Outcome flow

A Sankey-style view of every delivery: from the restart, through its delivery_outcomes() category, down to whether it ended in a goal. Every other chart in this gallery ranks, compares, or maps a set of numbers – this one shows a funnel: how much of the traffic at the left edge actually survives to a goal at the right edge.

Ribbon width is proportional to delivery count at every stage, so a category that rarely converts (most of them) reads as a wide band flowing straight into “No goal,” while anything that does convert shows up as a visible diagonal sliver breaking off toward “Goal.”

from pathlib import Path

from wa_setpieces import load_matches
from wa_setpieces.core.outcomes import delivery_outcomes
from wa_setpieces.viz.plots import plot_outcome_flow

try:
    _here = Path(__file__).resolve().parent
except NameError:
    _here = Path.cwd()
DATA = _here.parent / "tests" / "data" / "sample_match.json"

events = load_matches(DATA)

Corners in this sample match: no goal came from any of them, so every band flows straight through to “No goal” with no diagonal split – a real (if unglamorous) finding a bar chart of counts wouldn’t make as immediately obvious.

corner_outcomes = delivery_outcomes(events, "corner")
fig, ax = plot_outcome_flow(corner_outcomes, title="Corner outcome flow")

Free kicks, the same way:

fk_outcomes = delivery_outcomes(events, "free_kick")
fig, ax = plot_outcome_flow(fk_outcomes, title="Free-kick outcome flow")