Case Study 1: 2:41 — The Last Drive of the Season, Told Honestly

An Outlaws case (fictional composite). It failed. Lyle tells it because the situation, not the coverage, is the part most people miss.


Setup

The game: the divisional playoff at Sacramento. The Outlaws had come back from 17–10 to tie it at 17 on Vance's nine-yard boot to Harlan at 12:48 of the fourth quarter. The Sentinels kicked a field goal at about 7:20 to lead 20–17. We punted. They punted. And with just under four minutes left, Austin had the ball at its own 25, down three, with one timeout.

The plan's box: our game plan had a box for exactly this, which most plans forget: behind by three, about four minutes left, needing a field goal to tie. It's a hybrid. Four minutes is too much time to play pure two-minute — you'd be throwing sideline routes with time to burn — and not enough time to play a normal drive, where you might take three minutes to go 50 yards. The first line in the box was the range line: around the Sacramento 35, which gives Evan Dorsey a kick in the low fifties. From our 25, that was 40 yards away.

What the scouting said: Teo's "Kimbrough by situation" page had one row I'd circled before the game: "3rd-and-7+ vs. 2x2: stayed two-high, every snap since their bye." By the time we got to this drive, the count was 29 for 29. Our game-plan rule 1 for third-and-long said, "Against two-high on third-and-long, the post is there if the middle opens."

Personnel and look: every snap of the drive came from the same picture — 11 personnel, two-by-two, Harlan attached on the left, Sutton outside him on the left, Powers in the right slot, Gaines wide right, Ruffin in the backfield. The Sentinels showed their two-high "Home" look on every snap.


What happens

THE LAST DRIVE (Game 3), measured against the range line

 SAC 35  - - - - - - - - - - - - - - - - - - RANGE LINE (52-yd try)
          (27 yards from the third-down snap: never reached)
 AUS 38   3-9   2:41  Mills: post to Gaines .... Royce INT
 AUS 37   2-10        Smash, checkdown Ruffin +1 (Bell)
 AUS 37   1-10        back-shoulder to Sutton, incomplete
 AUS 32   2-3         Ruffin inside zone +5
 AUS 25   1-10  ~3:5x quick out to Powers +7 (Vo)

📊 Diagram (described): A ladder of the drive from the bottom up. Austin starts at its own 25 just under four minutes: a quick out to Powers gains seven to the 32; Ruffin's inside zone gains five to the 37 and a first down; a back-shoulder throw to Sutton falls incomplete; a smash concept ends in a one-yard checkdown to Ruffin; and on third-and-9 at the 38 with 2:41 left, Vance throws the post and Ike Royce intercepts. A dashed line at the Sacramento 35 marks the range line, 27 yards beyond the third-down snap. The drive never reaches it.

Snap 1 — 1st-and-10, AUS 25. Quick out to Powers. Kenny Vo tackled him after seven. The Sentinels played Quarters. Good start: the chains are moving and we haven't spent anything.

Snap 2 — 2nd-and-3, AUS 32. Ruffin, inside zone, five yards. First down at the 37. Sacramento was in Six — Cover 6, the half to the boundary.

Snap 3 — 1st-and-10, AUS 37. The back-shoulder to Sutton against single-high, the throw that had scored our first touchdown at the 9 in the second quarter. It hit his fingers and fell. Incomplete. The clock stopped, which was fine.

Snap 4 — 2nd-and-10, AUS 37. Smash against Quarters. The hitch and the corner were both carried; Vance checked it down to Ruffin, and Bell tackled him after one yard. Clock running.

Snap 5 — 3rd-and-9, AUS 38, 2:41, Sentinels 20–17. I called Mills, our post-dig: Gaines on the post at about twelve yards, Powers on the dig underneath it, Sutton on a go on the left, Harlan to the flat, Ruffin staying in to protect. Before the snap, the Sentinels showed two high safeties at thirteen yards. Two-high on third-and-long, for the thirtieth time.

At the snap, the strong safety on the left — the tight end's side — rotated down, and Ike Royce spun from the right side into the deep middle. Single-high. Cover 3. Vance had read two-high and was already throwing the post at about 2.4 seconds. Royce was standing in it. Interception, about twenty-two yards downfield at the Sacramento 40, returned to the 46.

We had one timeout. Sacramento ran three times, got a first down, and took knees. Season over.

On the sideline, before I could say anything, Vance said, "Rule four, Coach." Rome's scramble rule four — the free square: when the middle is in doubt, throw the ball you know. He was owning it. I wasn't ready to let him own it alone.


