.)
Worked answer
One strong call: Stick from 2x2 to the side of the nickel, with the H's slant as the hot. Against Quarters with off corners, Stick puts the hook defender in a conflict between the stick at six and the flat. The stick sits in the window between the nickel and the Will at the sticks; the flat beats a hook defender who widens. Against a nickel blitz from Vo's side, Powers's route becomes the quick slant into the space Vo left (our late-walk and hot rules), and the ball is out at about 1.0 second.
Out of ten, roughly: against Quarters (seven in ten), about 6 conversions; against the nickel blitz (two in ten), about 6 as well, because the hot is built in; some other look (one in ten), about 4. That's roughly 6 in 10 overall, with low sack risk because the ball comes out fast.
Why not Flood? Flood is very good against Cover 3 but slower to develop, and against off-Quarters it asks for a deeper throw on a down that only needs six. Why not Quick Outs? They convert against off corners only if the receiver makes a man miss after the catch; six yards is right at the edge of what the route earns. Either is defensible. The one call that isn't defensible is one with no answer to Vo, because your booth just told you about him.
25.6 The Other Coordinator
Every call I make is a guess about what Russ is about to call, and every call he makes is a guess about mine. The game is a conversation, and you can't hold up your end of a conversation if you don't know who you're talking to.
So I keep a model of him. It isn't a mystical "feel." It's a short list of beliefs, each with evidence, that I revise during the game. Here's the one I carried into the divisional round.
| What I believed about Russ | The evidence | What it meant for my calls |
|---|---|---|
| Coverage first; pressure on purpose. | ~62% two-high pre-snap; blitz rate in the high 20s; pressure lives mainly on third-and-long. | Expect four-man rushes on early downs. Save hot-throw answers for third down, but have them on every third-down call. |
| He lets you invest, then changes the price. | Week 15: our quick game went 11 of 13 in the first half, then he pressed and rotated about half our early downs. | Expect our best first-half concept to be taken away after halftime. Have the next answer on the sheet before he does it. |
| He saves his best call for the biggest snap. | His mug had been "can't call it" since his bye. | Assume his best pressure is waiting for a third-and-long in a close game. Build the protection to meet it, not the play to avoid it. |
| He reads our rules, not just our plays. | Week 15: every one of our protection rules fired exactly as written, and he'd built a call around them. | A rule that works twice is a rule he's studying. Change the rule before he changes the call. |
| He respects Vance against the blitz. | Our sample: 28 of 36 against five or more. | When he does pressure, it'll be a sim or a fire zone with a dropper in the hot window, not an all-out blitz. |
The model isn't a script. It's a set of starting odds. The first three third downs of a game tell you whether today's Russ matches the model, because a defensive coordinator shows his plan early on third down. When the game disagrees with the model, the game wins. "Never let a tendency outlive its evidence" has been on my whiteboard in red since the spring. That rule is about my model of Russ as much as it's about Teo's page.
🛡️ The defense's view: Russ keeps a model of me too, and in the spring he told me some of it. His staff's first whiteboard line on our offense was "They run to make you wrong, not to make yards." He knew I'd test before I attacked, and that I'd call the same play twice in a row if the first one told me something. What a model like that buys a defensive coordinator is timing. If you know the offense will show you its question on the first snap of a series and its answer on the third, you know which snap to disguise.
Tendencies: theirs and yours
Russ's model of me is built from my tendencies, the same way mine is built from his. So the play caller has two jobs with tendencies: use the opponent's, and manage his own.
Two of mine were sitting in the charting Russ's staff did on us before Week 6, and Marisol's self-scout told me before the playoff that neither had gone away:
- First play of a drive from 2x2: Inside Slant (the inside-zone RPO with the slant tag) on 7 of 11 in that sample.
- 2nd-and-1 to 3: a play-action shot on 6 of 17.
You don't break a tendency by never calling the play again. That just trades one tendency for another. You break it on purpose, at a cheap moment, and you break it with the same look. The divisional script opened with a jet sweep instead of the RPO, and its fifteenth call was "2nd-and-1 to 3 only: Inside Right," a plain run from the play-action picture. The goal is to make the defense's tendency chart wrong just often enough that it can't trust the chart. It shouldn't be wrong so often that you've walked away from your best plays.
