Chapter 14 Quiz: Expected Possession Value (EPV)

Instructions

This quiz contains 25 questions covering Expected Possession Value methodology and applications.


Question 1

What does EPV measure?

A) Points per game B) Expected points at any moment during a possession C) Defensive efficiency D) Player salary value

Answer: B

Explanation: EPV estimates the expected number of points a possession will yield at any given moment, updating continuously as the possession unfolds.


Question 2

EPV is an example of which type of analytical framework?

A) Box score analysis B) Spatiotemporal modeling C) Traditional statistics D) Contract valuation

Answer: B

Explanation: EPV uses spatial (player/ball position) and temporal (time remaining) data to model possession outcomes, making it a spatiotemporal framework.


Question 3

What is the typical EPV value at the start of a possession?

A) 0.0 points B) About 1.0-1.1 points C) 2.0 points D) 3.0 points

Answer: B

Explanation: League average offensive efficiency is approximately 1.0-1.1 points per possession, so EPV starts near this value at possession start.


Question 4

When does EPV increase during a possession?

A) Only when a shot is taken B) When the offensive team improves its scoring probability C) At fixed time intervals D) Never during a possession

Answer: B

Explanation: EPV increases when offensive actions (driving, passing to open shooter) increase scoring probability, and decreases when defense forces worse situations.


Question 5

The derivative of EPV with respect to time is called:

A) Points per second B) EPV Added or EPV velocity C) Defensive rating D) Usage rate

Answer: B

Explanation: EPV Added (or EPV velocity) measures how quickly a player is increasing or decreasing the team's expected points through their actions.


Questions 6-10

[Continue with questions about EPV calculation, tracking data requirements, modeling approaches, player evaluation with EPV, and limitations]


Question 11

Which data source is essential for calculating EPV?

A) Box scores only B) Play-by-play data only C) Player tracking (optical/chip) data D) Survey data

Answer: C

Explanation: EPV requires continuous spatial data about player and ball positions, which comes from optical tracking systems or player-worn sensors.


Question 12

When a player drives to the basket, EPV typically:

A) Stays constant B) Decreases C) Increases if the drive creates advantage D) Becomes negative

Answer: C

Explanation: Successful drives that create closer shots or defensive breakdowns increase EPV by improving expected scoring probability.


Question 13

EPV for a wide-open corner three is approximately:

A) 0.5 points B) 1.0 points C) 1.2-1.3 points D) 3.0 points

Answer: C

Explanation: Wide-open corner threes have ~40% make probability, yielding EPV of ~1.2 points (0.40 × 3 = 1.2).


Question 14

The main advantage of EPV over traditional plus-minus is:

A) Easier to calculate B) Attributes value to specific actions, not just outcomes C) Requires less data D) Works for all sports

Answer: B

Explanation: EPV can credit or debit specific actions (passes, drives, defensive rotations) based on how they change expected value, unlike plus-minus which only measures outcomes.


Question 15

What machine learning approach is commonly used to model EPV?

A) Linear regression only B) Gradient boosted trees or neural networks C) K-means clustering D) Principal component analysis

Answer: B

Explanation: Modern EPV models typically use gradient boosted trees or neural networks to capture complex, non-linear relationships between spatial configurations and outcomes.


Questions 16-20

[Continue with questions about defensive EPV, pass EPV, shot selection evaluation, and practical applications]


Question 21

If a player consistently takes shots with EPV lower than passing EPV, this suggests:

A) They are an excellent shooter B) They should pass more often C) They have high basketball IQ D) Their teammates are poor

Answer: B

Explanation: If available passes yield higher EPV than the shots taken, the player is making suboptimal decisions and should pass more.


Question 22

EPV models must account for which "counterfactual"?

A) What would happen if the player weren't on the team B) What alternative actions were available at each moment C) Historical averages from 10 years ago D) Playoff performance only

Answer: B

Explanation: To properly credit actions, EPV models compare actual choices to alternative options that were available (the counterfactual).


Question 23

The biggest limitation of EPV for player evaluation is:

A) It's too simple B) Requires expensive tracking infrastructure and complex modeling C) Only works for guards D) Ignores offense

Answer: B

Explanation: EPV requires tracking data (expensive), sophisticated modeling (complex), and significant computational resources, limiting its accessibility.


Question 24

EPV can be used to evaluate which of the following?

A) Shot selection quality B) Passing decisions C) Defensive rotations D) All of the above

Answer: D

Explanation: EPV provides a common currency (expected points) for evaluating all actions that affect possession value.


Question 25

When might EPV undervalue a player's contribution?

A) When they score a lot B) When their value comes from creating options that aren't taken C) When they play many minutes D) When they have high usage

Answer: B

Explanation: EPV credits realized actions; if a player creates valuable passing options that teammates don't utilize, their contribution may be understated.


Scoring Guide

  • 23-25: Excellent
  • 19-22: Good
  • 15-18: Satisfactory
  • Below 15: Review chapter