National Hockey League Statistics¶

We can look at NHL statistics by team or by player, using data from hockey-reference.com or ESPN NHL Statistics.

Statistics by Team¶

team = 'EDM'
year = '2019'

# download the data
import pandas as pd
team_stats_url = 'https://www.hockey-reference.com/teams/'+team+'/'+year+'_games.html'
team_stats = pd.read_html(team_stats_url)[0]
# clean up the data
team_stats = team_stats[team_stats['Date']!='Date'].set_index('GP').drop(columns=['W','L','OL','Streak','Notes'])
team_stats.columns = ['Date', 'Away', 'Opponent', 'Goals For', 'Goals Against', 'Win', 'Overtime', 'Attendance', 'Duration']
team_stats = team_stats.fillna(0).replace('@', 1).replace('OT', 1).replace('W',1).replace('SO',1).replace('L',0)
# convert text string columns to number columns
team_stats['Goals For'] = pd.to_numeric(team_stats['Goals For'])
team_stats['Goals Against'] = pd.to_numeric(team_stats['Goals Against'])
team_stats['Attendance'] = pd.to_numeric(team_stats['Attendance'])
# convert duration in h:mm to duration in minutes
duration_values = team_stats['Duration'].str.split(':', expand=True).astype(int)
team_stats['Duration'] = duration_values[0]*60 + duration_values[1]
# display the data
team_stats

	Date	Away	Opponent	Goals For	Goals Against	Win	Overtime	Attendance	Duration
GP
1	2018-10-06	1	New Jersey Devils	2	5	0	0	12044	151
2	2018-10-11	1	Boston Bruins	1	4	0	0	17565	148
3	2018-10-13	1	New York Rangers	2	1	1	0	17085	144
4	2018-10-16	1	Winnipeg Jets	5	4	1	1	15321	149
5	2018-10-18	0	Boston Bruins	3	2	1	1	18347	146
...	...	...	...	...	...	...	...	...	...
78	2019-03-30	0	Anaheim Ducks	1	5	0	0	18347	141
79	2019-04-01	1	Vegas Golden Knights	1	3	0	0	18367	144
80	2019-04-02	1	Colorado Avalanche	2	6	0	0	17021	142
81	2019-04-04	0	San Jose Sharks	2	3	0	0	18347	147
82	2019-04-06	1	Calgary Flames	3	1	1	0	19289	145

82 rows × 9 columns

Statistics by Player¶

This data set contains the following columns for each player in the NHL:

GP: Games Played
G: Goals
A: Assists
PTS: Points
+/-: Plus/Minus Rating
PIM: Penalty Minutes
PTS/G: Points Per Game
SOG: Shots on Goal
PCT: Shooting Percentage
GWG: Game-Winning Goals
G.1: Power-Play Goals
A.1: Power-Play Assists
G.2: Short-Handed Goals
A.2: Short-Handed Assists

This will take a while to run, since it needs to get data from multiple pages.

# download the data
points_url = 'http://www.espn.com/nhl/statistics/player/_/stat/points'
import pandas as pd
for i in range(20):
    try:
        p = pd.read_html(points_url+'/count/'+str(1+40*i), header=1)[0]
        p = p[p['PLAYER']!='PLAYER'].dropna(subset=['PLAYER']).fillna(method='ffill')
        if i == 0:
            points = p
        else:
            points = points.append(p).reset_index().drop(columns='index')
    # if the site has run out of data
    except:
        pass
# convert text string columns to number columns
for column in points.columns:
    if column != 'PLAYER' and column != 'TEAM':
        points[column] = pd.to_numeric(points[column])
# split the player name and position into two columns
points['POSITION'] = points['PLAYER'].str.split(',', expand=True)[1]
points['PLAYER'] = points['PLAYER'].str.split(',', expand=True)[0]
# display the data
points

	RK	PLAYER	TEAM	GP	G	A	PTS	+/-	PIM	PTS/G	SOG	PCT	GWG	G.1	A.1	G.2	A.2	POSITION
0	1	Nikita Kucherov	TB	20	6	20	26	13	20	1.30	72	8.3	1	0	7	0	0	RW
1	2	Nathan MacKinnon	COL	15	9	16	25	13	12	1.67	65	13.8	0	3	6	0	0	C
2	2	Brayden Point	TB	18	9	16	25	10	8	1.39	54	16.7	2	1	3	0	0	C
3	4	Miro Heiskanen	DAL	22	5	18	23	5	2	1.05	47	10.6	0	2	6	0	0	D
4	5	Mikko Rantanen	COL	15	7	14	21	11	6	1.40	55	12.7	0	2	6	0	0	RW
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
370	361	Marcus Foligno	MIN	4	0	1	1	-2	5	0.25	4	0.0	0	0	0	0	0	LW
371	361	Carl Hagelin	WSH	8	0	1	1	-4	2	0.13	4	0.0	0	0	0	0	0	LW
372	361	Jake Evans	MTL	6	0	1	1	-1	0	0.17	3	0.0	0	0	0	0	0	C
373	361	Ilya Kovalchuk	WSH	8	0	1	1	0	2	0.13	5	0.0	0	0	0	0	0	LW
374	361	Morgan Geekie	CAR	8	0	1	1	-1	0	0.13	4	0.0	0	0	0	0	0	C

375 rows × 18 columns

National Basketball Association Olympics Data