kristiker/match_history_analyse.py

## match_history_analyse.py
"""
Analyse your CS:GO Match History Data

Original by DrMarioCSGO
https://pastebin.com/v5EVZMTq
https://www.reddit.com/r/GlobalOffensive/comments/qvdf4v/

Requirements:
pip install pandas
pip install matplotlib
pip install html5lib
pip install bs4

Usage:
    You'll need to load your full match history (automatically possible
    with Ban Checker for Steam), Ctrl+S(ave) it, then put the python file
    in the same directory, put in your nickname at the top (me variable)
    and just run the script (might take a minute).
"""

import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
import matplotlib as mpl
import numpy as np
from collections import Counter, OrderedDict
import pprint
from datetime import datetime
import re
import pickle
import unicodedata
import warnings


###########
## Enter your in-game nick here:
me = "ϟ kристи'"


file_path = "Steam Community __ Counter-Strike_ Global Offensive __ Personal Game Data.html"

# because of shit special character nicknames some warnings are thrown when plotting
# you can disable them here
warnings.filterwarnings("ignore", message="Glyph 119967 missing from current font")


def generator():
    # Generating list of dataframes
    with open(file_path, "r",encoding="utf8") as f:
        data = pd.read_html(f.read(),flavor="bs4",attrs = {"class":["csgo_scoreboard_inner_left","csgo_scoreboard_inner_right"]})

    # Remove surrenders, short matches and scrimmage mirage
    selected_for_deletion = []
    for idx, dataframe in enumerate(data):
        if (idx % 2) == 0:
            if data[idx].at[0,0] == "Competitive Mirage Scrimmage":
                selected_for_deletion.append(idx)
                selected_for_deletion.append(idx+1)
        else:
            if data[idx].at[5,"Player Name"].split()[0] != str(16) and \
               data[idx].at[5,"Player Name"].split()[0] != str(15) and \
               data[idx].at[5,"Player Name"].split()[2] != str(16) and \
               data[idx].at[5,"Player Name"].split()[2] != str(15):
                selected_for_deletion.append(idx)
                selected_for_deletion.append(idx-1)
            if data[idx].at[5,"Player Name"].split()[0] == str(15) and \
               data[idx].at[5,"Player Name"].split()[2] != str(15):
                selected_for_deletion.append(idx)
                selected_for_deletion.append(idx-1)
            if data[idx].at[5,"Player Name"].split()[2] == str(15) and \
               data[idx].at[5,"Player Name"].split()[0] != str(15):
                selected_for_deletion.append(idx)
                selected_for_deletion.append(idx-1)

    selected_for_deletion.sort(reverse=True)
    for each in selected_for_deletion:
        del data[each]


    # Generating match dictionaries
    matches = [{} for i in range(len(data)//2)]


    for i in range(len(data)):

        if (i % 2) == 0:
            world = data[i].at[0,0]         # map seems to be a reserverd python word
            world = world.split(' ', 1)[1]  # remove Competitive word
            matches[i//2]["world"] = world

            time = data[i].at[1,0]
            matches[i//2]["time"] = datetime.strptime(time,"%Y-%m-%d %H:%M:%S %Z")

            waittime = data[i].at[3,0]
            waittime = waittime.split()[2]
            int(waittime[:2])*60+int(waittime[3:5])
            matches[i//2]["waittime"] = int(waittime[:2])*60+int(waittime[3:5])

            duration = data[i].at[4,0]
            duration = duration.split()[2]
            matches[i//2]["duration"] = int(duration[:2])*60+int(duration[3:5])

        else:
            ### Playerlist
            players = []
            for j in range(0,5):
                name = data[i].at[j,"Player Name"]
                players.append(name)
            for j in range(6,11):
                name = data[i].at[j,"Player Name"]
                players.append(name)
            matches[i//2]["players"] = players

