Chapter 34 Exercises — Analytics and Defensive Decision-Making

Difficulty legend: ⭐ basic recall · ⭐⭐ application · ⭐⭐⭐ analysis and judgment · ⭐⭐⭐⭐ extension and research

All Sentinels, Outlaws, Stallions, and Ospreys numbers are fictional and match the chapter. Where an exercise uses expected-points values, use the rounded, illustrative values from 34.2 unless the exercise gives others. Selected answers are in Appendix J; short answers to Part C appear below each item.


Part A — Conceptual ⭐

A1. In one sentence each, define blitz rate, pressure rate, and time to pressure. Which of the three does the defensive coordinator control completely?

A2. What are Owen's three questions for any defensive number, and what fourth question does he always add?

A3. Explain expected points (EP) and expected points added (EPA) to a high school linebacker in no more than four sentences, without using the words "model" or "regression."

A4. Give the two common definitions of success rate and one situation where they might disagree on the same play.

A5. Using the square-root rule of thumb, how fuzzy is a percentage built from 25 snaps? From 144? From 400?

A6. What is the book's definition of an explosive play? How does Owen classify a 17-yard quarterback scramble, and why?

A7. Why are fumble recoveries considered a poor measure of a defense's skill, while forced fumbles are a better one?

A8. What is a turnover-worthy play (PFF's term), and how is Owen's "dangerous ball" similar and different?

A9. Name three numbers that are deliberately not on the front page of the Sentinels' weekly dashboard, and give the reason for each.

A10. What is "resulting," and why does Owen's receipt have separate rows for the result and the decision?


Part B — Film & Diagram Recognition ⭐⭐

B1. You're charting a Week 11 snap against St. Louis. Pre-snap, Kenny Vo is walked up inside the slot and both safeties are at 12 yards. At the snap Vo drops to the flat, the four linemen rush, and Holloway rotates down. Haugen throws at 2.7 seconds as Rollins hits him. How does Owen chart this snap for (a) blitz, (b) pressure, (c) the coverage family on the coverage table?

B2. Same snap as B1, but Bell had a Hug on the back: the back stayed in to block, Bell waited, then rushed and crossed the line at about 1.4 seconds, before the throw. How does that change the blitz charting under Owen's rule? How might another charting service code it differently?

B3. Here's a post-snap picture. Identify the coverage family and say which row of Owen's coverage table it belongs in.

POST-SNAP (B3)

      [DEEP HALF]                              [DEEP HALF]
           F                                        $
   ^                    ^                 ^                 ^
   |X                   |H                |Y                |Z
   C (trail)            * (trail)         D (trail)         C (trail)

                              M (on R)
               E     N     T     E
                         Q    R

📊 Diagram (described): Two safeties retreat to deep halves. Four receivers — X, H, Y, and Z — run vertical routes, and each is followed from underneath by a defender in trail position: a corner on X, the nickel on H, the dime back on Y, and the other corner on Z. The Mike linebacker is assigned to the running back. Four linemen rush.

B4. A broadcast graphic says a defense "leads the league with a 41 percent pressure rate." List three questions you'd need answered before comparing that number with the Sentinels' 34 percent.

B5. On the end-zone film of a Week 9 Ospreys drive, you clock six dropbacks: releases at 2.2, 2.4, 2.1, 2.6, 2.3, and 2.5 seconds; first-rusher arrival at 2.7, 2.5, 2.9, 2.4, 2.8, and 2.6. On how many snaps did pressure arrive before the release? What does that tell you about grading this pass rush by pressure rate?

B6. Code the first cause of this explosive: Austin play-action on 1st-and-10, Holloway takes two hard steps downhill toward the run fake, Harlan runs past him up the seam, and Royce, in the deep middle, arrives late; 27-yard completion. Which code from the 34.6 table, and which earlier chapter's fix applies?

B7. Code the first cause of this one: Ospreys 3rd-and-6, quick slant to the slot, caught at 5 yards; Vo was in position but lunged and missed; Ashby ran for 19 more. Which code, and which dashboard numbers does it move?

