Numbers & models

BookSharp

Read basketball statistics through the games behind the numbers.

Interactive study

Explore the example
Game 114.0Game 220.0Game 318.0Game 424.0Game 516.0
From the example below.

Why I’m interested

A threshold changes how we read a player's recent games. Three hits out of five looks convincing until you examine which games cleared the line and how widely the scores vary. I want basketball analysis to show that reasoning, with the observations beside the summary statistics.

Progress

The historical platform contains player-prop analysis, injury and data pipelines, and strategy code. The Rust example below calculates a mean, standard deviation, and threshold counts across five synthetic games. The next integration uses the original BookSharp analysis path.

See what the threshold changes

Choose a points threshold and inspect the same five synthetic games, their mean, and their spread.

Choose an example

Five synthetic games

Keep the five scores fixed and move the threshold. The hit count changes with the line, while the mean and standard deviation stay the same. The fraction describes these five observations.

Game 114.0Game 220.0Game 318.0Game 424.0Game 516.0
Mean
18.4
Above threshold
3 / 5
Standard deviation
3.44

Input

Game points: 14, 20, 18, 24, 16

Result

3 / 5 observations exceed 16.5
Mean: 18.40
Population standard deviation: 3.4409

Five synthetic games

Keep the five scores fixed and move the threshold. The hit count changes with the line, while the mean and standard deviation stay the same. The fraction describes these five observations.

Game 114.0Game 220.0Game 318.0Game 424.0Game 516.0
Mean
18.4
Above threshold
2 / 5
Standard deviation
3.44

Input

Game points: 14, 20, 18, 24, 16

Result

2 / 5 observations exceed 18.5
Mean: 18.40
Population standard deviation: 3.4409

Five synthetic games

Keep the five scores fixed and move the threshold. The hit count changes with the line, while the mean and standard deviation stay the same. The fraction describes these five observations.

Game 114.0Game 220.0Game 318.0Game 424.0Game 516.0
Mean
18.4
Above threshold
1 / 5
Standard deviation
3.44

Input

Game points: 14, 20, 18, 24, 16

Result

1 / 5 observations exceed 22.5
Mean: 18.40
Population standard deviation: 3.4409

Rust calculates the mean, population standard deviation, and observed threshold fraction from the synthetic scores. The next integration runs the sample through BookSharp's analysis path.

What comes next

Run the same sample through the BookSharp analysis path and compare its statistics with the Rust result.

Implementation and credits

How does a threshold change the observed hit rate across the same games?

Five synthetic games supply the points, mean, standard deviation, and observed threshold counts.

Choose a threshold and compare the games that clear it with the mean and spread.

Rust calculates the synthetic example. The original service retains its existing implementation.

A basketball analytics experiment with synthetic data for the public example.

Back to the collection