            ### Outcome & Result
            result = data[i].at[5,"Player Name"]
            result_team1 = int(result.split()[0])
            result_team2 = int(result.split()[2])

            if result_team1 == result_team2:
                matches[i//2]["outcome"] = "draw"
            else:
                if players.index(me) <= 4:
                    if result_team1 < result_team2:
                        matches[i//2]["outcome"] = "loss"
                    else:
                        matches[i//2]["outcome"] = "won"
                else:
                    if result_team1 < result_team2:
                        matches[i//2]["outcome"] = "won"
                    else:
                        matches[i//2]["outcome"] = "loss"

            if result_team1 < result_team2:          # 16:2 = 2:16 etc. for later analysis
                result = str(result_team2) + " : " + str(result_team1)
            matches[i//2]["result"] = result

            ### Personal Stats
            if players.index(me) > 4:
                myindex = players.index(me)+1
            else:
                myindex = players.index(me)
            matches[i//2]["ping"] = int(data[i].at[myindex,"Ping"])
            matches[i//2]["kills"] = int(data[i].at[myindex,"K"])
            matches[i//2]["assists"] = int(data[i].at[myindex,"A"])
            matches[i//2]["deaths"] = int(data[i].at[myindex,"D"])
            #matches[i//2]["mvps"] = data[i].at[myindex,"★"].strip("★")    # to do: NaN (is float type)
            #matches[i//2]["HSP"] = data[i].at[myindex,"HSP"].strip("%")    # dito
            matches[i//2]["score"] = int(data[i].at[myindex,"Score"])

    return matches

## The following two lines need only be run for the first time
## afterwards you can comment them and uncomment the following line for faster compilation

matches = generator()
pickle.dump( matches, open( "save.p", "wb" ) )

#matches = pickle.load( open( "save.p", "rb" ) )


#pprint.pprint(matches)

def debugging():
    pd.set_option("display.max_columns", 101)
    pd.set_option('display.expand_frame_repr', False)
    for i in range(len(data)):
        print("####### "+str(i)+" #######")
        print(data[i])
        print("\n\n")
#debugging()


def analysing_matchtimes():
    x = []
    times = []

    for item in matches:
        x.append(mpl.dates.date2num(item["time"]))  # list of all datetimes

    for i in range(len(x)):
        time = x[i] % 1 + int(x[0])     # set all times to a day
        times.append(time)


    fig, ax = plt.subplots()

    for i in range(len(times)):
        if matches[i]["outcome"] == "loss":
            ax.plot_date(x[i], times[i], 'ro',markersize=2)
        elif matches[i]["outcome"] == "won":
            ax.plot_date(x[i], times[i], 'go',markersize=2)
        else:
            ax.plot_date(x[i], times[i], 'mo',markersize=2)


    ax.yaxis_date()
    ax.yaxis.set_major_formatter(mdates.DateFormatter('%H:%M'))
    ax.xaxis.set_major_formatter(mdates.DateFormatter('%b %Y'))

    fig.autofmt_xdate()
    plt.show()

analysing_matchtimes()


def analysing_matches(StatOfInterest):
    x = []
    y = []
    kills = []

    for item in matches:
        x.append(mpl.dates.date2num(item["time"]))  # list of all datetimes

    for i in range(len(x)):
        if StatOfInterest == "kd":
            kill = matches[i]["kills"]/max(matches[i]["deaths"], 1)      #k/d
        else:
            kill = matches[i][StatOfInterest]
        kills.append(kill)


    fig, ax = plt.subplots()

    for i in range(len(kills)):
        if matches[i]["outcome"] == "loss":
            ax.plot(x[i], kills[i], 'ro',markersize=2, label=StatOfInterest + " of Loss")
        elif matches[i]["outcome"] == "won":
            ax.plot(x[i], kills[i], 'go',markersize=2, label=StatOfInterest + " of Win")
        else:
            ax.plot(x[i], kills[i], 'mo',markersize=2, label=StatOfInterest + " of Draw")

