Statistical Analysis

Ping Pong Serve Precision Model

Samir Leonardo Caizapasto Hernández
ESPOL
January 2025
Updated: January 2026
309
Observations
k=3, p=0.3
Negative Binomial Model
r=0.65
Correlation

Executive Summary

This study analyzes ping pong serve precision through a controlled experiment with 309 observations. We validated a Negative Binomial model (k=3, p=0.3) and found significant patterns in participant performance.

Key Finding

The data fits the proposed model with p-value = 0.660, validating the Negative Binomial distribution hypothesis.

Study Objectives

General Objective

Analyze the statistical behavior of ping pong serve precision using a Negative Binomial model.

Specific Objectives

  • Validate the Negative Binomial model fit (k=3, p=0.3)
  • Evaluate differences in average serve time
  • Analyze relationships between categorical variables
  • Quantify correlation between attempts and time

Descriptive Analysis

General Statistics

Metric Value Description
Total Observations 309 Total number of experiments conducted
Mean Attempts 3.2 Average attempts until success
Mean Serve Time (s) 1.945 Average serve time in seconds
Std Dev Attempts 2.1 Variability in number of attempts
Std Dev Time 0.32 Variability in serve time

Categorical Variable Distribution

Dominant Hand

Right-handed 64.4%
Left-handed 35.6%

Serve Height

Above Waist 45.9%
Above Shoulders 34.6%
Below Waist 19.4%

Data Visualizations

Charts generated automatically by the R analysis pipeline

Attempts Distribution

Distribution of Attempts

Histogram showing the distribution of attempts until success. The red dashed line indicates the mean.

Serve Time Distribution

Serve Time Distribution

Distribution of serve time in seconds. Mean is significantly lower than the theoretical value of 2 seconds.

Attempts vs Time Correlation

Correlation Scatter Plot

Scatter plot showing the relationship between number of attempts and serve time. Moderate positive correlation suggests a fatigue effect.

Serve Time by Handedness

Boxplot by Handedness

Comparison of time distribution between right-handed and left-handed players.

Inferential Statistics

Statistical Test Results

Test Statistic P-Value Conclusion Significance
Goodness of Fit (Chi²) χ² = 0.193 0.660 Fits the model Not Significant
Time vs 2 seconds t = -3.18 0.002 Significantly less than 2s Significant
Attempts-Time Correlation r = 0.65 < 0.001 Significant correlation Very Significant

Conclusions

Key Findings

1

Model Validated

Data fits the Negative Binomial model (k=3, p=0.3) with p = 0.660

2

Temporal Efficiency

Average serve time (1.945s) is significantly lower than theoretical 2 seconds

3

Positive Correlation

Moderate correlation (r = 0.65) between attempts and serve time

4

Natural Distribution

Right-handed player proportion (64.4%) is consistent with population distribution

Practical Implications

  • Players tend to be more temporally efficient than expected
  • Fatigue or technique adjustment influences time as attempts increase
  • The Negative Binomial model is appropriate for predicting success patterns in this context

Samir Caizapasto

Junior Data Engineer & Analyst

Escuela Superior Politécnica del Litoral

Building automated data pipelines and transforming raw data into actionable insights.