B8. Tess shows you two clips from the same game. Clip 1: a five-man pressure, the quarterback throws hot at 1.1 seconds for 6 yards on 3rd-and-9; nobody touches him. Clip 2: a four-man rush, Coleman pushes the guard into the quarterback's lap at 2.3 seconds, and the throw sails incomplete on 2nd-and-7. Chart each clip for blitz and pressure, estimate the sign of each play's EPA, and say which clip is the better advertisement for the call that was made.


Part C — Make the Call / Draw It ⭐⭐–⭐⭐⭐

C1. ⭐⭐ Compute EPA. Using the illustrative ladder in 34.2, estimate the EPA of each play for the offense: (a) 1st-and-10 at their own 25, a 15-yard completion to their own 40; (b) 1st-and-10 at our 35, a 15-yard completion to our 20; (c) 1st-and-10 at our 20, a sack for −8 to our 28 (treat 2nd-and-18 at our 28 as about +2.6). Which play hurt the defense most?

Answer

(a) Own 25 is about +0.8; own 40 is about +1.6 → +0.8. (b) Our 35 (the offense's "opp 35") is about +3.0; our 20 is about +4.0 → +1.0. (c) Our 20 is about +4.0; 2nd-and-18 at our 28 is about +2.6 → −1.4 (good for the defense). The same 15 yards was worth more near our goal line, which is why explosives in the red zone and short fields hurt most. Play (b) hurt the defense most.

C2. ⭐⭐ Make the call. Week 9 at Oklahoma City, 3rd-and-6 at the OKC 41, Ospreys 11 personnel 2x2. Owen's report: Oduya releases in about 2.4 seconds; against our four-man rush last season our pressure rate vs. Oklahoma City was about 19 percent; against off corners he completed 41 of 52 with no 20+ completions, against press 17 of 29 with three 20+ completions. Choose between Nickel Over Star Fire, Nickel Over Quarters with Boundary Hands, and Nickel Over Spin 1 Field Press. Justify with the numbers.

Answer

Nickel Over Spin 1, Field Press is defensible on 3rd-and-6, because the down is decided short — off coverage concedes the completion that converts. Press denies completions (59 percent against 79) at the cost of a few more explosives; with a post safety (Spin 1), the explosive risk is capped. Star Fire is the weakest choice: Oduya's 2.4-second release beats even our blitz pressure (median about 2.1 to 2.5 depending on who's free) on many snaps, and the fire zone leaves soft spots underneath for the quick game. Quarters with Boundary Hands is a reasonable second call — soft press to one side — but on 3rd-and-6 the full-coverage off side is a gift. Every sample here is a lean; the mechanism (release time versus arrival time) is the strongest evidence.

C3. ⭐⭐⭐ Situation bin. Owen's season coverage table shows plain Two at +0.06 EPA per dropback and Two Man at −0.12. A new assistant proposes replacing Two with Two Man on first-and-10. Write the two-paragraph response you'd give in the staff meeting.

Answer

Paragraph one: the season numbers aren't comparable, because the coverages are called in different situations. Two Man lives on third-and-long, where almost every coverage posts good numbers because short completions end in punts; plain Two lives on first-and-10 and in the red zone. Inside a situation bin, Two Man was our worst coverage on 3rd-and-7 to 10 (6 of 18 converted), because trail technique concedes the short throw.

Paragraph two: on first-and-10 the short throw is exactly what the offense wants, and man coverage turns backs to the quarterback against scrambles and invites rubs. If you want to test Two Man on first down, pull every first-down snap where we played it (probably a story-sized sample) and look at the film for a reason. Until then, the proposal is a selection effect dressed as a finding.

C4. ⭐⭐⭐ Diagram it. Draw Dime Over Two Man against an 11-personnel 2x2 set on 3rd-and-12, pre-snap and post-snap, with the sticks line at 12 yards. Label where trail technique concedes the throw and explain why that concession is acceptable on 3rd-and-12 but not on 3rd-and-8.

Answer

Pre-snap: two safeties at about 13 on the hashes; corners and the two slot defenders (nickel and dime) in inside, underneath leverage; Bell alone in the middle on the back; four rushers. Post-snap: safeties to deep halves; five underneath defenders trailing a step inside and under their men; the concession zone is roughly 5 to 10 yards — underneath and outside of each trailer. On 3rd-and-12 a completion in that zone ends short of the sticks and the offense punts; on 3rd-and-8 the same completion is a first down. The coverage didn't change. The sticks did.