    N=50
    y_padded = np.pad(kills, (N//2, N-1-N//2), mode='edge')
    y_smooth = np.convolve(y_padded, np.ones((N,))/N, mode='valid')
    plt.plot(x,y_smooth,"b-",label="Moving average")

    plt.axvline(datetime(2020, 9, 11))
    ax.xaxis.set_major_formatter(mdates.DateFormatter('%b %Y'))
    plt.title(StatOfInterest + " of matches throughout the years")
    plt.xlabel("Date")
    plt.ylabel(StatOfInterest)

    handles, labels = plt.gca().get_legend_handles_labels() # no legend for every single match
    by_label = dict(zip(labels, handles))
    plt.legend(by_label.values(), by_label.keys())

    fig.autofmt_xdate()
    plt.show()

analysing_matches("waittime")
analysing_matches("duration")
analysing_matches("ping")
analysing_matches("kills")
#analysing_matches("assists") boring
analysing_matches("deaths")
analysing_matches("kd")


def analysing_matches_by_daytime(StatOfInterest):
    x = []
    y = []
    kills = []
    times=[]

    for item in matches:
        x.append(mpl.dates.date2num(item["time"]))  # list of all datetimes

    for i in range(len(x)):
        time = x[i] % 1 + int(x[0])     # set all times to a day
        times.append(time)

    for i in range(len(x)):
        if StatOfInterest == "kd":
            kill = matches[i]["kills"]/max(matches[i]["deaths"], 1)      #k/d
        else:
            kill = matches[i][StatOfInterest]
        kills.append(kill)


    fig, ax = plt.subplots()

    for i in range(len(kills)):
        if matches[i]["outcome"] == "loss":
            ax.plot(times[i], kills[i], 'rx',markersize=2, label=StatOfInterest + " of Loss")
        elif matches[i]["outcome"] == "won":
            ax.plot(times[i], kills[i], 'gx',markersize=2, label=StatOfInterest + " of Win")
        else:
            ax.plot(times[i], kills[i], 'yx',markersize=2, label=StatOfInterest + " of Draw")

    zipped = zip(times,kills)
    zipped = sorted(zipped)
    times,kills = list(zip(*zipped))
    N=50
    y_padded = np.pad(kills, (N//2, N-1-N//2), mode='edge')
    y_smooth = np.convolve(y_padded, np.ones((N,))/N, mode='valid')
    plt.plot(times,y_smooth,"b-",label="Moving average")

    #plt.axvline(datetime(2020, 9, 11))
    ax.xaxis.set_major_formatter(mdates.DateFormatter('%H-%M'))
    plt.title(StatOfInterest + " of matches throughout time of day")
    plt.xlabel("Time of day")
    plt.ylabel(StatOfInterest)

    handles, labels = plt.gca().get_legend_handles_labels() # no legend for every single match
    by_label = dict(zip(labels, handles))
    plt.legend(by_label.values(), by_label.keys())

    fig.autofmt_xdate()
    plt.show()

analysing_matches_by_daytime("waittime")
analysing_matches_by_daytime("duration")
analysing_matches_by_daytime("ping")
analysing_matches_by_daytime("kills")
#analysing_matches_by_daytime("assists") boring
analysing_matches_by_daytime("deaths")
analysing_matches_by_daytime("kd")


def analysing_players(top_x_players,readout="no"):
    c = Counter()
    for item in matches:
        for player in item["players"]:
            c[player] += 1


    if readout == "yes":
        pprint.pprint(c.most_common(top_x_players)[1:])
        y = [count for tag, count in c.most_common(top_x_players)[1:]]
        x = [tag for tag, count in c.most_common(top_x_players)[1:]]

        fig, ax = plt.subplots()
        width = 0.85
        ind = np.arange(len(y))
        ax.barh(ind, y, width)
        ax.set_yticks(ind)
        ax.set_yticklabels(x)
        plt.tight_layout()
        for i, v in enumerate(y):
            plt.text(v, i, " "+str(v), color='blue', va='center')
        plt.subplots_adjust(top=0.9, right=0.9)
        plt.title('Number of times a teammate joined your ' +\
                  str(c.most_common(top_x_players)[0][1]) + " matches")
        plt.show()

    return c.most_common(top_x_players)
analysing_players(25,"yes")


def analysing_players_cumsum():

