BlackJack Smart Effect of Removal ML Predicting Expected Value Changes in Blackjack Through Card Removal Analysis

Goal

Predict the Expected Value (EV) changes in Blackjack based on card removal patterns from a 6-deck shoe. Your model should accurately estimate how removing specific combinations of cards affects the remaining deck's EV, providing a machine learning alternative to traditional combinatorial analysis.

The Challenge You are provided with data showing:

Number of cards removed for each value (1-10) The corresponding EV of the remaining deck Training data (20,500 rows): 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, ev 0, 1, 0, 1, 2, 2, 3, 0, 1, 3, -0.024345 1, 11, 5, 5, 12, 7, 8, 10, 9, 30, -0.024339 …

Test data (8,655 rows): 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 12, 6, 10, 7, 8, 10, 11, 5, 8, 43 11, 5, 9, 5, 7, 9, 9, 10, 8, 27 …

Evaluation Models are evaluated using Mean Squared Error (MSE). Successful solutions should balance accuracy with computational efficiency.

Here is the link to the competition: https://www.kaggle.com/competitions/black-jack-smart-effect-of-removal-ml