Pass Sonar Alternative

This example shows how to make an alternative to the pass sonar. The pass sonar was introduced by Eliot McKinley. The idea is to use a kernel density estimator (KDE) instead of a polar bar chart that is used in the usual pass sonar.

The steps to produce this are:

• convert the pass start/ end points to angles (radians) and distance

• convert to Cartesian coordinates so each pass start point is centered at coordinate x=0, y=0

• plot the kernel density estimate of the new pass end points on a square inset axes

• draw circles at intervals of ten, which that act as grid lines

• cut the kernel density estimate contours to the last circle

I tried a lot of different techniques, but couldn’t get it to work plot on polar axes so gave up. I eventually stumbled on this solution after reading a stats stackexchange post.

import matplotlib.pyplot as plt
import numpy as np
from matplotlib.patches import Circle
from scipy.stats import gaussian_kde

from mplsoccer import VerticalPitch, Sbopen, FontManager

# data parser, fonts and path effects for giving the font an edge
parser = Sbopen()
pitch = VerticalPitch(goal_type='box', line_alpha=0.5, goal_alpha=0.5)
fm_rubik = FontManager('https://raw.githubusercontent.com/google/fonts/main/ofl/'
                       'rubikmonoone/RubikMonoOne-Regular.ttf')

Load StatsBomb data

Load the Starting XI and pass receptions for a Barcelona vs. Real Madrid match for plotting Barcelona’s starting formation.

event, related, freeze, tactics = parser.event(69249)
# starting players from Barcelona
starting_xi_event = event.loc[((event['type_name'] == 'Starting XI') &
                               (event['team_name'] == 'Barcelona')), ['id', 'tactics_formation']]
# joining on the team name and formation to the lineup
starting_xi = tactics.merge(starting_xi_event, on='id')
# replace player names with the shorter version
player_short_names = {'Víctor Valdés Arribas': 'Víctor Valdés',
                      'Daniel Alves da Silva': 'Dani Alves',
                      'Gerard Piqué Bernabéu': 'Gerard Piqué',
                      'Carles Puyol i Saforcada': 'Carles Puyol',
                      'Eric-Sylvain Bilal Abidal': 'Eric Abidal',
                      'Gnégnéri Yaya Touré': 'Yaya Touré',
                      'Andrés Iniesta Luján': 'Andrés Iniesta',
                      'Xavier Hernández Creus': 'Xavier Hernández',
                      'Lionel Andrés Messi Cuccittini': 'Lionel Messi',
                      'Thierry Henry': 'Thierry Henry',
                      "Samuel Eto''o Fils": "Samuel Eto'o"}
starting_xi['player_name'] = (starting_xi['player_name']
                              .replace(player_short_names)
                              .str.replace(' ', '\n')
                              )
# filter only succesful passes the starting XI
event = event.loc[((event['type_name'] == 'Pass') &
                   (event['outcome_name'].isnull()) &
                   (event['player_id'].isin(starting_xi['player_id']))
                   ), ['player_id', 'x', 'y', 'end_x', 'end_y']]
# merge on the starting positions to the events
event = event.merge(starting_xi, on='player_id')
formation = event['tactics_formation'].iloc[0]

Plot the data

fig, ax = pitch.grid(endnote_height=0, title_height=0.08, figheight=14, grid_width=0.9,
                     grid_height=0.9, axis=False)
title = ax['title'].text(0.5, 0.5, 'Passes\nBarcelona vs. Real Madrid', fontsize=25,
                         va='center',
                         ha='center', color='#161616', fontproperties=fm_rubik.prop)
player_names = pitch.formation(formation, positions=starting_xi.position_id,
                               xoffset=[-15, -12, -12, -12, -12, -6, -12, -12, -12, -12, -12],
                               text=starting_xi['player_name'],
                               kind='text', va='center', ha='center', fontproperties=fm_rubik.prop,
                               fontsize=12, ax=ax['pitch'], color='#353535')
axs = pitch.formation(formation, positions=starting_xi.position_id, height=20, aspect=1,
                      xoffset=[-3, 0, 0, 0, 0, 6, 0, 0, 0, 0, 0],
                      kind='axes', ax=ax['pitch'])

# grid for evaluating the kernel density estimator
X, Y = np.mgrid[-50:50:100j, -50:50:100j]
grid = np.vstack([X.ravel(), Y.ravel()])

for position in axs:

    # set the inset axes to square, here I made the goalkeeper larger as they tend to kick longer
    lim = 50 if position == 1 else 40
    num_lines = 5 if position == 1 else 4
    axs[position].set_xlim(-lim - 1, lim + 1)  # added some padding for plotting the last circle
    axs[position].set_ylim(-lim - 1, lim + 1)
    axs[position].axis('off')

    event_position = event[event.position_id == position].copy()
    angle_position, distance_position = pitch.calculate_angle_and_distance(event_position.x,
                                                                           event_position.y,
                                                                           event_position.end_x,
                                                                           event_position.end_y)
    x = distance_position * np.cos(angle_position)
    y = distance_position * np.sin(angle_position)
    xy = np.vstack([y, x])
    kde = gaussian_kde(xy)
    density = kde(grid).T.reshape(X.shape)

    # note on vertical pitches the kde needs to be rotated and X, Y switched
    if pitch.vertical:
        contours = axs[position].contourf(Y, X, np.rot90(density, k=3), cmap='viridis', levels=50,
                                          zorder=2)  # switch Y and X as plotted on a vertical pitch
    else:
        contours = axs[position].contourf(X, Y, density, cmap='viridis', levels=50, zorder=2)

    # add circles at intervals of ten
    for i in range(num_lines):
        lw = 3 if i == num_lines else 1  # make the last circle thicker
        circ = Circle((0, 0), (i + 1) * 10, ec='#a19e9d', lw=lw, alpha=0.5, fc='None', zorder=3)
        circ_artist = axs[position].add_artist(circ)

    # clip to the last circle. borrowed from the following stackoverflow post:
    # https://stackoverflow.com/questions/50099960/draw-plot-only-inside-a-circle-with-matplotlib
    for contour in contours.collections:
        contour.set_clip_path(circ_artist)

plt.show()  # If you are using a Jupyter notebook you do not need this line

/home/docs/checkouts/readthedocs.org/user_builds/mplsoccer/checkouts/latest/examples/tutorials/plot_pass_sonar_kde.py:133: MatplotlibDeprecationWarning: The collections attribute was deprecated in Matplotlib 3.8 and will be removed two minor releases later.
  for contour in contours.collections: