Quiz: Special Teams Analytics

Target: 70% or higher to proceed.

Section 1: Multiple Choice (1 point each)

1. What is the approximate field position value per 10 yards?

  • A) 0.1 expected points
  • B) 0.4 expected points
  • C) 1.0 expected points
  • D) 2.0 expected points
Answer **B)** 0.4 expected points *Explanation:* Each 10 yards of field position is worth approximately 0.4 EP.

2. Why is raw field goal percentage a flawed metric?

  • A) Kickers don't try on every kick
  • B) It doesn't account for distance or conditions
  • C) Field goals are worth different points
  • D) The NFL doesn't track it properly
Answer **B)** It doesn't account for distance or conditions *Explanation:* A 50-yard FG is much harder than a 25-yarder; raw FG% treats them equally.

3. What is "FG over expected"?

  • A) Field goals made minus expected makes based on distance
  • B) Field goals longer than 50 yards
  • C) Extra points converted
  • D) Kicks in overtime
Answer **A)** Field goals made minus expected makes based on distance *Explanation:* This adjusts for difficulty by comparing actual makes to expected based on kick distance.

4. What is the typical league FG% from 41-50 yards?

  • A) 95%
  • B) 88%
  • C) 75%
  • D) 55%
Answer **C)** 75% *Explanation:* Kicks from 41-50 yards convert at approximately 75% league-wide.

5. Which punting metric best captures field position value?

  • A) Gross average
  • B) Net average combined with inside-20 rate
  • C) Hangtime
  • D) Number of punts
Answer **B)** Net average combined with inside-20 rate *Explanation:* Net average accounts for returns, and inside-20 rate measures pinning ability.

6. Why is year-to-year kicker FG% correlation relatively low (~0.35)?

  • A) Kickers get injured frequently
  • B) Small sample sizes create high variance
  • C) Teams switch kickers often
  • D) The rules change every year
Answer **B)** Small sample sizes create high variance *Explanation:* With only ~30 attempts per season, significant variance exists that isn't skill-based.

Section 2: True/False (1 point each)

7. A kicker who makes 25/30 field goals (83.3%) is definitively better than one who makes 22/28 (78.6%).

Answer **False** *Explanation:* Without knowing distances, conditions, and given sample size variance, we cannot conclude one is better. The difference could be entirely explained by luck and difficulty.

8. Touchback on kickoffs gives the opponent the ball at the 25-yard line.

Answer **True** *Explanation:* As of 2016, touchbacks on kickoffs result in the ball being placed at the 25-yard line.

9. Punt return average is more reliable year-to-year than kick return average.

Answer **False** *Explanation:* Both have low year-to-year correlations (~0.20-0.25), though punt return is slightly lower due to fewer opportunities.

Section 3: Short Answer (2 points each)

10. Explain why special teams plays are considered high-leverage despite being only ~17% of total plays.

Sample Answer **High-leverage reasons:** 1. **Field position swings**: A single punt can change field position by 50+ yards, worth ~2 EP 2. **Scoring plays**: Return TDs, blocked kicks are worth 6-7 points directly 3. **Close game impact**: 78% of games within 7 points have a decisive ST play 4. **Discrete outcomes**: FG make/miss is binary with large EP swings 5. **Low frequency, high variance**: Each play has outsized impact potential A blocked punt for a TD can swing a game by 10+ points in a single play.

11. What metrics would you use to evaluate a punt coverage unit (not the punter)?

Sample Answer **Coverage unit metrics:** 1. **Average return yards allowed**: Lower is better 2. **Fair catch rate forced**: Higher indicates good coverage speed 3. **Long return rate**: Percentage of returns over 15/20 yards 4. **Tackles made inside 20**: Pinning ability 5. **Return TD rate**: Should be near zero 6. **Net minus gross**: Difference attributable to coverage **Formula:** 
Coverage Score = fair_catch_rate * 100
              - avg_return * 2
              - long_return_rate * 50
              - return_td_rate * 700
Key insight: Coverage is a team metric, not individual.

Section 4: Application (3 points each)

12. A team is at the opponent's 35-yard line, 4th and 3. Calculate the expected value of attempting a 52-yard FG vs punting vs going for it.