    top_x_players = 25         # how many players to look at

    all_datetimes = []
    for item in matches:
        all_datetimes.append(item["time"])
    oldest_datetime=max(all_datetimes)
    print(oldest_datetime)

    fig, ax = plt.subplots()
    list_of_most_common_players = analysing_players(top_x_players)

    for player in list_of_most_common_players[1:]:      # skip "me"
        r = []
        for i in range(len(matches)-1,-1,-1):           # for some reason need to go through reversed
            if player[0] in matches[i]["players"]:
                r.append((matches[i]["time"],1))

        r.append((oldest_datetime, 0))       # draw line to end of diagramm

        x , v = [d[0] for d in r], [d[1] for d in r]
        v = np.array(v).cumsum()                        # magic from stackoverflow
        ax.plot(x, v, '-',label=player[0])  # replace shit special character names
        ax.legend()


    plt.title('Number of times a mate joined you for a match over time')
    ax.xaxis.set_major_formatter(mdates.DateFormatter('%b %Y'))

    fig.autofmt_xdate()
    plt.show()
analysing_players_cumsum()


def analysing_results():
    c = Counter()

    for item in matches:
        c[item["result"]] += 1

##    for item in matches:
##        if item["world"] == "Nuke":       # Check for specific map
##            c[item["result"]] += 1        # (comment out above for loop)

    c = c.most_common()

    SORT_ORDER = {'16 : 0': 0, '16 : 1': 1, '16 : 2': 2, '16 : 3': 3,\
                  '16 : 4': 4, '16 : 5': 5, '16 : 6': 6, '16 : 7': 7,\
                  '16 : 8': 8, '16 : 9': 9, '16 : 10': 10, '16 : 11': 11,\
                  '16 : 12': 12, '16 : 13': 13, '16 : 14': 14, '15 : 15': 15}

    c.sort(key=lambda val: SORT_ORDER[val[0]])
    pprint.pprint(c)

    y = [count for tag, count in c]
    x = [tag for tag, count in c]

    fig, ax = plt.subplots()
    width = 0.85
    ind = np.arange(len(y))
    ax.barh(ind, y, width)
    ax.set_yticks(ind)
    ax.set_yticklabels(x)
    plt.tight_layout()
    for i, v in enumerate(y):
        plt.text(v, i, " "+str(v), color='blue', va='center')
    plt.subplots_adjust(top=0.9, right=0.9)
    plt.title('Number of times a specific result has occured')
    plt.show()
analysing_results()


def analysing_maps():
    c = Counter()
    for item in matches:
        c[item["world"]] += 1
    pprint.pprint(c.most_common())

    y = [count for tag, count in c.most_common()]
    x = [tag for tag, count in c.most_common()]

    fig, ax = plt.subplots()
    width = 0.85
    ind = np.arange(len(y))
    ax.barh(ind, y, width)
    ax.set_yticks(ind)
    ax.set_yticklabels(x)
    plt.tight_layout()
    for i, v in enumerate(y):
        plt.text(v, i, " "+str(v), color='blue', va='center')
    plt.subplots_adjust(top=0.9, right=0.9)
    plt.title('Number of times a map has been played')
    plt.show()
analysing_maps()
	"""
	Analyse your CS:GO Match History Data

	Original by DrMarioCSGO
	https://pastebin.com/v5EVZMTq
	https://www.reddit.com/r/GlobalOffensive/comments/qvdf4v/

	Requirements:
	pip install pandas
	pip install matplotlib
	pip install html5lib
	pip install bs4

	Usage:
	You'll need to load your full match history (automatically possible
	with Ban Checker for Steam), Ctrl+S(ave) it, then put the python file
	in the same directory, put in your nickname at the top (me variable)
	and just run the script (might take a minute).
	"""

	import pandas as pd
	import matplotlib.pyplot as plt
	import matplotlib.dates as mdates
	import matplotlib as mpl
	import numpy as np
	from collections import Counter, OrderedDict
	import pprint
	from datetime import datetime
	import re
	import pickle
	import unicodedata
	import warnings


