Match timeline and value flow

Two ways of looking at when things happened, not just how many. First, every set piece in the match on one shared timeline – one row per restart type, one dot per delivery, positioned by match minute. Good for spotting patterns like a team leaning on throw-ins early or corners piling up late.

Second, cumulative set-piece added value over the match – a step chart showing when each team’s set-piece threat was actually created, not just the final total.

from pathlib import Path

from wa_setpieces import load_matches, XTModel
from wa_setpieces.viz.plots import plot_match_timeline, plot_set_piece_value_flow

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

events = load_matches(DATA)

fig, ax = plot_match_timeline(events)

Fitting on one match, as below, is illustrative only – see By value model on why a real xT model needs a season’s worth of data. Corners and free kicks are the only two types with a value model, so they’re what gets combined here by default.

model = XTModel.fit(events)
fig, ax = plot_set_piece_value_flow(events, model)