💡 Coach's point: A tendency is only a problem if the defense knows it, and it's strong enough to bet on, and the defense has a call that punishes it. All three conditions are necessary. Check them in that order before you spend a good call breaking a habit.
🔄 Check your understanding: Why does Lyle say that a defensive coordinator's model of the offense mostly buys him timing?
Answer
Because a defense can't disguise or pressure on every snap without giving up soundness. If it knows when the offense asks its questions and when it cashes the answers (for example, the third play of a series, or third-and-long after a successful first down), it can spend its disguises and its best pressures on those snaps. A good model doesn't tell the defense what you'll call so much as when the call will matter.
25.7 Sequencing: Calls as Investments
I don't call plays. I call the next play's reason to work.
That's the line people quote back to me, and it's the honest center of how I call a game. A call has two payoffs: what it gains now, and what it teaches the defense to do next. Some calls are worth more for the second payoff than the first. I call those investments.
The clearest investment I've ever made is Anchor D from Game 1, the four-play sequence that tied the game at 27. From the same look all four times (Eleven, Gun Doubles Right, Harlan attached right, Ruffin offset left), we ran inside zone right for five, then for seven, with Holloway filling the alley harder each time. Then we ran play-action off the identical action, and Gaines ran the post into the deep quarter Holloway had just left, 35 yards to their 4. Then inside zone again, and Ruffin walked in because Holloway didn't trust the run anymore. (Chapter 19 breaks down every frame.) The first two runs were good plays on their own. Their real value was what they did to Holloway.
Now apply the threshold concept to the investment. The two inside zones weren't bets on the inside zones alone. They were bets on a pair: run, run, play-action. Grade each one by itself and you'll call them "five-yard runs." Grade the sequence and they're the setup for a 35-yard throw. This is why we have series cards on the call sheet: chains of calls with the trigger for each step written down. It's also why Chapter 19's "two receipts" rule is printed on top of every play-action box: the defender you're going to lie to has to show you his reaction twice, in the same picture, in this game, before you cash it.
Rhythm versus predictability
Quarterbacks talk about rhythm, and they mean something real. After three quick completions the ball comes out faster, the feet are calmer, and the eyes go where they belong. Coordinators talk about predictability, and they mean something real too. Three quick completions in a row is a pattern, and the defensive coordinator on the other headset has noticed it.
Week 15 is the lesson. Our first-half quick game went 11 of 13 for 74 yards, and the longest gain was 11. That's rhythm. Russ told me in the spring that he wanted us to complete them, because eleven of thirteen meant we weren't throwing over his corners. After halftime he pressed about half our early downs and rotated out of two-high on about half of them. Our rhythm became his scouting report.
Here's how I balance the two, as rules on the sheet rather than vibes:
- Rhythm on the first two downs of a drive, variety by the third. A quick completion on first down is rhythm. The same concept on the third straight first down is a tendency.
- Rhythm by the quarterback's eyes, not by the play. Different concepts that give Vance the same read (the same defender, the same first look) keep his rhythm without repeating the picture.
- After a big play, go fast. We aren't a tempo team, but after an explosive we'll snap quickly with a call from the fast-mode panel. That's rhythm the defense can't sub against.
- When the defense changes the price, change the purchase. If they start pressing the quick game, the quick game's answer (the stack release, the Z Sluggo, the zoom motion) is already on the sheet.
📋 Try it — Make the call (sequencing): A home game next season against St. Louis, a single-high, eight-man-box defense. First quarter, 0–0. You've run "Eleven, Gun Doubles Right, Inside Right" twice on this drive, for +3 and +4. Both times the strong safety, rolled down at eight yards over the tight end, came downhill on the first step. Now it's 1st-and-10 at the St. Louis 45. Your call sheet has Inside Right, Inside Slant, "Inside PA, Z Post," and Wide Right. Choose your call and write the next two calls in the series, with the trigger for each. (This is a hypothetical drive.)