	###########
	## Enter your in-game nick here:
	me = "ϟ kристи'"


	file_path = "Steam Community __ Counter-Strike_ Global Offensive __ Personal Game Data.html"

	# because of shit special character nicknames some warnings are thrown when plotting
	# you can disable them here
	warnings.filterwarnings("ignore", message="Glyph 119967 missing from current font")


	def generator():
	# Generating list of dataframes
	with open(file_path, "r",encoding="utf8") as f:
	data = pd.read_html(f.read(),flavor="bs4",attrs = {"class":["csgo_scoreboard_inner_left","csgo_scoreboard_inner_right"]})

	# Remove surrenders, short matches and scrimmage mirage
	selected_for_deletion = []
	for idx, dataframe in enumerate(data):
	if (idx % 2) == 0:
	if data[idx].at[0,0] == "Competitive Mirage Scrimmage":
	selected_for_deletion.append(idx)
	selected_for_deletion.append(idx+1)
	else:
	if data[idx].at[5,"Player Name"].split()[0] != str(16) and \
	data[idx].at[5,"Player Name"].split()[0] != str(15) and \
	data[idx].at[5,"Player Name"].split()[2] != str(16) and \
	data[idx].at[5,"Player Name"].split()[2] != str(15):
	selected_for_deletion.append(idx)
	selected_for_deletion.append(idx-1)
	if data[idx].at[5,"Player Name"].split()[0] == str(15) and \
	data[idx].at[5,"Player Name"].split()[2] != str(15):
	selected_for_deletion.append(idx)
	selected_for_deletion.append(idx-1)
	if data[idx].at[5,"Player Name"].split()[2] == str(15) and \
	data[idx].at[5,"Player Name"].split()[0] != str(15):
	selected_for_deletion.append(idx)
	selected_for_deletion.append(idx-1)

	selected_for_deletion.sort(reverse=True)
	for each in selected_for_deletion:
	del data[each]


	# Generating match dictionaries
	matches = [{} for i in range(len(data)//2)]


	for i in range(len(data)):

	if (i % 2) == 0:
	world = data[i].at[0,0] # map seems to be a reserverd python word
	world = world.split(' ', 1)[1] # remove Competitive word
	matches[i//2]["world"] = world

	time = data[i].at[1,0]
	matches[i//2]["time"] = datetime.strptime(time,"%Y-%m-%d %H:%M:%S %Z")

	waittime = data[i].at[3,0]
	waittime = waittime.split()[2]
	int(waittime[:2])*60+int(waittime[3:5])
	matches[i//2]["waittime"] = int(waittime[:2])*60+int(waittime[3:5])

	duration = data[i].at[4,0]
	duration = duration.split()[2]
	matches[i//2]["duration"] = int(duration[:2])*60+int(duration[3:5])

	else:
	### Playerlist
	players = []
	for j in range(0,5):
	name = data[i].at[j,"Player Name"]
	players.append(name)
	for j in range(6,11):
	name = data[i].at[j,"Player Name"]
	players.append(name)
	matches[i//2]["players"] = players

	### Outcome & Result
	result = data[i].at[5,"Player Name"]
	result_team1 = int(result.split()[0])
	result_team2 = int(result.split()[2])

	if result_team1 == result_team2:
	matches[i//2]["outcome"] = "draw"
	else:
	if players.index(me) <= 4:
	if result_team1 < result_team2:
	matches[i//2]["outcome"] = "loss"
	else:
	matches[i//2]["outcome"] = "won"
	else:
	if result_team1 < result_team2:
	matches[i//2]["outcome"] = "won"
	else:
	matches[i//2]["outcome"] = "loss"

	if result_team1 < result_team2: # 16:2 = 2:16 etc. for later analysis
	result = str(result_team2) + " : " + str(result_team1)
	matches[i//2]["result"] = result

