Tutorial for plotting data using the hockey-rink library
Intro
Use the chickenstats library to scrape play-by-play data and plot shot events using the hockey-rink library.
Parts of this tutorial are optional and will be clearly marked as such. For help, or any questions, please don't hesitate to reach out to chicken@chickenandstats.com or @chickenandstats.com on Blue Sky.
Housekeeping
Import dependencies
Import the dependencies we'll need for the guide
from pathlib import Path
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import seaborn as sns
from hockey_rink import NHLRink
import chickenstats.utilities
from chickenstats.chicken_nhl import Scraper, Season
from chickenstats.chicken_nhl.helpers import norm_coords
from chickenstats.chicken_nhl.info import NHL_COLORS
from chickenstats.chicken_nhl.helpers import charts_directory
Pandas options
Sets different pandas options. This cell is optional
Folder structure
Chickenstats matplotlib style
chickenstats.utilities includes a custom style package - this activates it. This cell is also optional
Scrape data
Schedule and game IDs
Scrape the schedule and standings using the Season object. Then, create a list of game IDs to scrape
Play-by-play
Scrape the play-by-play data for the list of game IDs. First instantiate the Scraper object,
then call the play_by_play attribute
Stats
Aggregate data using the Scraper object's built-in methods
scraper.prep_stats(level="season", disable_progress_bar=True)
stats = scraper.stats.reset_index(drop=True)
scraper.prep_lines(level="season", disable_progress_bar=True)
lines = scraper.lines.reset_index(drop=True)
scraper.prep_team_stats(level="season", disable_progress_bar=True)
team_stats = scraper.team_stats.reset_index(drop=True)
Scatter plots
Drawing the rink
Drawing the rink with the NHLRink class from the hockey-rink library
Filter conditions
Set the team, strength state, minimum TOI, and maximum number of lines to plot, then filter the data
team = "NSH"
strength_state = "5v5"
toi_min = 12
max_lines = 3
conds = np.logical_and.reduce([lines.team == team, lines.strength_state == strength_state, lines.toi >= toi_min])
plot_lines = lines.loc[conds].sort_values(by="toi", ascending=False).head(max_lines).reset_index(drop=True)
Top-N forward line combos
Plot the top-N forward line combinations at the chosen strength state, from the prior cell
# Setting overall figures
fig, axes = plt.subplots(nrows=2, ncols=max_lines, dpi=650, figsize=(16, 8))
fig.tight_layout(pad=1.5)
axes = axes.reshape(-1)
for row, line in plot_lines.iterrows():
ax = axes[row]
ax_zone = "dzone" if row > 5 else "ozone"
rink.draw(ax=ax, display_range=ax_zone)
shot_events = ["MISS", "SHOT", "GOAL"]
plot_conds = np.logical_and.reduce(
[
pbp.forwards == line.forwards,
pbp.forwards_api_id == line.forwards_api_id,
pbp.strength_state == strength_state,
pbp.event.isin(shot_events),
]
)
plot_data = pbp.loc[plot_conds].reset_index(drop=True)
plot_data = norm_coords(data=plot_data, norm_column="event_team", norm_value=line.team)
size_multiplier = 500
plot_data["pred_goal_size"] = plot_data.pred_goal * size_multiplier
for shot_event in shot_events:
conds = np.logical_and(plot_data.forwards_api_id == line.forwards_api_id, plot_data.event == shot_event)
plot_data2 = plot_data.loc[conds]
if plot_data2.empty:
continue
colors = NHL_COLORS[plot_data2.iloc[0].event_team]
facecolor = colors[shot_event]
if shot_event == "MISS" or shot_event == "SHOT":
edgecolor = "#FFFFFF"
elif shot_event == "GOAL":
edgecolor = colors["SHOT"] if facecolor == "#FFFFFF" else "#FFFFFF"
rink.plot_fn(
sns.scatterplot,
data=plot_data2,
x="norm_coords_x",
y="norm_coords_y",
color=facecolor,
edgecolor=edgecolor,
lw=0.75,
s=plot_data2.pred_goal_size,
# sizes=(0, size_multiplier),
size_norm=(0, size_multiplier),
zorder=100,
alpha=0.75,
ax=ax,
)
ax.set_title(f"{line.forwards}", x=0.5, y=1.01, ha="center", fontweight="bold", fontsize=10)
for row, line in plot_lines.iterrows():
row = row + max_lines
ax = axes[row]