C5. ⭐⭐⭐ The pressure trade. Owen's breakdown: five-man man-free pressures, 41 dropbacks, 14 pressures (34 percent), about +0.08 EPA per dropback; five-man fire zones with a lineman drop, 46 dropbacks, 22 pressures (48 percent). Your linebackers coach wants more man-free pressure "because it's simpler to teach." Make the case, with these numbers, for which five-man pressure family to grow.

Answer

Grow the fire zones with a lineman drop. They produced more pressure (48 percent against 34) and better coverage behind them, because three-deep, three-under zone keeps eyes on the quarterback and puts a dropper in the hot window. The man-free pressures were positive EPA for the offense: the pressure they added didn't pay for the man-coverage holes, especially with no underneath help. Simplicity matters — the fix is to teach one or two fire zones well (Chapter 13), not to keep calling a pressure family that loses. Both samples are in the 40s: plans, but keep checking.

C6. ⭐⭐⭐ Whose fault — the number. After three games your missed-tackle rate is 13 percent, up from 9. Tess's chart shows the extra misses are almost all "angle" misses by safeties in the alley, on runs that bounced outside. Before you schedule more tackling drills, what other cause should you check, and in which chapter's rules?

Answer

Check the run fits in front of the safeties. Bounced runs that reach the alley often begin with a fit error — an edge squeezing, a linebacker filling the wrong gap — that hands the safety a runner at full speed with a bad angle (Anchor C is exactly that). Check Chapters 5, 20, and 21: force and alley assignments, and whether the edge players are keeping the ball inside. If the fits are sound, then it's a tackling problem (Chapter 23's Track angle drill). Code the first cause before choosing the drill.

C7. ⭐⭐⭐ Rare events. Owen's report says Haugen threw 2 interceptions in 31 attempts against Cover 1 in the Stallions sample and 0 in 48 against quarters. A coach wants to "play more Cover 1 to get picks." Using the chapter's rules, explain what Owen would want to see instead.

Answer

Two interceptions is a story — rare events are the fuzziest numbers there are, and 2 in 31 versus 0 in 48 could easily reverse in the next sample. Owen would want the dangerous-ball counts upstream: how many throws did Haugen put where a defender had a realistic play against each coverage? He'd also want a mechanism on film (for example, Haugen staring down a single receiver against man). If dangerous balls are high against Cover 1, there's a lean worth building on; if they're similar, the picks were bounces.

C8. ⭐⭐⭐ Read the dashboard. Here is a three-week stretch from a fictional opponent's defense you're scouting:

Week EPA/play allowed Success rate allowed Explosives Pressure rate (4-man / 5+) Takeaways (fumbles recovered / fumbles)
1 −0.12 41% 2 30% / 38% 4 (3 / 3)
2 −0.09 43% 3 27% / 41% 3 (2 / 2)
3 −0.10 42% 2 29% / 39% 4 (3 / 4)

Their headline says "the best defense in the league." As the offensive coordinator attacking them — or as their own coordinator worried about next month — what do you see, and which number do you expect to change first?

Answer

The every-down numbers (success rate in the low 40s, pressure around 30 percent on four-man rushes) are good but ordinary. The EPA is excellent mainly because of takeaways: 11 in three games, including 8 fumble recoveries on 9 opponent fumbles. That recovery rate is far above the near-coin-flip most analysts find, so takeaways are the number most likely to fall, and when they do, EPA will climb toward ordinary. An offensive coordinator should attack them normally and protect the ball; their coordinator should stop treating the EPA as the defense's true level and look at what produces the success rate. Three weeks is a story-to-lean sample on everything here — but the direction of regression is predictable.


Part D — Judgment & Trade-offs ⭐⭐⭐

D1. Owen's receipt priced Russ's Cover 0 decision at "about a quarter of a point per snap." A fan says, "A quarter of a point is nothing — the call just didn't work." Write the reply, using expected value over a season and the win-probability layer.