	### Personal Stats
	if players.index(me) > 4:
	myindex = players.index(me)+1
	else:
	myindex = players.index(me)
	matches[i//2]["ping"] = int(data[i].at[myindex,"Ping"])
	matches[i//2]["kills"] = int(data[i].at[myindex,"K"])
	matches[i//2]["assists"] = int(data[i].at[myindex,"A"])
	matches[i//2]["deaths"] = int(data[i].at[myindex,"D"])
	#matches[i//2]["mvps"] = data[i].at[myindex,"★"].strip("★") # to do: NaN (is float type)
	#matches[i//2]["HSP"] = data[i].at[myindex,"HSP"].strip("%") # dito
	matches[i//2]["score"] = int(data[i].at[myindex,"Score"])

	return matches

	## The following two lines need only be run for the first time
	## afterwards you can comment them and uncomment the following line for faster compilation

	matches = generator()
	pickle.dump( matches, open( "save.p", "wb" ) )

	#matches = pickle.load( open( "save.p", "rb" ) )



	#pprint.pprint(matches)

	def debugging():
	pd.set_option("display.max_columns", 101)
	pd.set_option('display.expand_frame_repr', False)
	for i in range(len(data)):
	print("####### "+str(i)+" #######")
	print(data[i])
	print("\n\n")
	#debugging()


	def analysing_matchtimes():
	x = []
	times = []

	for item in matches:
	x.append(mpl.dates.date2num(item["time"])) # list of all datetimes

	for i in range(len(x)):
	time = x[i] % 1 + int(x[0]) # set all times to a day
	times.append(time)


	fig, ax = plt.subplots()

	for i in range(len(times)):
	if matches[i]["outcome"] == "loss":
	ax.plot_date(x[i], times[i], 'ro',markersize=2)
	elif matches[i]["outcome"] == "won":
	ax.plot_date(x[i], times[i], 'go',markersize=2)
	else:
	ax.plot_date(x[i], times[i], 'mo',markersize=2)


	ax.yaxis_date()
	ax.yaxis.set_major_formatter(mdates.DateFormatter('%H:%M'))
	ax.xaxis.set_major_formatter(mdates.DateFormatter('%b %Y'))

	fig.autofmt_xdate()
	plt.show()

	analysing_matchtimes()


	def analysing_matches(StatOfInterest):
	x = []
	y = []
	kills = []

	for item in matches:
	x.append(mpl.dates.date2num(item["time"])) # list of all datetimes

	for i in range(len(x)):
	if StatOfInterest == "kd":
	kill = matches[i]["kills"]/max(matches[i]["deaths"], 1) #k/d
	else:
	kill = matches[i][StatOfInterest]
	kills.append(kill)


	fig, ax = plt.subplots()

	for i in range(len(kills)):
	if matches[i]["outcome"] == "loss":
	ax.plot(x[i], kills[i], 'ro',markersize=2, label=StatOfInterest + " of Loss")
	elif matches[i]["outcome"] == "won":
	ax.plot(x[i], kills[i], 'go',markersize=2, label=StatOfInterest + " of Win")
	else:
	ax.plot(x[i], kills[i], 'mo',markersize=2, label=StatOfInterest + " of Draw")

	N=50
	y_padded = np.pad(kills, (N//2, N-1-N//2), mode='edge')
	y_smooth = np.convolve(y_padded, np.ones((N,))/N, mode='valid')
	plt.plot(x,y_smooth,"b-",label="Moving average")

	plt.axvline(datetime(2020, 9, 11))
	ax.xaxis.set_major_formatter(mdates.DateFormatter('%b %Y'))
	plt.title(StatOfInterest + " of matches throughout the years")
	plt.xlabel("Date")
	plt.ylabel(StatOfInterest)

	handles, labels = plt.gca().get_legend_handles_labels() # no legend for every single match
	by_label = dict(zip(labels, handles))
	plt.legend(by_label.values(), by_label.keys())

	fig.autofmt_xdate()
	plt.show()

	analysing_matches("waittime")
	analysing_matches("duration")
	analysing_matches("ping")
	analysing_matches("kills")
	#analysing_matches("assists") boring
	analysing_matches("deaths")
	analysing_matches("kd")


	def analysing_matches_by_daytime(StatOfInterest):
	x = []
	y = []
	kills = []
	times=[]

	for item in matches:
	x.append(mpl.dates.date2num(item["time"])) # list of all datetimes