Sample Answer **Analysis:** **Field Goal Attempt (52 yards):** - Expected make probability: ~60% - Make: +3 points, opponent gets ball at 25 = +3 - 0 EP ≈ +3.0 - Miss: opponent gets ball at 42 = -0.6 EP - FG EV = 0.60 * 3.0 + 0.40 * (-0.6) = 1.8 - 0.24 = **+1.56 EP** **Punt:** - Expected net: ~35 yards - Opponent starts at own 5: EP = -0.5 - Punt EV = -(-0.5) = **+0.5 EP** **Go for it:** - Conversion probability (4th & 3): ~55% - Convert: New 1st down at 32, EP ≈ +2.0 - Fail: Opponent at 35, EP = -0.6 - Go EV = 0.55 * 2.0 + 0.45 * (-0.6) = 1.1 - 0.27 = **+0.83 EP** **Recommendation:** Attempt the field goal (+1.56 EP) Note: This analysis is simplified; actual decision depends on game situation, kicker accuracy, etc.

13. How would you design an analysis to determine if a kicker performs better or worse under pressure?

Sample Answer **Pressure Kicking Analysis:** **Step 1: Define pressure situations** - 4th quarter, margin ≤ 7 points - Final 2 minutes, any margin - Playoff games - Kicks that would tie or take lead **Step 2: Calculate metrics** 
pressure = fgs[
    (fgs['qtr'] == 4) &
    (abs(fgs['score_differential']) <= 7)
]

non_pressure = fgs[~pressure_condition]

pressure_fg_pct = pressure['made'].mean()
normal_fg_pct = non_pressure['made'].mean()
clutch_differential = pressure_fg_pct - normal_fg_pct
**Step 3: Sample size considerations** - Require minimum 10 pressure kicks - Calculate confidence intervals - Use Bayesian approach for small samples **Step 4: Year-over-year analysis** - Does clutch performance persist? - Correlation between years **Key finding (likely):** Clutch performance shows little persistence - it's mostly variance, not skill.

Section 5: Critical Thinking (2 points)

14. Why might a team's special teams EPA be misleading as an evaluation metric?

Sample Answer **Potential issues with team ST EPA:** 1. **Sample size**: Only ~150-200 ST plays per team per season 2. **Return TD variance**: One or two return TDs massively skew EPA 3. **Opponent quality**: Facing elite return specialists affects results 4. **Kicker/punter dominance**: Individual specialists drive team numbers 5. **Situational bias**: Some teams have more/fewer FG opportunities **Example:** - Team A: 0 return TDs allowed, solid coverage = +15 ST EPA - Team B: 1 return TD allowed, same coverage = +8 ST EPA - That single play swings 7 EPA despite same overall quality **Better approach:** - Separate components (kicking, punting, returns, coverage) - Regress toward mean for rare events (return TDs) - Account for opportunity differences

15. What are the key differences between evaluating a kicker and evaluating a punter?

Sample Answer **Kicker evaluation:** - **Binary outcomes**: Make or miss - **Distance as primary difficulty**: Can model expected % - **Points directly scored**: Clear value - **Higher sample stability**: More consistent year-to-year - **Metrics**: FG%, FG over expected, EPA **Punter evaluation:** - **Continuous outcomes**: Net yards on spectrum - **Situation-dependent**: Different goals by field position - **Field position value**: Indirect point impact - **Coverage unit interaction**: Team affects outcomes - **Metrics**: Net average, inside-20%, hangtime (if available) **Key differences:** 1. Kickers have clearer individual attribution 2. Punters depend more on coverage unit 3. Kicker value is direct (points), punter is indirect (field position) 4. Situational demands differ (punter from own 10 vs opp 40) 5. Punter evaluation needs more context (starting position) **Both share:** Small samples, year-to-year variance, environmental effects

Scoring

Section Points Your Score
Multiple Choice (1-6) 6 ___
True/False (7-9) 3 ___
Short Answer (10-11) 4 ___
Application (12-13) 6 ___
Critical Thinking (14-15) 4 ___
Total 23 ___

Passing Score: 16/23 (70%)