In [8]:

#import packages
import pandas as pd
import matplotlib as mpl
import matplotlib.pyplot as plt

In [3]:

#read in data from csv this is from a premier league game Southampton vs Bournemouth in 2019
#I scrpaed this data from understat. I would recommend checking out the tutorial on scraping understat on my Youtube channel
#https://www.youtube.com/c/McKayJohns
df = pd.read_csv('xgtutorial.csv')

In [4]:

df

Out[4]:

	Unnamed: 0	minute	team	xG	player	result
0	0	1	Southampton	0.105402	Nathan Redmond	BlockedShot
1	1	2	Southampton	0.136320	Jan Bednarek	BlockedShot
2	2	2	Southampton	0.037217	Oriol Romeu	BlockedShot
3	3	2	Southampton	0.021404	James Ward-Prowse	BlockedShot
4	4	3	Southampton	0.066363	Pierre-Emile Højbjerg	BlockedShot
5	5	9	Southampton	0.363826	Nathan Redmond	SavedShot
6	6	11	Southampton	0.354095	Shane Long	Goal
7	7	17	Southampton	0.036274	Nathan Redmond	MissedShots
8	8	21	Southampton	0.076804	Danny Ings	MissedShots
9	9	23	Southampton	0.380378	Shane Long	SavedShot
10	10	23	Southampton	0.608956	Shane Long	ShotOnPost
11	11	47	Southampton	0.027636	Ryan Bertrand	MissedShots
12	12	49	Southampton	0.026454	Nathan Redmond	MissedShots
13	13	54	Southampton	0.033908	James Ward-Prowse	Goal
14	14	57	Southampton	0.429492	Danny Ings	SavedShot
15	15	58	Southampton	0.021995	Danny Ings	SavedShot
16	16	60	Southampton	0.050620	Matt Targett	BlockedShot
17	17	60	Southampton	0.053497	Nathan Redmond	MissedShots
18	18	66	Southampton	0.086690	Matt Targett	Goal
19	19	71	Southampton	0.011735	Pierre-Emile Højbjerg	MissedShots
20	20	79	Southampton	0.023274	Nathan Redmond	MissedShots
21	21	81	Southampton	0.018172	Pierre-Emile Højbjerg	MissedShots
22	22	19	Bournemouth	0.516069	Dan Gosling	Goal
23	23	31	Bournemouth	0.016307	David Brooks	SavedShot
24	24	31	Bournemouth	0.515468	Callum Wilson	Goal
25	25	51	Bournemouth	0.025648	Jack Simpson	MissedShots
26	26	72	Bournemouth	0.034942	Ryan Fraser	MissedShots
27	27	82	Bournemouth	0.017407	Joshua King	MissedShots
28	28	85	Bournemouth	0.973102	Callum Wilson	Goal
29	29	87	Bournemouth	0.506683	Callum Wilson	SavedShot
30	30	94	Bournemouth	0.024924	Joshua King	MissedShots

In [7]:

#now that we have our dataframe set up, we are going to create some lists to plot the different xG values
#4 lists - home and away xg and minutes
#We start these with zero so our charts will start at 0
a_xG = [0]
h_xG= [0]
a_min = [0]
h_min = [0]


#this finds our team names from the dataframe. This will only work as long as both teams took a shot
hteam = df['team'].iloc[0]
ateam = df['team'].iloc[-1]

for x in range(len(df['xG'])):
    if df['team'][x]==ateam:
        a_xG.append(df['xG'][x])
        a_min.append(df['minute'][x])
    if df['team'][x]==hteam:
        h_xG.append(df['xG'][x])
        h_min.append(df['minute'][x])
        
#this is the function we use to make our xG values be cumulative rather than single shot values
#it goes through the list and adds the numbers together
def nums_cumulative_sum(nums_list):
    return [sum(nums_list[:i+1]) for i in range(len(nums_list))]

a_cumulative = nums_cumulative_sum(a_xG)
h_cumulative = nums_cumulative_sum(h_xG)

#this is used to find the total xG. It just creates a new variable from the last item in the cumulative list
alast = round(a_cumulative[-1],2)
hlast = round(h_cumulative[-1],2)

In [12]:

fig, ax = plt.subplots(figsize = (10,5))
fig.set_facecolor('#3d4849')
ax.patch.set_facecolor('#3d4849')

#set up our base layer
mpl.rcParams['xtick.color'] = 'white'
mpl.rcParams['ytick.color'] = 'white'

ax.grid(ls='dotted',lw=.5,color='lightgrey',axis='y',zorder=1)
spines = ['top','bottom','left','right']
for x in spines:
    if x in spines:
        ax.spines[x].set_visible(False)
        
plt.xticks([0,15,30,45,60,75,90])
plt.xlabel('Minute',fontname='Andale Mono',color='white',fontsize=16)
plt.ylabel('xG',fontname='Andale Mono',color='white',fontsize=16)

#plot the step graphs
ax.step(x=a_min,y=a_cumulative,color='#d3d3d3',label=ateam,linewidth=5,where='post')
ax.step(x=h_min,y=h_cumulative,color='#fd3607',label=ateam,linewidth=5,where='post')

Out[12]:

[<matplotlib.lines.Line2D at 0x10013cb8>]

In [ ]: