Case Study 1 — Week 15, the Fourth Quarter: Three Calls, Three Grades

The Austin Outlaws, Sacramento Sentinels, and their coaches are fictional composites. The NFL rules described are real.

Setup

Week 15 at Sacramento. Our second game against Russ Kimbrough's Sentinels this season, the one we lost 23–20. You've seen the moment the second half turned: 3rd-and-7 at our own 38, 6:51 left in the third quarter, tied 13–13. The double-A mug, Weller's walk at ten seconds, Coleman dropping into our hot window, Bell free. Minus nine. Punt.

This case study covers what came after. It's a quarter and a half of football, and three decisions in it are worth grading by the method of Chapter 25: by process first, then by result, never by result alone.

The cast on the headset: me on the sideline, calling plays; Teo Arriaga in the booth with the quiet window; Wade Harmon, our head coach, who owns fourth downs and timeouts; Marisol Vega, our analytics director, on Wade's channel with the fourth-down chart.

What the scouting said at halftime of that game (my halftime sheet, three lines): They're giving us the quick game, and they'll take it away. 94 is beating our left tackle. Twelve isn't working. My three changes: Trips with Sutton alone into the boundary and the tight end as #3 on a drag; chip Rollins with Ruffin; less 12 personnel. In the spring Russ told me his three halftime changes: corners pressed on about half our early downs, rotation from two-high on about half our early downs, and the mug corner fire zone live on 3rd-and-7 or longer. We'd just met the third one.

What Happens

Grade 1: The series after the sack (my own tilt test)

When Vance got up from that sack, I wanted to call five straight max-protect shots and make Russ pay for embarrassing my line. That's the panic shot from 25.9, the oldest tilt there is.

What I did instead was make one pencil mark on the result column of my call sheet: R. A rule had failed. I didn't know yet which one. That came on Wednesday, with Rome's stopwatch and the "6 blockers + 1 ball = 7" whiteboard. But I knew the concept hadn't failed and Vance's read hadn't failed. One letter told me what not to change for the rest of the game: the base offense.

Sacramento scored a touchdown on their next drive, with a little over two minutes left in the third: 20–13.

Process grade: I'll give myself a B. The one-letter routine kept me from tilting. But an R in the third quarter of a game should mean "change the rule now," and I didn't. I kept the same mug rule on the card for the rest of the game because you don't install a new protection on the sideline. That's defensible. It also meant that if Russ had called the mug again, I had no answer. He didn't.

Grade 2: Fourth-and-1 at our 36 (Wade's call)

Early in the fourth quarter, down 20–13, we had 3rd-and-4 at our own 33. Sacramento played Quarters with the backside corner locked on Sutton. We ran Flood. Harlan caught the flat for three. Fourth-and-1 at our 36.

On Wade's channel, Marisol's chart said go. I heard the decision when Wade came onto mine.

Wade: "Punting."

Me: "I like the go. Heavy's on Green."

Wade: "I heard you. Punting."

Eight seconds. That was the whole conversation, as it should be. Wade's reasoning, which he gave me on Monday, not on the headset: Sacramento's defense had been better than our offense for a quarter, and he'd rather make Russ's offense drive the length of the field than hand them a short field if we failed. Sacramento drove and kicked a field goal: 23–13.

Process grade: This is the hard one, because the result tempts you both ways. "We punted and they scored anyway, so the punt was wrong." That's resulting. "Marisol's model said go, so the punt was wrong." That's closer to an argument, but a model is one input, and the head coach owns the decision.

Here's the honest version. On fourth-and-1 near your own 36, most public fourth-down models of recent seasons lean toward going for it, and Marisol's did. Wade's input, that our offense was struggling, is real information the model doesn't have, but it's also the kind of feel that's easy to overweight after a sack. My grade: a defensible decision against the chart's lean, made by the right person in eight seconds. That's a C-plus to B-minus on process, with a result that tells you almost nothing. On Monday I said that once, in Wade's office, with the chart on the table. He said, "Write it down. If I'm wrong about this kind of fourth down, Marisol will show me in a month." That's the right answer from a head coach: not "you're wrong," but "show me." (We went for it 26 times that season and converted 15. Wade isn't a timid fourth-down coach. He just owns the call.)

Grade 3: The slant-and-go (my call)

We got the ball back at our 25, down 23–13. Three snaps got us to the 41. Then Teo's one thing:

Teo: "Two high. Six. Simmons pressed six of the last seven first downs."

That number was the whole call. A corner who presses on first down is playing the slant. A corner who's been taught by six straight first downs that our slant is coming is a corner who will bite on a slant fake. And a two-high shell with Royce responsible for the middle means the seam, if we run one, will carry him away from Gaines.