Worked answer
Call: "Inside PA, Z Post" from the identical look. You have two receipts: the same defender, the same reaction, the same picture, this game. Against single-high, the post runs at the middle safety, so Vance must hold him with his eyes first. The read is the rolled-down safety's first step. If he comes downhill again, the space behind him belongs to the post or to the Y running behind the fake.
Next call if the PA hits: go fast. "Eleven, Gun Doubles Right, Inside Right," the same look again. The safety who just got burned will hesitate, and that hesitation is your alley.
Next call if the PA is covered (the safety stayed home): he's stopped trusting the run fake, so the run is honest again. Either Inside Right with a box count you like, or Inside Slant if the Will is the one now fitting downhill. The trigger is what the safety did on the fake.
Calling Wide Right here isn't wrong, but it spends a fresh look before you've cashed the two receipts you earned. That's a sequencing mistake, not a scheme mistake.
25.8 The Headset, the Booth, and the Head Coach
A play caller's headset is the busiest room in the stadium, and the more voices it carries, the worse the calls get. Here's who's on ours and when each person is allowed to speak.
GAME-DAY CHANNELS (Austin)
COACHES' BOOTH (upstairs) SIDELINE
+---------------------+ +-----------------------------+
| Teo -- the booth |==== OC ====> | Lyle (calls) |
| read: shell, box, | channel | | |
| one thing | | +--radio (dead at :15)--> |
| | | Vance's helmet |
| Marisol -- numbers |==== HC ====> | Wade (4th downs, timeouts) |
| | channel | joins OC channel only |
+---------------------+ | to give a decision |
| Gus, Rome, Cal, Lonnie -- |
| OC channel between series |
+-----------------------------+
📊 Diagram (described): Two boxes, the coaches' booth upstairs and the sideline. In the booth, Teo talks to me on the offensive-coordinator channel, and Marisol talks to Wade on the head-coach channel. On the sideline, I call the plays and talk to Vance over the radio until it cuts off at 15 seconds. Wade owns fourth downs and timeouts and comes onto my channel only to give a decision. Gus, Rome, Cal, and Lonnie use my channel between series, not between plays.
Between plays, only Teo. The quiet window runs from the whistle to the call, and nobody else talks. I learned that rule the hard way. In my first season calling plays at the college level, I had five voices on the headset. On one third down I heard three of them at once, "the end's soft inside," "they're in man," and "watch the clock," and I called a play none of them had suggested.
Between series, the room opens. When the defense takes the field, the channel opens for about two minutes. Teo summarizes the series in three lines. Gus gives me the line's word ("Fitch says the 3-technique is slanting on every long down"). Rome reports on Vance after looking at the sideline tablet with him. Cal gives me one receiver note. Then it closes again.
Wade on fourth down and timeouts. Wade and Marisol run the fourth-down conversation on their own channel. When we're facing third down in territory where fourth down matters, Teo's "one thing" sometimes becomes Wade's decision, relayed early: "Wade's going on fourth-and-two or less." That changes my third-down call, and I want it before I call third down, not after.
Who owns what
| Decision | Owner | Input from | Notes |
|---|---|---|---|
| The play call | Lyle | Teo (quiet window), Gus/Rome/Cal (between series) | No votes. One voice in the quiet window. |
| Fourth down | Wade | Marisol's chart, Lyle's view | Lyle states a preference once and doesn't argue on the headset. |
| Timeouts | Wade | Lyle, Vance | Vance may call one at :06 or later rather than take a delay, our standing rule. |
| Two-point tries | Wade | Marisol's chart | The base play is on the Green panel. |
| Challenges | Wade | Booth replay | |
| Protection and hot rules | Lyle and Gus | Rome, Teo | Set by Friday; no new protections after Friday. |
The relationship between a coordinator and his head coach is the least-taught part of this job. Mine works because Wade and I settled the boundaries early. He's a former defensive coordinator, so he's the "how would I defend that?" voice in our building, and on game day he lets me call the offense. I let him own the decisions that belong to the whole team: fourth downs, timeouts, the clock. When we disagree, and we have, I say it once on the headset and twice in his office on Monday.
