Poker
Poker has been a serious study for over five years. What started as curiosity became a habit once I realised poker is really a game of incomplete information and probability, the same territory I work in professionally. The analytical skills transfer remarkably well between data science and competitive play.
5+
Years of Study
GTO+
Primary Solver
EV
Decision Framework
10,000
simulations per calculation
I built a Monte Carlo calculator that estimates win probability for any two hands against any board. It runs the same style of simulation-based estimation I use professionally: play out the scenario thousands of times, read off the distribution.
Open the win-probability calculator →My Philosophy
I approach poker the same way I approach ML models: build a solid prior, update on new information, and stay aware of your own biases. The biggest edge isn't hand-reading genius. It's discipline, staying process-focused, and not letting one outcome distort your read on the next decision.
Study Approach
- Hand history review with PioSOLVER / GTO+
- Range construction and frequency analysis
- Monthly review sessions to spot recurring decision errors
- Theory reading: "The Mathematics of Poker" and "Modern Poker Theory"
- Solver work on key spots (3-bet pots, turn decisions)
The data science connection
Poker sharpened skills that apply directly to ML: comfort with uncertainty, expected value thinking, updating beliefs on new evidence, and keeping emotions out of analytical decisions. Both reward rigorous process over results-oriented thinking.
Recommended Resources
The Mathematics of Poker · Book
Modern Poker Theory · Book
Run It Once Training · Training Site
PioSOLVER · Software
Upswing Poker Lab · Training Site