{% extends "base.html" %} {% block title %}Model Info — Cricket AI{% endblock %} {% block content %}
Technical details about the four XGBoost classifiers powering this system.
Predicts whether the next delivery will be a dot ball (no run scored). Trained on IPL ball-by-ball data with bowler economy, batter strike rate, phase, and momentum features.
Predicts the probability of a boundary (4 or 6) being scored. Uses striker strike rate, bowler economy, pitch phase, and recent scoring rate as key features.
Multi-class classifier outputting probability distribution over outcomes 0–5 runs per ball. Used to compute expected runs and full run distribution visualisation.
Predicts win probability for the chasing team in the 2nd innings. Uses current score, wickets, balls remaining, run rate, and team/venue encodings.
{
"batting_team" : "Mumbai Indians",
"bowling_team" : "Chennai Super Kings",
"venue" : "Wankhede Stadium",
"innings" : 1,
"over" : 14,
"ball_in_over" : 3,
"current_score" : 110,
"wickets_fallen": 2,
"batter_sr" : 148,
"bowler_eco" : 7.4,
"last_6_runs" : 11,
"last_12_runs" : 19
}
{
"dot_ball_prob" : 24.3, // % probability of dot ball
"boundary_prob" : 38.7, // % probability of 4 or 6
"expected_runs" : 2.41, // weighted mean runs
"run_distribution": [0.24, 0.22, 0.08, 0.06, 0.28, 0.12],
"win_probability" : null, // populated in innings 2
"phase" : "Middle Overs (Ov 7-15)",
"run_rate" : 7.86
}