🎙️ On the headset: Game 1, first quarter, fourth-and-goal at their 3, up 7–0. Marisol's chart said go, close. I wanted to go. I said, "I like the go. I've got Jumbo on Green." Wade said, "Kick it." Dorsey kicked it, 10–0. That's the whole exchange, eight seconds. If I'd spent another twenty arguing, I'd have been arguing during my own forty seconds for the next series. We won 34–27, which proves nothing about the kick either way. You've read 25.5.
Booth or sideline?
Coordinators call plays from both places, at every level. In the NFL you'll find play callers on the sideline and in the booth, and a head coach who calls his own plays is on the sideline by definition. In college, coordinators in the booth are common. Neither is right. Each one buys something.
| Booth | Sideline | |
|---|---|---|
| Sees | The whole field, both safeties, the shell before the snap, the rotation after it | The line, the bench, the players' faces, the quarterback between series |
| Feels | Calm. Less noise, less emotion. | The game: who's tired, who's rattled, who's angry, what the crowd is doing to the cadence |
| Talks to | The headset only | The quarterback face to face between series; the line on the bench |
| Risk | Disconnected from the players; harder to read a quarterback's confidence | Emotional, a bad view of coverage, more voices nearby |
I call from the sideline, and Teo is my eyes upstairs. The sideline is where I can see Vance's face after a sack, and that's information no coverage chart has. The cost is a terrible view of the safeties, which is exactly why the quiet window exists. On the sideline, I'm the feel. Teo, in the booth, is the picture.
🏫 Any level: In high school, the "booth" may be an assistant in the press box with a phone headset, or nobody at all, and the play caller may also be the head coach. Scale the system down rather than skipping it. One spotter upstairs who says only "shell and box" between plays gives you most of what Teo gives me. Give your quarterback a sideline "result card" (what the safeties did on the last third down) instead of a tablet. High-school hashes sit 53 feet 4 inches from each sideline under NFHS rules, compared with about 70 feet 9 inches in the NFL, so field-and-boundary calls matter far more. Make "Field" and "Boundary" the first split on your call sheet, before down and distance in some boxes. And in college and high school, check your code's rules on coach-to-player communication. They differ from the NFL's helmet radio and change often.
🔄 Check your understanding: Why does Lyle want Wade's fourth-down decision before he calls third down, and why does he state his own fourth-down preference only once on the headset?
Answer
Before: because the fourth-down decision sets what each third-down result is worth (25.3). A four-yard gain means something different if Wade will go on fourth-and-3. Only once: because the headset is part of the play caller's forty seconds. An argument with the head coach eats the time and attention the next call needs, and the decision belongs to Wade anyway. The place to argue is Monday, with the film and Marisol's chart.
25.9 Failure, Tilt, and the Next Call
You'll fail on the headset. A good play caller in the NFL will call a few dozen failed third downs a season, a sack or two a game, and an interception every couple of weeks, and that's a good season. The question isn't whether you fail. It's what the failure does to the next call.
Poker players call it tilt: making worse decisions because of what just happened rather than what's about to happen. Play-calling tilt has a few recognizable shapes.
- Revenge calling. They just stopped your best run, so you call it again to prove a point. It's the same play, but now the reason is your pride.
- The panic shot. You just took a sack, so you call a deep ball to "get it back." Now you're in third-and-19 instead of punting from third-and-10.
- Abandonment. Two runs fail, and you stop calling runs for the rest of the half. The defense stops respecting the run fake, and your play-action dies with it.
- Freezing. After an interception, you call nothing but checkdowns for a quarter. The defense sits on the short routes, and you've traded one bad play for twelve bad snaps.
Curtis Bale, who worked with me on Walt Dupree's staff, is the cautionary version of a brilliant coach. In his first season as a coordinator his offense was the most talked-about in the league. The next season, defenses played deep. He stopped running the ball when games got tight, his pocket quarterback took sack after sack, and every failure produced a bigger, cooler call. In his second year, a choice route was intercepted on third-and-6 near midfield in a 13-point game. He was fired before the next season. Curtis isn't a villain. He's the best play designer I've worked with, and he never learned that the call after a failure is the most important call of the series.