D2. A head coach wants you to put sacks on the weekly dashboard because "sacks win games and the fans understand them." Make the case for pressure rate instead — and say whether you'd still track sacks somewhere and why.

D3. Your defensive line coach grades rushers on pressures. In two weeks your scramble yards allowed have doubled. Using Goodhart's law, explain the likely link and design a grading rule that rewards pressure without destroying rush lanes. Include a player-safety note about how far a rusher should go to "finish" on a quarterback.

D4. You're a high school coordinator with no film exchange for one opponent and a 140-snap sample on another from three games. How would you use Owen's confidence ladder differently for the two opponents? What would you put on the call sheet as a "plan," and what as a "lean"?

D5. A charting service you don't subscribe to publishes a coverage table for your next opponent's quarterback: "73 percent completions against man, 61 percent against zone." What three questions do you ask before you let that change the game plan — and what's the danger of using it at all?


Part M — Mixed / Interleaved ⭐⭐–⭐⭐⭐

M1. (Chapter 34 + Chapter 18) Write the six-line protection ledger for Nickel Over Mac Star Fire Thin against six-man protection (five linemen plus the back), then use the timing chart in 34.4 to say whether the free rusher, if any, arrives before a hot throw. What does the combination tell you about calling Thin against Vance?

M2. (Chapter 34 + Chapter 25) Chapter 25's illustrative mixing table said the offense converts 40 percent if you always play coverage and that the best mix was about one pressure in four. Using this chapter's Vance data (28 of 36 against five-plus), explain in plain words why the best mix shrinks to about one in five or six, and how you'd check on the dashboard whether your mix was being followed.

M3. (Chapter 34 + Chapter 6) Explain how the dangerous-ball counts for Vance (1 in 36 against pressure, 6 in 52 against rotation) are evidence for Chapter 6's threshold concept, "the shell is not the coverage."

M4. (Chapter 34 + Chapter 26) Chapter 26 bins third downs into 1–3, 4–6, 7–10, and 11+. Last season the Sentinels allowed 33 of 58, 26 of 54, 15 of 62, and 4 of 44 in those bins. Compute each rate. Then explain why a single season "third-down conversion rate allowed" of 36 percent can hide a real weakness, and which bin you'd watch first against Austin, whose trips dig/flood lives on 3rd-and-4 to 6.

M5. (Chapter 34 + Chapter 5 and Chapter 23) Last season's 18 explosive runs: 11 fit errors, 4 sound fits followed by a missed tackle, 3 good blocking. Russ's first season had a 14 percent missed-tackle rate; last season, 9 percent. Write a short paragraph arguing which coaching investment — fits or tackling — did more to cut explosive runs, and what extra data you'd need to be sure.

M6. (Chapter 34 + Chapter 16) Chapter 16 says volatility is a tool when you're losing and a tax when you're tied or ahead. Using the illustrative per-100 table in 34.8, describe a game state in which the six-man Cover 0 column, despite its worse expected points, might be the right call. What changes — the EPA or the win probability?


Part E — Research & Film Study ⭐⭐⭐⭐

E1. Build a public-data dashboard. Using the free nflverse play-by-play for one recent NFL season, compute for one defense: EPA per dropback and per designed run allowed, success rate allowed (EPA > 0), explosive plays allowed by the book's definition (runs 12+, passes 20+; decide and state how you'll treat scrambles), third-down conversion rate allowed by Chapter 26's bins, and red-zone touchdown rate allowed. Write one paragraph on which of your numbers you trust most and least, and why. (If you can't run the data yourself, do the same for three games by hand from the official play-by-play.)

E2. Time the race. Watch one full game's end-zone or broadcast film for one defense. On every dropback you can see clearly, record the number of rushers, the release time (stopwatch from snap to throw), and whether a rusher reached the quarterback's spot before the release. Compute blitz rate, your own pressure rate, and the median release time. Compare your pressure rate with a charting service's number for the same game, if available, and explain the difference in terms of definitions.

E3. Code the explosives. Find every explosive play (runs 12+, passes 20+) one defense allowed over three games. Watch each one and assign a first cause from the 34.6 table. Present the counts as a small table and write the "fixed by" column for that defense — what you'd install or drill first if you were its coordinator.