calculations.matches_12_prob_win

Computes match-outcome probabilities (home win / draw / away win) by treating each side’s goals as independent Poisson random variables with means stats_goals_expected_home and stats_goals_expected_away. Joint probability of (h, a) is P(H=h) * P(A=a), summed over the 11×11 grid 0..10 × 0..10.

Lineage

flowchart LR
m9[/matches_9_expected_goals/] --> me[/calculations.matches_12_prob_win/]
me --> m15[/matches_15_strat_expected_value/]
me --> m17[/matches_17_strat_predicted/]
me --> mart[/mart.matches/]
click m9 "/calculations/matches_9_expected_goals/" "Open"
click m15 "/calculations/matches_15_strat_expected_value/" "Open"
click m17 "/calculations/matches_17_strat_predicted/" "Open"
click mart "/mart/matches/" "Open"

Columns

Column	Description	Formula
event_id	Bet365 event identifier.	passthrough
event_date	Event date.	passthrough
event_partition_key	ClickHouse partition key.	passthrough
score_total	Actual total goals (carried for downstream join).	passthrough
score_difference	Actual goal difference.	passthrough
stats_goals_expected_home	Expected home goals (carried).	passthrough
stats_goals_expected_away	Expected away goals (carried).	passthrough
stats_goals_expected_total	Expected total goals (carried).	passthrough
prob_home_win	Home-win probability.	round(sum(if(home_goals > away_goals, joint_prob, 0)), 4) over home_goals, away_goals ∈ 0..10, where joint_prob = poisson_pmf(stats_goals_expected_home, home_goals) * poisson_pmf(stats_goals_expected_away, away_goals)
prob_draw	Draw probability.	round(sum(if(home_goals = away_goals, joint_prob, 0)), 4)
prob_away_win	Away-win probability.	round(sum(if(home_goals < away_goals, joint_prob, 0)), 4)
prob_win_sum	Sanity check — should be ≈ 1.	round(sum(joint_prob), 4)

poisson_pmf is the project macro macros/poisson_pmf.sql.