Our anti-tilt routine
Here is what I do, in order, every time a snap goes badly. It takes about four seconds, and it fits inside the quiet window.
- Listen to Teo first. Shell, box, one thing. Facts before feelings.
- One mark. On the result column of my call sheet, I make one pencil mark next to the failed call. C means the call was wrong for the look. R means a rule failed (protection, hot, sight). X means execution: we called it right and didn't do it. One letter, no sentences.
- Call the next play's reason. I ask what the defense just learned and which call uses what it learned. I don't ask what would make me feel better.
- At halftime, read the marks. Three Cs in a half means my model of the defense is wrong. Two Rs means a rule needs changing now, not next week. Xs go to the position coaches.
After the Week 15 sack I made an R. I didn't know yet which rule had failed, but I knew it wasn't the concept and it wasn't Vance's read. That one letter kept me from doing what I wanted to do on the next series, which was call five straight max-protect shots to punish Russ for embarrassing my line.
⚠️ Common mistake: fixing the last play instead of the next one. After a sack, a caller goes to max protection on every remaining dropback. He's fixed the sack and broken his offense: two-receiver routes against eight droppers on every third down. The fix: diagnose with one letter, change only what that letter points to, and leave the rest of the game plan alone until halftime gives you evidence.
📋 Try it — Whose fault?: A hypothetical game next season at Portland, against the Pioneers' Tampa 2. Third quarter, you're up 17–10. On 2nd-and-6 at the Portland 40, a smash concept to the left is broken up. The corner, Josh Kiel, sank under the corner route instead of squatting on the hitch, and Vance threw the corner route anyway. On the next snap, 3rd-and-6, the caller (frustrated, the booth said later) calls "All Go" against Portland's Tampa 2. The ball is overthrown deep down the middle, and the Pioneers' Mike carries the seam. Punt. Assign the fault for each snap — call, rule, or execution — and name the tilt pattern, if any.
Answer
Snap 1 (smash): mostly X. Smash against Cover 2 is the right call, and its rule is "the corner chooses, the quarterback throws the other one." Kiel sank, so the hitch to Sutton was open. Vance threw the route he expected, not the one the corner gave him. One letter, X, goes to Rome.
Snap 2 (All Go on 3rd-and-6): C, and it's tilt: the panic shot. Against Tampa 2, the Mike runs the deep middle precisely to take away the seam that four verticals depend on. Third-and-6 needs a concept that beats Cover 2 underneath: levels, a dig, the hitch-seam window. The caller was trying to win back the last snap instead of winning this one. The fix isn't a new play; it's the routine. Teo's read, one mark, then the next play's reason.
🔄 Check your understanding: What are the three letters in Lyle's result column, and why does he force himself to use only one?
Answer
C (the call was wrong for the look), R (a rule failed: protection, hot, or sight), and X (execution). One letter forces a diagnosis in about a second instead of an argument with yourself during the next call's window, and it separates "fix it now" (two Rs) from "fix it at halftime" (a pattern of Cs) from "fix it in the position room" (Xs).
25.10 Art and Science
Somebody always asks whether play calling is an art or a science, and the honest answer is that it's a science of the things you can count and an art of the things you can't, and that the line between them moves every year.
What the science does well. It counts. Teo's page told me Russ had stayed two-high twenty-nine straight times on third-and-long. Marisol's model told me a conversion at our 41 was worth about two points over a punt. Our charting told me Vance was 28 of 36 against pressure in the sample. Those numbers were better than my memory, and better than my feel, because my feel remembers the dramatic snaps and forgets the boring ones. In our building, the science sets the odds.
What the science can't see. Our numbers didn't know that Ben Achterberg had taken a speed rush on three straight dropbacks and that his next set would be a yard deeper. They didn't know Vance had been hit late on the previous series and was, for one series, getting rid of the ball a beat early. They didn't know that the scout team had shown two-high on every Thursday rep, so our quarterback's eyes had been trained on one picture. The data doesn't know your tackle is tired. It doesn't know your quarterback is rattled. It only knows what was tagged.
