NHL Analysis (1963 - 2022)

This is my very first project using pandas and matplotlib altogether. In this entry, I will show some analytics I made with this database. Shoutout to OP for the dataset!

Import the libraries and .csv file

python 

import pandas as pd #graphs 
import numpy as np 
import matplotlib.pyplot as plt

df = pd.read_csv('nhldraft.csv')
df.head()
output 






  
    

      
      id
      year
      overall_pick
      team
      player
      nationality
      position
      age
      to_year
      amateur_team
      ...
      points
      plus_minus
      penalties_minutes
      goalie_games_played
      goalie_wins
      goalie_losses
      goalie_ties_overtime
      save_percentage
      goals_against_average
      point_shares
    

  
  
    

      0
      1
      2022
      1
      Montreal Canadiens
      Juraj Slafkovsky
      SK
      LW
      18.0
      NaN
      TPS (Finland)
      ...
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
    

    

      1
      2
      2022
      2
      New Jersey Devils
      Simon Nemec
      SK
      D
      18.0
      NaN
      HK Nitra (Slovakia)
      ...
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
    

    

      2
      3
      2022
      3
      Arizona Coyotes
      Logan Cooley
      US
      C
      18.0
      NaN
      USA U-18 Development Team (USDP/USHL)
      ...
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
    

    

      3
      4
      2022
      4
      Seattle Kraken
      Shane Wright
      CA
      C
      18.0
      NaN
      Kingston Frontenacs (OHL)
      ...
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
    

    

      4
      5
      2022
      5
      Philadelphia Flyers
      Cutter Gauthier
      SE
      LW
      18.0
      NaN
      USA U-18 Development Team (USDP/USHL)
      ...
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
      NaN
    

  



5 rows × 23 columns

Players drafted by age

python 

ax=df['age'].plot(kind='hist',ylabel='Q',xlabel='age',bins=20,color='pink',title='PLAYERS DRAFT AGE')

#axis name change

ticks = np.arange(df['age'].min(), df['age'].max()+1,2)
ax.set_xticks(ticks)

As expected, most of these guys are 18 years old. You can also see that a few players were younger than 18 — these correspond to the years 1955–1967, before the Junior Hockey regulations changed to set the age range to 18–20 for Canadian and American players, and 18–21 for international players. You can read more about this here.

Teams with less and most players drafted.

python 

# less
df['team'].value_counts().nsmallest(15).plot(kind='barh',figsize=(10,10),title='TEAMS WITH LESS PLAYERS DRAFTED',color='pink')

# most
df['team'].value_counts().nlargest(6).plot(kind='barh',figsize=(10,10),color='pink',title='TEAMS WITH MORE PLAYERS DRAFTED')

Still not a very surprising result — the teams that drafted the fewest players are the newest franchises (such as the Kraken, the Vegas Golden Knights, and the Minnesota Wild), as well as teams whose franchises are no longer active.

Here, I purposely created a Top 6 to see whether all the "Original 6" teams would appear — but surprisingly, they didn’t. The Leafs barely made the list, and the Flyers ended up taking the Bruins’ spot instead.

The Flyers ranking so high reflects their drafting approach during the 80s and 90s, where they accumulated more mid-round picks than most Original Six teams. Of course, drafting many players doesn’t necessarily mean drafting successful players.

The gap between the Leafs and the top teams is also smaller than expected, showing how consistent their draft activity has been across decades despite mixed on-ice results.