Case Study 34.2 — Vic Maddox's Three Seasons, Read Through a Dashboard

Vic Maddox, his team, his players, and every number in this case are fictional. The public data source Owen uses — league play-by-play as distributed by nflverse — is real, and so are the methods.

Setup

The week after Game 1, Vic Maddox called me. You know about that call from Chapter 16: he'd seen the 52-yard touchdown, and he'd recognized his own pressure with my words on it. "Zero Thunder," he said. "Zero Thunder," I said. Then we talked for twenty minutes about everything except football, which is how old coaches apologize to each other.

Vic is not a cautionary tale because he was bad. He was brilliant. In his first season as a defensive coordinator, his defense blitzed on roughly 44 percent of dropbacks, led its conference in sacks, forced 31 takeaways, and made the playoffs. He interviewed for two head coaching jobs that winter. By his third season, his best corner had left in free agency, the quarterbacks in his division had quick releases, offenses had two seasons of his film, and he was fired with his defense near the bottom of the league in points allowed.

After the phone call I walked down the hall and asked Owen a question I'd wondered about for years.

Russ: If you'd had Vic's three seasons on your dashboard, would you have seen Year 3 coming in Year 1?

Owen: I can rebuild most of it. Play-by-play is public back to 1999 — EPA, success rate, explosives, takeaways, third downs, red zone. I can't rebuild pressure rate or coverage from play-by-play alone; that needs charting. For blitz rate, I'll use the numbers Vic used to quote you.

Russ: He quoted them a lot.

Owen: Then we'll have them.

It took him two evenings.

What happens

Owen's three-season dashboard

Vic's defense (fictional) Year 1 Year 2 Year 3
Blitz rate (Vic's own number) 44% 46% 47%
Sacks Led the conference Middle of the pack Lower half
Takeaways 31 (15 INT, 16 fumble recoveries) 21 (12 INT, 9 FR) 16 (9 INT, 7 FR)
Opponent fumbles → recovered 16 of 24 (67%) 9 of 20 (45%) 7 of 15 (47%)
Success rate allowed (EPA > 0) 45% 45% 46%
EPA per play allowed −0.03 +0.02 +0.08
Explosives allowed (runs 12+, passes 20+) 64 69 77
Points allowed Top half Middle Near the bottom

Owen also drew the two numbers that told the story on one page.

VIC MADDOX AS A COORDINATOR -- TAKEAWAYS AND EXPLOSIVES ALLOWED
Owen's rebuild from public-style play-by-play (fictional numbers).
Each # = 2 plays.

TAKEAWAYS (good for the defense)
  Year 1  ################               31
  Year 2  ###########                    21
  Year 3  ########                       16

EXPLOSIVES ALLOWED (bad for the defense)
  Year 1  ################################         64
  Year 2  ##################################       69
  Year 3  ######################################   77

SUCCESS RATE ALLOWED:   45%  ->  45%  ->  46%   (barely moved)

📊 Diagram (described): Two horizontal bar charts for Vic Maddox's three seasons, with each mark worth two plays. The first chart shows takeaways falling from 31 in Year 1 to 21 in Year 2 to 16 in Year 3. The second shows explosive plays allowed rising from 64 in Year 1 to 69 in Year 2 to 77 in Year 3. A line beneath both charts notes that success rate allowed barely moved, from 45 percent to 45 percent to 46 percent.

The conversation

Tess: His success rate barely moved. Forty-five, forty-five, forty-six. So he wasn't getting beaten more often.

Owen: Right. He was winning about the same share of snaps all three years. That's the first thing everybody misses. The defense didn't fall apart snap to snap. What changed was the size of the plays at the two ends — the ones that go his way and the ones that go against him.

Russ: Show me.

Owen did the arithmetic on the whiteboard, using rounded values from the chapter: a takeaway is worth roughly 4 expected points to the defense (it ends a drive and hands over the ball), and an explosive is worth roughly 3 to 4 to the offense.

  • Takeaways, Year 1 to Year 3: 31 down to 16 is 15 fewer. At about 4 points each, that's about 60 expected points the defense lost over a season.
  • Explosives, Year 1 to Year 3: 64 up to 77 is 13 more. At about 3.5 points each, that's about 45 expected points the offenses gained.
  • Together: about 105 expected points over roughly 1,050 snaps, or about 0.10 per play.
  • The actual change in EPA per play allowed: −0.03 to +0.08, which is 0.11 per play.

Owen: Almost the whole collapse is two rows. Fewer takeaways, more explosives. The every-down defense was the same defense.

Tess: And the takeaways — sixteen fumble recoveries out of twenty-four?

Owen: That's the tell. Two-thirds of his loose balls in Year 1. Most analysts find recovering a fumble is close to a coin flip. If he'd recovered half — twelve — he'd have had 27 takeaways, not 31. Still very good. But four of his Year 1 takeaways were bounces, and bounces don't come back on purpose. Year 2 he recovered 45 percent. That's not a decline. That's the coin landing normally.

Russ: So Year 1 was luckier than it looked, and Year 3 was about as unlucky as it looked.

Owen: Year 1 was a boom-or-bust defense that got the booms. It lived in the bottom-right box of the two-number map: wins a lot of snaps, gives up big plays. The booms covered the busts. In Year 3, the booms stopped and the busts didn't.