Where feel earns its keep. "Feel" isn't magic. It's information that hasn't been tagged yet: a player's body language, a defensive end's stance, the crowd noise that's going to make the snap count useless. A sideline play caller sees things the booth and the spreadsheet can't. Use them, and then check them against the counts, because feel is also where tendencies hide. After our Week 15 loss I asked Marisol to tag my own "gut" calls, the ones I made against the sheet's first choice. Over the rest of the season, they were about as successful as the sheet's calls, no better and no worse, on a small sample. That told me my gut wasn't a superpower. It was one more input, and a decent one.
Momentum. Every caller believes in it, and the evidence for momentum as a force is weaker than coaches think. Researchers have argued for decades about the "hot hand" in basketball. Early studies found little evidence for it, and later work found a small effect once a statistical bias in the original method was corrected. What's clearly real is the human part: confidence, fatigue, a crowd, a quarterback whose feet are calm. So I don't call plays because "we have momentum." I call plays because Vance's feet are calm and Rollins hasn't gotten home in two series. Those are facts. The word "momentum" just bundles them.
💡 Coach's point: Let the science set the odds and let the art adjust them, and write down every time the art overrules the science. In a season you'll know whether your art is worth what you think it is.
🎬 Film room: To study play calling on film, you need two things the broadcast never shows: the situation at the moment of the call, and the defense before it declared. Watch the All-22 with the game clock and play clock visible. For each third down, pause at the huddle break and write the call you'd make from what's visible. Then play it, and grade your call by process, not by what happened. Ten third downs a week for a month will teach you more about play calling than a season of watching results.
Spaced Review
Answer each question before you read the paragraph under it.
From Chapter 24: what was blank on the Game 3 contingency table, and why does it matter here? The row for "rotate out of two-high on 3rd-and-7+ from a quiet front" was blank, because the evidence said it wouldn't happen. In this chapter's language, the plan treated a twenty-nine-for-twenty-nine tendency as a certainty instead of as odds. A call is a bet on a distribution, and a plan with a blank row has quietly set one outcome's probability to zero. Russ called the thirtieth.
From Chapter 22: in Week 15, how many of our six blockers blocked a rusher, and what was our "seventh blocker"? Four of six; Ruffin and the right guard blocked air when Weller and Coleman dropped. The seventh answer was the hot throw to Powers, and Coleman covered it. Protection is arithmetic, and the quarterback's throw is part of the count. Anchor C fixed the arithmetic: five of six on rushers, one helping, plus a sight-adjust throw the dropper couldn't reach.
From Chapter 8: how many words can a radio call carry before you call the band, and why? Ten. Past ten words the call risks running past :15, when the radio cuts out. In the 25.2 timeline, the call has to be in Vance's helmet by about 20 so the huddle can break by 15. Our Anchor C call was nine words: "Eleven, Gun Doubles Left, Y Off, 60 Right, Mills."
From Chapter 19: what is the "two receipts" rule? The defender you're going to fool with play-action must show the same reaction twice, in the same picture, in this game, before you cash it. It's the investment idea from 25.7, written as a rule.
Project Checkpoint: Call Sheet v1 + Play-Caller's Pregame Checklist
Where you are. In Chapter 24 you built Game Plan v1 and a 15-play opening script against the Sentinels, with boxes, series cards, and contingencies. A plan is too big to use with a play clock running. This checkpoint turns it into the two things you'll actually carry onto the field.
Deliverable 1: Call Sheet v1 (one 11-by-17 page, front and back, or two letter pages).
- Front: openers (the script, with if A / if B), base runs by front, RPO and Kill calls, shots with your "when we call it" rule, early-down passes by shell, play-action (each under the run it imitates), screens, the four third-down boxes (long split into quiet front and pressure), the zero bin, and two or three series cards.
