Cricket Taught Me More About Data Science Than My Computer Science Degree

As an Indian tech enthusiast, I was destined to develop two obsessions: writing code and watching cricket. What I didn't expect was how much these two passions would converge, with my cricket obsession teaching me more about practical data analysis than four years of formal education.

It started innocently enough. I just wanted to predict match outcomes based on historical performance—a simple weekend project. Three years later, I'm running machine learning models on bowling patterns and can tell you the exact probability of Virat Kohli scoring a century based on whether he had a pre-match coffee. (The data suggests it's a statistically significant factor, by the way.)

Here are some surprising lessons cricket taught me about data:

Context is everything - A batting average of 45 is excellent, unless it's in a period where pitch conditions heavily favor batsmen. Similarly, business metrics mean nothing without contextual information about market conditions and competition.
Sample size matters - We cricket fans often declare a player "in form" after two good innings, ignoring the statistical insignificance of such a small sample. I've caught myself making the same mistake in A/B tests at work.
Correlation isn't causation - Just because India tends to win when Rohit Sharma eats vada pav before a match doesn't mean that's the causal factor (though I'm not ruling it out entirely).

The rabbit hole went deeper when I started analyzing bowling strategies. I built visualization tools to track ball placement, swing patterns, and batsmen's weaknesses—skills I later applied to visualize customer journey patterns at work. My colleagues were impressed, but little did they know I was basically repurposing code I'd written to analyze Jasprit Bumrah's yorker accuracy.

Cricket's complex scoring system and multitude of performance metrics make it the perfect training ground for data analysis. Where else can you find a sport with metrics like "economy rate," "strike rate," and the mysteriously named "Duckworth-Lewis-Stern method" that requires a mathematics degree to fully comprehend?

My cricket analytics hobby has given me a playground to experiment with techniques I'd be hesitant to try directly on work projects. I've tested various prediction models, visualization approaches, and even dabbled in computer vision to track fielder positioning—all skills that eventually benefited my professional work.

So if your manager ever questions your obsessive cricket watching during World Cup season, just explain that you're conducting advanced data research. And if India's performance causes emotional distress, that's just the occupational hazard of being a dedicated data scientist.

Share this article on X.com