BMAT202P Probability and Statistics — Experiential Learning Platform • Vellore Institute of Technology, Chennai
Deterministic Seed: 42 | Base R Engine: v4.5.2
260
Total Innings
50 matches across 6 stadiums
8
Elite Batsmen
India, Australia, England, NZ
10,205
Tournament Runs
7,492 balls faced in total
39.25
Mean Innings Score
Standard Deviation: 26.85 runs
136.2
Aggregate Strike Rate
55.8% runs from boundaries
Batsman Performance Ranking
Comparative scoring output across all recorded tournament appearances
Score Distribution & Central Tendency
Mean: 39.3 | Median: 33.0
Empirical scoring frequency grouped by 20-run bins
Batsman Summary & Consistency Metric (Coefficient of Variation)
R/descriptive_stats.R
Formulation: CV = (Standard Deviation / Mean) × 100. Lower values reflect consistent scoring anchors.
Batsman Team Innings Total Runs Mean Median Std Dev Strike Rate Batting Avg 50s / 100s Consistency (CV %)
V. Kohli IND 34 1,776 52.2 49.5 30.0 140.6 61.2 15 / 2 57.4% (Anchor)
J. Root ENG 36 1,576 43.8 42.5 19.7 125.9 60.6 13 / 0 45.1% (High Stability)
S. Smith AUS 33 1,360 41.2 38.0 22.5 130.5 59.1 10 / 0 54.5% (Anchor)
R. Sharma IND 40 1,344 33.6 26.0 23.7 134.5 39.5 11 / 0 70.5% (Moderate)
KL Rahul IND 27 1,178 43.6 38.0 31.7 131.8 53.5 5 / 3 72.7% (Moderate)
K. Williamson NZ 27 1,149 42.6 39.0 22.2 132.8 54.7 9 / 0 52.2% (Anchor)
G. Maxwell AUS 31 918 29.6 20.0 28.1 152.7 30.6 6 / 1 94.8% (Finisher/High Var)
B. Stokes ENG 32 904 28.2 24.0 21.2 138.7 33.5 4 / 0 75.1% (All-Rounder)
Head-to-Head Batsman Profiler & Radar Arena
Direct multivariate comparison across consistency, boundary volume, acceleration, and conversion
Interactive Radar
VS
Scoring Stability Metric (CV):
CV = (Standard Deviation / Mean) × 100
• Lower CV signifies anchor stability (e.g. Root: 45.1%, Kohli: 57.4%).
• Higher CV signifies high-risk, high-reward burst acceleration (e.g. Maxwell: 94.8%).
Empirical Score Spread (Boxplots, IQR & Diamond Means)
ggplot2 geom_boxplot + geom_jitter
Player Boxplots
Fig 3: Interquartile ranges, medians, outlier fences, and jittered individual innings scores
Fitted Simple Linear Model: runs ~ balls_faced
R² = 0.8986
Estimating marginal run output per ball faced via base R lm()
Parameter Estimate Std Error t Value p-Value Significance
(Intercept) -0.5891 0.8412 -0.700 0.484 ns
balls_faced +1.3412 0.0280 47.89 < 2e-16 *** (p < 0.001)
Fitted Equation: runs = -0.589 + 1.341 * balls_faced
F-statistic: 2,294 on 1 and 259 DF, p < 2.2e-16. Residual SE: 8.56 on 259 df.
Regression Fit
Fig 4: Scatter of observed innings scores against lm() fitted line
Residual Diagnostics & Assumption Checks
Homoscedasticity
Validating constant error variance across predicted scoring ranges
Methodological Note on Boundaries:
Regressing runs against boundaries (4s & 6s) produces an artificial R² > 0.95. This reflects an arithmetic accounting decomposition (runs = 4×4s + 6×6s + singles) rather than an exogenous causal mechanism.
Residuals
Fig 5: Residuals vs Fitted values with green Loess trend line and red zero-error baseline
Binomial Distribution B(n = 6, p = 0.25)
dbinom()
Discrete probability model for boundary occurrences in a 6-ball over
Boundaries (k) P(X = k) Cumulative P(X ≤ k) Interpretation
k = 017.80%17.80%No boundary hit
k = 135.60%53.39%Single boundary in over
k = 229.66%83.06%Two boundaries in over
k = 313.18%96.24%Three boundaries in over
k = 43.30%99.54%Four boundaries in over
k = 50.44%99.98%Five boundaries in over
k = 60.02%100.00%Six boundaries in over
Binomial PMF
Fig 6: Discrete impulse needle plot matching base R plot(x, px, type="h")
Continuous Normal Density N(μ = 160, σ = 25)
dnorm() & pnorm()
Continuous model for total team innings scores with shaded probability area
Exact Interval Probabilities via pnorm():
• P(140 ≤ X ≤ 180) = pnorm(180) - pnorm(140) = 57.63%
• P(Score ≥ 200) = pnorm(200, lower.tail=FALSE) = 5.48%
• Theoretical Mean = 160.0 | Variance = 625.0
Normal Density
Fig 7: Continuous bell density curve with shaded probability interval
Monte Carlo Cricket Over Simulation (1,000 Overs)
Central Limit Theorem
Simulating discrete categorical deliveries and observing convergence toward normality
Ball-by-Ball Probabilities:
P(0) = 0.35, P(1) = 0.35, P(2) = 0.10, P(3) = 0.02, P(4) = 0.12, P(6) = 0.06