- Back: fringe, high red, low red, goal line, two-minute, four-minute, backed up, two-point, and a result column for your one-letter marks.
- Rules: every box has at least two calls; every third-down call names its hot; every call fits in ten words or carries a wristband number.
Deliverable 2: the Play-Caller's Pregame Checklist (one index card, 12–15 lines). Ours includes: radio and backup radio checked with the quarterback by 90 minutes before kickoff; the head coach's fourth-down chart for today (go ranges by field zone); wristband numbers read aloud and matched to the sheet; the first three third-down calls decided; the quiet-window rule restated to the booth; the "two calls ready" habit; the three things I must learn by the end of the first quarter; and the opponent coordinator's model in five lines.
A worked piece (pro level): the Sentinels' third-long box as we'd write it now, including the row we didn't have in January.
| 3rd LONG (7–10) | Call | Hot / sight | Out of ten (est.) |
|---|---|---|---|
| Quiet front, two-high | Eleven, Gun Doubles Left, 60 Right, Mills (Blue 23) | Read the field safety: jumps the dig → post; sits on the post → dig | ~5 |
| Mug | Eleven, Gun Doubles Left, Y Off, 60 Right, Mills (Blue 24) | C on M, R on W, RT widens; Z hitch if the corner comes | ~5½ |
| Rotation from two-high | Middle-read tag: middle closes → dig or flat | R checkdown | ~4 |
Four levels. - High school: one letter-size page, front only; boxes split by hash first; 40–50 calls; a checklist of eight lines, including who signals and who spots. - College: the full two-sided sheet with 70–90 calls; add your code's ineligible-downfield limit beside every RPO, and a line in the checklist for the communication rules in your conference. - Pro: the full sheet plus the band; 100-plus calls; out-of-ten estimates on every third-down call. - Film room: pick one real team and one game. Build the third-down boxes you think its coordinator carried from that team's previous four games, then grade each of his third-down calls by process before you look at the result.
Next. In Chapter 26 you'll build the Situational Package: third down (short, medium, long), red zone, two-minute, and four-minute. It's the part of your sheet that decides close games.
Chapter Summary
A play call is a choice from a prepared menu, made in about six seconds of a forty-second clock, by one person who usually can't see the defense he's calling against yet. Everything a play caller does before Sunday exists to make that choice fast and sound: the call sheet sorted by what you'll know, two calls ready before the spot, a quiet window with one voice in it, and rules that speak for you after the radio dies at 15.
The threshold concept: a call is a bet on a distribution of outcomes, not a promise of one. Grade the decision by the information and alternatives the caller had, not by the single result.
The 40-second routine
| Play clock | Job | Owner |
|---|---|---|
| 40–34 | Result; booth read: shell, box, one thing | Teo (only voice) |
| 34–28 | Pick the ready call; personnel out | Play caller |
| 28–20 | Call in, ten words or the band number | Play caller → QB |
| 20–15 | Huddle, once; radio cuts at 15 | QB |
| 15–snap | Mike ID, motion, walk-ups, rules | QB, center, everyone |
The decision framework for any third down
- Situation first: down, distance, field zone, score, time, timeouts, and the head coach's fourth-down rule (what does failure cost?).
- Find the box; list the two or three defenses you might get, with rough odds.
- For each call in the box, estimate the outcomes out of ten: convert, punt, sack, turnover.
- Choose the call that's acceptable against every likely defense and very good against at least one. Prefer a fatter tail only when your rules are rehearsed for it.
- Make sure the call carries its hot and sight rules, because those are your voice after 15.
- After the snap: one letter (C, R, X), then the next play's reason.
Grading calls: two grades every Monday, result and process. Keep the "earned" and "bad beat" calls. Fix the "earned, unfortunately" calls. Throw out the "dumb luck" calls, even though they worked.
What's Next
Third-and-seven at our 41 was one situation. Chapter 26 takes the whole family: third down at every distance, the red zone where the field shrinks, two-minute and four-minute football, and the last drive of our season, from our 25 with just under four minutes left to the third-and-nine at our 38. I'll tell that drive the way I wish I'd planned it.