Why the booms stopped

The dashboard can show that the takeaways and sacks fell. Owen went looking for why, and every reason was one this chapter teaches.

  1. The coin landed normally. Fumble recoveries fell to about half, as they usually do. Nothing Vic did caused it, and nothing he could have done would have prevented it.
  2. The quarterbacks sped up. By Year 3, Vic's division had quick-release quarterbacks. Look at the timing chart in 34.4: a free blitzer from depth arrives at 2.4 to 2.8 seconds; a quarterback who releases at 2.4 throws before most blitzes land. A blitz rate of 47 percent against quarterbacks who get the ball out fast buys holes in the coverage without buying pressure in time. The sack total fell even though Vic blitzed more.
  3. Offenses had the film. You saw this in Chapter 18's case study, which happened in Vic's second season: an offensive coordinator named Dennis Okerlund charted 31 of Vic's "Avalanche" pressures, found 27 came from an identical walk-down look, and installed a protection and a quick tight-end seam for it. The second time the teams met, that seam went 58 yards for a touchdown at 2.2 seconds. A pressure's value comes partly from rarity. Vic's best pressures weren't rare anymore.
  4. The man coverage lost its best man. His top corner left in free agency. The explosives against man coverage — "lost in man," in Owen's first-cause codes — are the ones that grow when you play the same amount of man with a worse corner.

Russ: And the dashboard would have caught it?

Owen: Not with certainty. One season is a small sample for takeaways. But in Year 1, three things on the page were yellow even while the season was green. Explosives allowed were high — 64 is almost four a game. Fumble recoveries were far above half. And the blitz rate was so high that any change in the quarterbacks or the film was going to hit the whole defense at once. I'd have written: "Expect takeaways to fall toward 24 to 27. The explosives will still be here when they do."

Russ: He'd have thrown you out of his office.

Owen: Probably. He'd have been in first place.

The part Russ owns

The conversation didn't end there, because Tess asked the question a good quality control coach asks.

Tess: Coach, in Week 6 — were you calling his game or ours?

Russ: For one snap, his. The rest of the day, ours. That's the honest answer, and that's why the zero card is on the call sheet now.

Here's what I meant. Our defense isn't built like Vic's. We blitz on 27 percent of dropbacks, not 44. We make our living on the every-down numbers — a 40 percent success rate, explosives held to about three a game, a 9 percent missed-tackle rate — and we take our big swings in small, planned doses, the way Case Study 34.1 shows with Mug Cat Fire. Our dashboard is built to keep us in the "smothering" box, not to chase booms. When I called Zero Thunder on 3rd-and-8, I borrowed a boom-or-bust call for a smothering defense at exactly the moment a bust cost the most. Owen's receipt priced it at about a quarter of a point. Vic's three seasons show what happens when that quarter-point becomes the system.

Analysis

What the dashboard sees that the headlines don't. Vic's Year 1 headlines were sacks and takeaways. Those are real and valuable, and they are also the two noisiest numbers on any defensive page — sacks because they're a small slice of pressures, takeaways because recoveries are close to a coin flip and interceptions are rare. Success rate and explosives are steadier, and they told a different story: a defense that won snaps at an ordinary rate and gave up big plays at a high one.

Why regression isn't a criticism. When a number that depends partly on luck is extreme, the next season's number is usually closer to normal — analysts call this regression to the mean. It's not a moral judgment. It's what happens to counts like fumble recoveries. A coach who knows it plans for it; a coach who doesn't is surprised by it, and surprised coaches make desperate calls.

What changed and what didn't. Vic's every-down defense — the success rate — was nearly identical for three years. His results changed because the extremes changed, and the extremes changed for reasons the chapter teaches: luck reverting, quarterbacks getting faster than his pressure, and offenses learning his looks. A dashboard doesn't predict those things perfectly. It tells you which of your good numbers are standing on a coin.

Respect where it's due. Vic's approach won a playoff berth and two head coaching interviews. Volatility is a tool, and in a season where the booms come, it's a powerful one. The lesson isn't "never be Vic." It's "know which numbers are carrying you, and how much of them is luck."

Discussion Questions

  1. Owen could rebuild EPA, success rate, explosives, and takeaways from public play-by-play, but not pressure rate or coverage. Which of Vic's problems would charting data have shown earlier, and how?
  2. Vic recovered 16 of 24 fumbles in Year 1. Explain to a skeptical head coach why that number should lower, not raise, his expectations for Year 2 — without using the word "luck."
  3. Vic's blitz rate rose each year while his sacks fell. Using the timing chart in 34.4, give the mechanism.
  4. Is there a version of Vic's defense — still pressure-heavy — that would have aged better? What would you change on his dashboard and his call sheet in Year 2?
  5. Russ says the Week 6 call borrowed "a boom-or-bust call for a smothering defense." Explain what he means with the two-number map from 34.3.

Your Turn

Take any NFL defense from a recent season (or your own team's last three seasons if you have the charts). Using public play-by-play or your own charting, fill in four rows for at least two seasons: takeaways (with fumbles and recoveries split out), success rate allowed, explosives allowed, and EPA per play allowed (or, at the high school level, points per drive). Then write three sentences: which number is most likely to regress next season and why, which number is steadiest, and what one call-sheet change you'd make because of it.