ax_zone = "dzone" if row > max_lines - 1 else "ozone"
rink.draw(ax=ax, display_range=ax_zone)
shot_events = ["MISS", "SHOT", "GOAL"]
plot_conds = np.logical_and.reduce(
[
pbp.opp_forwards == line.forwards,
pbp.opp_forwards_api_id == line.forwards_api_id,
pbp.strength_state == strength_state,
pbp.event.isin(shot_events),
]
)
plot_data = pbp.loc[plot_conds].reset_index(drop=True)
plot_data = norm_coords(data=plot_data, norm_column="event_team", norm_value=line.team)
size_multiplier = 500
plot_data["pred_goal_size"] = plot_data.pred_goal * size_multiplier
for shot_event in shot_events:
conds = np.logical_and(plot_data.opp_forwards_api_id == line.forwards_api_id, plot_data.event == shot_event)
plot_data2 = plot_data.loc[conds]
if plot_data2.empty:
continue
colors = NHL_COLORS[plot_data2.iloc[0].opp_team]
facecolor = colors[shot_event]
if shot_event == "MISS" or shot_event == "SHOT":
edgecolor = "#FFFFFF"
elif shot_event == "GOAL":
edgecolor = colors["SHOT"] if facecolor == "#FFFFFF" else "#FFFFFF"
rink.plot_fn(
sns.scatterplot,
data=plot_data2,
x="norm_coords_x",
y="norm_coords_y",
color=facecolor,
edgecolor=edgecolor,
lw=0.75,
s=plot_data2.pred_goal_size,
# sizes=(0, size_multiplier),
size_norm=(0, size_multiplier),
zorder=100,
alpha=0.75,
ax=ax,
)
fig_title = "Nashville forward lines aren't converting 5v5 scoring chances"
fig.suptitle(fig_title, x=0.5, y=1.1, fontweight="bold", fontsize=14)
subtitle = f"NSH top-{max_lines} forward line combinations by 5v5 TOI | 5v5 unblocked shot attempts for & against, sized for xG"
fig.text(s=subtitle, x=0.5, y=1.05, fontsize=12, ha="center")
attribution = "Data & xG model @chickenandstats | Viz @chickenandstats"
fig.text(s=attribution, x=0.95, y=-0.02, fontsize=12, ha="right", style="italic")
savepath = Path(f"./charts/{team.lower()}_forwards_rink_maps.png")
fig.savefig(savepath, transparent=False, bbox_inches="tight")
Contour plots
Drawing the rink
Drawing the rink with the NHLRink class from the hockey-rink library
Filter conditions
Set the team, strength state, minimum TOI, and maximum number of players to plot, then filter the data
team = "NSH"
strength_states = ["5v4", "5v3", "4v3"]
positions = ["L", "C", "R", "L/R", "C/R", "R/L", "R/C"]
toi_min = 5
max_players = 6
group_columns = ["player", "api_id", "team"] # "strength_state"]
conds = np.logical_and.reduce(
[
stats.team == team,
stats.strength_state.isin(strength_states),
stats.position.isin(positions),
stats.toi >= toi_min,
]
)
plot_stats = (
stats.loc[conds]
.groupby(group_columns, as_index=False)
.agg({"ixg": "sum", "g": "sum", "toi": "sum"})
.sort_values(by="toi", ascending=False)
.head(max_players)
.reset_index(drop=True)
)
Top-N forward shooters
Plot the top-N forwards' xG using the hockey-rink library and seaborn's kdeplot
cmap = "rocket_r"
weights = "pred_goal"
fig, axes = plt.subplots(nrows=2, ncols=int(max_players / 2), dpi=650, figsize=(12, 8))
fig.tight_layout(pad=1.5)
axes = axes.reshape(-1)
for row, player in plot_stats.iterrows():
ax = axes[row]
rink.draw(ax=ax, display_range="ozone")
shot_events = ["MISS", "SHOT", "GOAL"]
plot_conds = np.logical_and.reduce(
[
pbp.player_1 == player.player,
pbp.player_1_api_id == player.api_id,
pbp.strength_state.isin(strength_states),
pbp.event.isin(shot_events),
]
)
plot_data = pbp.loc[plot_conds].reset_index(drop=True)
plot_data = norm_coords(data=plot_data, norm_column="player_1_api_id", norm_value=player.api_id)
rink.plot_fn(
sns.kdeplot,
data=plot_data,
x="norm_coords_x",
y="norm_coords_y",
cmap=cmap,
fill=True,
levels=12,
weights=plot_data[weights],
zorder=100,
alpha=0.75,
ax=ax,
legend=True,
)
rink.plot_fn(
sns.kdeplot,
data=plot_data,
x="norm_coords_x",
y="norm_coords_y",
cmap=cmap,
fill=False,
levels=12,
linewidths=2,
weights=plot_data[weights],
zorder=110,
alpha=1,
ax=ax,
)
ax.set_title(f"{player.player}", x=0.5, y=1.01, ha="center", fontweight="bold", fontsize=10)
save_path = Path(f"./charts/{team}_top_{max_players}_pp.png")
fig.savefig(save_path, transparent=False, bbox_inches="tight")