Analysis

Most retellings of this drive are about the coverage: the 29 straight two-high snaps, the 30th that rotated, the quarterback baited by post-snap rotation — which our own sample had flagged as his weakness. All true. Chapter 25 deals with that decision as a call. This case is about the situation, because there's a lesson in it that would hold even if the safety had stayed put.

The price list changed; the call didn't

Here's what each result of that third-and-9 was worth, in that spot.

Result What happens next How bad
Conversion (9+ yards) First down near midfield; ~2:30; one timeout plus the two-minute warning; about 15–20 yards to the range line The goal
Short completion (1–8 yards) Fourth-and-short to fourth-and-long around our 40–45; Wade decides; the clock is running Bad, but alive
Incompletion Fourth-and-9 at our 38, clock stopped at ~2:35; Wade decides between a long fourth down and a punt with one timeout and the warning left Bad, but alive
Sack Fourth-and-long from our 30-something Very bad, still technically alive
Interception Opponent ball, one timeout to stop them with Season over

In the second quarter of a regular-season game, the gap between "incompletion" and "interception" on a third-and-9 is a few points of expected value and a possession. At 2:41 down three in a playoff game, the gap between them is the gap between alive and done. The cost of the worst outcome had multiplied, and the call I made had the same chance of that outcome as it did in any other quarter. A post thrown into the deep middle is the throw most likely to be intercepted if the quarterback reads the middle wrong. On that snap, it was the throw most likely to end the season.

That's what I mean when I say situational football is pricing the results before you call the play. I priced the conversion. I didn't price the failure.

"If the middle opens" is a condition, not a promise

Our game-plan rule said the post is there if the middle opens. That "if" was supposed to be a read. But the whole week, and the whole season, the middle had opened every time on that down, so the "if" had quietly become "when." The call didn't have a built-in answer for the middle closing — no tag that said, "if a safety spins to the middle, the ball goes to the dig or the flat." That answer is what I spent the offseason building: the middle-read tag, a post-snap middle-of-the-field read on every third-and-long call. You can call it a coverage lesson. I think of it as a situational one too, because in that situation, the wrong read had to be survivable.

What a situational call would have looked like

There were other ways to attack third-and-9 there, and I'll name them without pretending any was guaranteed.

  • A call whose bad outcome is an incompletion. A throw to the sideline — the go or back-shoulder to Sutton, a deep out — fails as an incompletion far more often than as an interception. Wade then makes a fourth-down call with the clock stopped and our timeout in hand.
  • A call that gets half of it. If Wade had told us before the series that we were in four-down territory, third-and-9 becomes two downs to get nine. Then a seven-yard dig is a good result, not a failure.
  • Mills with the middle-read tag — the same call with the answer to rotation built into it. That's next season's version.

None of those was obviously right. The point is that I didn't weigh them against the situation. I weighed them against the tendency, and the tendency was 29 for 29.

The fourth-down letter

Go back to the rule from the chapter: the fourth-down answer belongs in the third-down call. I didn't say a letter before that snap. If I had, the conversation would have gone, "Wade, if this doesn't convert, what are we doing on fourth-and-9 at our 38 with 2:35 and a timeout?" And whatever he answered would have changed what I called. If he says "punt," then third down is the last real chance and the call has to convert. If he says "go," then third down only has to get close, and the interception is the only result that truly kills us — so you choose the call least likely to throw one.

Russ's side

At the second-morning breakfast in the spring, Russ told me the rotation on that snap was their first on third-and-long against two-by-two since their bye. They'd been banking it. I told him, "You let us get good at the wrong test." He laughed. He'd earned the laugh.


Discussion questions

  1. Lay out the "price list" for the same third-and-9 at the same spot, but in the middle of the second quarter, tied 7–7. Which results change value the most, and why?
  2. Vance said "Rule four, Coach." Lyle says he "wasn't ready to let him own it alone." Divide the responsibility for the interception among the quarterback, the play-caller, and the game plan. Defend your split.
  3. The drive's plan said the range line was the first goal. At the third-down snap, the range line was 27 yards away. Does that make Mills, a deep throw, more defensible or less? Consider what a nine-yard conversion would have left.
  4. If Wade had said "go" before the snap — four-down territory — name a call you'd prefer to Mills, and explain how its failures differ.
  5. Russ banked a rotation for 29 snaps. What does that tell you about how much a defense will pay to create a situation in which one wrong read is fatal?

Your turn

Take one third-and-long call from your Situational Package (the Project Checkpoint). Write its price list twice: once for the second quarter of a tied game, and once for 2:40 left, down three, at your own 38, one timeout. For each, list the five results (conversion, short completion, incompletion, sack, interception) and what happens next. Then add one tag to the call — a post-snap rule — that makes its worst outcome less likely in the second setting without ruining it in the first.