The call went in as a band number, because it was twelve words: "Eleven, Gun Trips Right, 60 Right, Z Sluggo, H Seam, Y Drag." Ruffin was on Rollins's side to chip him, which was halftime change number two.

POST-SNAP: Week 15, Q4 -- 1st-and-10, AUS 41, SAC 23-13 (Z Sluggo)

                                F (Royce): carries H's seam
                                        ^              ^ Z go
                                        | H seam       | 44 yds,
                                        |              | caught at
 C (Brennan)            *               |              | the SAC 25
                                        |    C (Simmons)
          E   T   N   T   E             |       \  jumps the
                                        |        \ slant fake
                    <------------------ Y drag    \
 X        LT  LG  C   RG  RT  Y         H          Z
     R (chips E, then out)   Q

📊 Diagram (described): Austin is in trips to the right: Gaines wide right, Powers in the slot, Harlan attached as the third receiver; Sutton is alone on the left. Ruffin chips Rollins, the left edge, before releasing. At the snap Powers runs straight up the seam, and Royce, the deep safety, carries him toward the middle. Harlan runs a drag underneath. Gaines takes two hard steps on a slant inside; Simmons, the pressed corner, jumps the slant, and Gaines turns up the sideline on the go. Vance throws it over the top. Gaines catches it at the Sacramento 25 and runs to the 15, 44 yards.

Forty-four yards to the Sacramento 15. Inside zone for five. Harlan on a stick for eight to the 2. Then "Jumbo, Under I Wing Left, Power Left," the goal-line call Wade had saved since Week 6, and Ruffin scored. 23–20, 9:24 left.

Process grade: Earned. The booth counted something real (six of seven), the call was built to use exactly that, the protection had the halftime chip in it, and the quarterback made the throw. This is the top-left box of the process grid, and the only thing to do with an earned call is keep the process that produced it. Notice that I don't keep the call itself as a habit. Russ saw it too.

The last possession: a postscript

We never scored again. On our last possession, 3rd-and-7 at the Sacramento 38, Sacramento was in Home, their two-high Quarters look (Russ's name, which I learned in the spring), and we threw Harlan up the seam. Holloway broke it up at the 20. Fourth-and-7 at the 38 is a 55-yard field goal try, beyond Evan Dorsey's range, so Wade punted and Sacramento took over at its 8. We didn't get it back with a chance.

Process grade: A good call against a great play. Harlan's seam against Quarters asks the safety to carry a 255-pound tight end and stay on top of him, and Holloway did it perfectly. That's a bad beat. Keep the process, check the execution (Harlan's release was clean; the ball was a touch inside), and don't let the replay talk you out of the concept.

Analysis

Line up the four grades and you get a quarter and a half of football that a results-only reviewer would describe as "the offense collapsed after the sack and scored once." A process reviewer sees something different:

Decision Result Process Lesson
Series after the sack (no tilt) Sacramento TD B The routine worked; the rule fix had to wait for the week.
4th-and-1 punt (Wade) Sacramento FG Defensible, against the chart's lean Argue once, on Monday, with evidence.
Z Sluggo 44 yards, then TD Earned Keep the process (the booth count), not the play.
Last 3rd-and-7 seam Broken up, punt Bad beat Keep the concept; check the throw.

The mechanism behind the earned call is worth naming, because it's the same one behind Anchor C. The defense's habit became the offense's input. Simmons's press on six of seven first downs was Russ's own halftime change, and Teo counted it in real time and handed it to me in the quiet window. Russ changed the price on our quick game, and we bought something else with the new price.

Discussion Questions

  1. The Week 15 rule fix waited until the following week. Was there any change Lyle could have made on the sideline after the R mark, without installing a new protection? What would it cost?
  2. Wade's fourth-and-1 reasoning ("their defense has been better than our offense for a quarter") is a feel input. How would you test whether that kind of input deserves weight against a fourth-down model?
  3. Why does Lyle say to keep the process behind the Sluggo but not the call itself?
  4. The last 3rd-and-7 is graded a bad beat. What evidence would change that grade to "bad process"?
  5. Which of the four decisions would a television broadcast most likely have graded wrong, and why?

Your Turn

Take one drive from a real NFL game this week, a fourth quarter you watched. For every snap, write two grades in two columns: result (success or not, using the common success-rate rule: 40 percent of the needed yards on first down, 60 percent on second, all of them on third or fourth) and process (was the call right for the situation and the likely defenses?). Then find the one snap where the two grades disagree most, and write three sentences explaining why.