NBA · BASKETBALL STRATEGY & ANALYTICS

Aansh Jha

Data Analyst — turning play-by-play into decisions

I build the models, pipelines, and dashboards that help teams evaluate lineups, scout the transfer portal, and read a season one possession at a time. I currently validate league-wide hustle-stat data at the NBA, and previously built analytics infrastructure for UConn Men's Basketball.

2,073Possessions Backtested
31-ScriptReproducible Pipeline
R² 0.78MVP Prediction Model
40Dashboard Views Shipped
6,280Player-Seasons Clustered
19Opponents Auto-Scouted
2,073Possessions Backtested
31-ScriptReproducible Pipeline
R² 0.78MVP Prediction Model
40Dashboard Views Shipped
6,280Player-Seasons Clustered
19Opponents Auto-Scouted
aanshjha@gmail.com +1 (617) 682-5880 Jersey City, NJ medium.com/@SprtsStatsAJ

Box Score

Experience — 3 stops

Data Analyst

National Basketball Association (NBA)
Sep 2025 – Present
  • Analyze and validate live NBA hustle-stat data with the Basketball Strategy and Analytics team, ensuring accuracy and consistency of league-wide datasets used for official reporting and downstream team analytics
  • Run real-time quality control on game-level data, flagging statistical discrepancies and Basketball Operations events for review under strict accuracy and turnaround standards
  • Maintain and verify structured player- and game-level datasets distributed across all 30 NBA teams, supporting scouting, player evaluation, roster strategy, and contract-related analysis
  • Apply standardized league definitions and validation procedures to resolve data inconsistencies and improve reliability for league and team analytics personnel

Data Analyst Intern

UConn Men's Basketball
Jun 2024 – May 2025
  • Automated daily data pipelines in R (dplyr, tidyverse), tracking 12+ player performance metrics across 30+ games and cutting manual reporting time
  • Built shot-classification and expected-points models using logistic regression and feature engineering, powering pre- and post-game scouting reports adopted by the coaching staff
  • Developed an interactive R Shiny dashboard with practice shot charts (sportyR, ggplot2) so coaches could monitor shooting development and drill efficiency in real time
  • Analyzed 200+ lineup combinations using plus/minus, net rating, and opponent-adjusted efficiency to identify the highest-efficiency units and inform rotation decisions

Software Engineering Intern

Birdi Systems, Inc.
Jun 2023 – Aug 2023
  • Enhanced airport security monitoring systems by integrating backend features with Java, Maven, and API integration, improving detection accuracy and reliability
  • Delivered performance upgrades to asset management platforms through cross-functional work with software architects, aligning engineering output with client requirements

Featured Projects

Independent & applied work
31
Script reproducible pipeline

UConn Men's Basketball Analytics Platform

Oct 2024 – Apr 2026
  • Built a 31-script R pipeline ingesting, cleaning, transforming, and analyzing play-by-play data across 109 structured files, with automated validation at every stage
  • Built Bayesian hierarchical models in Stan to evaluate lineup and defensive efficiency, using partial pooling to quantify uncertainty in small samples
  • Deployed an R Shiny dashboard with 40 pre-built views for lineup optimization, player roles, scouting, and roster contingency planning
  • Backtested predictions across 2,073 possessions and 27 games, reporting Brier score, log loss, and calibration error
  • Automated scouting-report generation for 19 opponents across 32 games via ESPN ETL pipelines, producing shot profiles and matchup summaries
200+
Portal targets evaluated

CBB Transfer Portal Strategy & Roster Construction

Apr 2026 – May 2026
  • Built a data-driven player evaluation board integrating Sports-Reference, BartTorvik, and live portal status through API integration and automated cleaning
  • Developed position-specific fit models for stretch 5s, spacing wings, rim bigs, and secondary creators using feature engineering across shooting, efficiency, and defensive indicators
  • Aligned portal targets to specific budget and roster construction constraints
0.78
R² — MVP prediction model

Career Trajectories & Player Development Patterns of NBA Players (2010–2020)

Jan 2025 – Apr 2025 · STAT 3494W Research Seminar
  • Clustered 1,287 NBA players across 6,280 player-season records using K-Means and hierarchical clustering to identify long-term development trajectories
  • Built MVP-level performance prediction models using Random Forest regression, identifying key drivers through feature engineering and hypothesis testing
  • Developed reproducible Python dashboards (Jupyter, Quarto) to visualize career clusters and evaluation insights

Draft History

Education
AUG 2026 — DEC 2027

Boston College

M.S. in Sports Analytics
Coursework: Math for Machine Learning, Data Analysis, AI/ML Software Tools and Platforms, Econometrics, Sports Analytics, AI Algorithms I
AUG 2021 — AUG 2025

University of Connecticut

B.S. in Statistics
Coursework: Applied Linear Regression, Statistical Methods, Design of Experiments, Introduction to Data Science, Mathematical Statistics I & II, Statistical Programming (R, Python), Principles of Databases (SQL), Statistical Visualization, Data Science Capstone

Scouting Report

Skills & interests

Programming & Tools

RPythonSQLStanC / C++BashDockerGitTableauR Shiny

Statistical Modeling & ML

Bayesian Hierarchical ModelingPartial PoolingRegression (Linear/Logistic/Ridge/Lasso)Regularized APMCausal InferenceRandom ForestsXGBoostClustering

Data Engineering

Reproducible PipelinesETL / ELTDBTAPI IntegrationModel VersioningData Validation

Sports Analytics

Play-by-Play & Stint ModelingLineup OptimizationOpponent-Adjusted EfficiencyDefensive Scheme AnalysisPlayer EvaluationScouting

Languages

Hindi — NativeSpanish — IntermediateGerman — Beginner

Interests

Table Tennis — State & National ChampionBoston SportsWriting (Medium/Substack)ChessViolin