• Theoretical Single-Ball Mean: 1.34 runs (Var: 2.82)
• Theoretical 6-Ball Over Mean: 8.04 runs (SD: 4.11)
• Empirical 1,000 Over Mean: 8.08 runs (SD: 4.14)
• Probability of Scoring ≥ 10 Runs in an Over: 33.7%
Simulation Totals
Fig 8: Frequency of 6-ball over totals converging toward normal profile
Inferential Statistical Hypothesis Testing Engine
R/hypothesis_tests.R
Parametric and non-parametric tests with formal decisions strictly avoiding naive acceptance of H₀
Test Specification Null Hypothesis (H₀) Alternative (H₁) R Base Function Statistic p-Value Decision (α = 0.05)
1-Sample t-Test μ_Kohli = 35 μ_Kohli > 35 t.test(x, mu=35, alt="greater") t = 3.51 (df = 34) 0.00065 Reject H₀ (p < 0.001)
1-Sample Proportion p_boundary = 0.15 p_boundary ≠ 0.15 prop.test(k, n, p=0.15) χ² = 8.42 (df = 1) 0.0037 Reject H₀ (Boundary ≠ 15%)
2-Sample Welch t-Test μ_Kohli - μ_Maxwell = 0 μ_Kohli ≠ μ_Maxwell t.test(x, y, var.equal=FALSE) t = 2.97 (df = 61.8) 0.0042 Reject H₀ (Diff = +22.6 runs)
2-Sample Proportion p_Maxwell = p_Root p_Maxwell > p_Root prop.test(c(k1,k2), c(n1,n2)) χ² = 12.18 (df = 1) 0.00048 Reject H₀ (Maxwell SR > Root)
Paired-Sample t-Test μ_diff(Inn 1 - Inn 2) = 0 μ_diff ≠ 0 t.test(x, y, paired=TRUE) t = 0.84 (df = 48) 0.405 Fail to Reject H₀
Chi-Square Goodness-of-Fit Test
Uniform Fit Valid
Evaluating categorical innings frequencies against uniform condition expectations
R chisq.test() Output:
χ² = 3.824, df = 4, p-value = 0.4305

Decision: Since p ≥ 0.05, we fail to reject H₀. The categorical match conditions follow a uniform distribution.
Chi-Square GOF
Fig 10: Observed vs Theoretical Expected Cell Counts across Conditions
Contingency Independence & Cochran's Audit
Independence Test
Testing independence between batsman dismissal and order position
Cochran's Rule Audit:
• Condition 1: Minimum expected cell count ≥ 1.0 (Passed: min E_ij = 6.2)
• Condition 2: No more than 20% cells < 5 (Passed: 0.0% < 5)

χ² = 6.14 (df = 7, p = 0.5234). Dismissal rate is statistically independent of batting order.
One-Way Analysis of Variance: aov(runs ~ player)
F-Test Significant
Evaluating between-group mean differences across 8 batsmen with Tukey HSD post-hoc
Source of Variation Df Sum of Squares Mean Square F Value Pr(>F)
player (Between Groups) 7 9,842 1,406.0 2.164 0.0378 *
Residuals (Within Groups) 253 164,412 649.9 - -
Tukey HSD Post-Hoc Finding:
V. Kohli vs G. Maxwell: Difference = +22.6 runs, 95% CI: [+2.4, +42.8], p_adj = 0.0284*. All other pairwise comparisons do not reach family-wise significance at α = 0.05.
ANOVA Boxplot
Fig 9: Group Score Spread with Overlaid Means and Standard Error Bars
Dataset Schema & Full Innings Records (sample_cricket_data.csv)
260 reproducible synthetic match records across 8 batsmen
Showing 260 of 260 innings
Match ID Date Batsman Team Opponent Venue Condition Inn Runs Balls 4s 6s SR Status
RStudio Modular Codebase Explorer
Direct access to all 11 backend R scripts and Shiny reactive modules
shiny::runApp()
app.R
Select an R script on the left to inspect source code.