	for i in range(len(x)):
	time = x[i] % 1 + int(x[0]) # set all times to a day
	times.append(time)

	for i in range(len(x)):
	if StatOfInterest == "kd":
	kill = matches[i]["kills"]/max(matches[i]["deaths"], 1) #k/d
	else:
	kill = matches[i][StatOfInterest]
	kills.append(kill)


	fig, ax = plt.subplots()

	for i in range(len(kills)):
	if matches[i]["outcome"] == "loss":
	ax.plot(times[i], kills[i], 'rx',markersize=2, label=StatOfInterest + " of Loss")
	elif matches[i]["outcome"] == "won":
	ax.plot(times[i], kills[i], 'gx',markersize=2, label=StatOfInterest + " of Win")
	else:
	ax.plot(times[i], kills[i], 'yx',markersize=2, label=StatOfInterest + " of Draw")

	zipped = zip(times,kills)
	zipped = sorted(zipped)
	times,kills = list(zip(*zipped))
	N=50
	y_padded = np.pad(kills, (N//2, N-1-N//2), mode='edge')
	y_smooth = np.convolve(y_padded, np.ones((N,))/N, mode='valid')
	plt.plot(times,y_smooth,"b-",label="Moving average")

	#plt.axvline(datetime(2020, 9, 11))
	ax.xaxis.set_major_formatter(mdates.DateFormatter('%H-%M'))
	plt.title(StatOfInterest + " of matches throughout time of day")
	plt.xlabel("Time of day")
	plt.ylabel(StatOfInterest)

	handles, labels = plt.gca().get_legend_handles_labels() # no legend for every single match
	by_label = dict(zip(labels, handles))
	plt.legend(by_label.values(), by_label.keys())

	fig.autofmt_xdate()
	plt.show()

	analysing_matches_by_daytime("waittime")
	analysing_matches_by_daytime("duration")
	analysing_matches_by_daytime("ping")
	analysing_matches_by_daytime("kills")
	#analysing_matches_by_daytime("assists") boring
	analysing_matches_by_daytime("deaths")
	analysing_matches_by_daytime("kd")





	def analysing_players(top_x_players,readout="no"):
	c = Counter()
	for item in matches:
	for player in item["players"]:
	c[player] += 1


	if readout == "yes":
	pprint.pprint(c.most_common(top_x_players)[1:])
	y = [count for tag, count in c.most_common(top_x_players)[1:]]
	x = [tag for tag, count in c.most_common(top_x_players)[1:]]

	fig, ax = plt.subplots()
	width = 0.85
	ind = np.arange(len(y))
	ax.barh(ind, y, width)
	ax.set_yticks(ind)
	ax.set_yticklabels(x)
	plt.tight_layout()
	for i, v in enumerate(y):
	plt.text(v, i, " "+str(v), color='blue', va='center')
	plt.subplots_adjust(top=0.9, right=0.9)
	plt.title('Number of times a teammate joined your ' +\
	str(c.most_common(top_x_players)[0][1]) + " matches")
	plt.show()

	return c.most_common(top_x_players)
	analysing_players(25,"yes")


	def analysing_players_cumsum():

	top_x_players = 25 # how many players to look at

	all_datetimes = []
	for item in matches:
	all_datetimes.append(item["time"])
	oldest_datetime=max(all_datetimes)
	print(oldest_datetime)

	fig, ax = plt.subplots()
	list_of_most_common_players = analysing_players(top_x_players)

	for player in list_of_most_common_players[1:]: # skip "me"
	r = []
	for i in range(len(matches)-1,-1,-1): # for some reason need to go through reversed
	if player[0] in matches[i]["players"]:
	r.append((matches[i]["time"],1))

	r.append((oldest_datetime, 0)) # draw line to end of diagramm

	x , v = [d[0] for d in r], [d[1] for d in r]
	v = np.array(v).cumsum() # magic from stackoverflow
	ax.plot(x, v, '-',label=player[0]) # replace shit special character names
	ax.legend()


	plt.title('Number of times a mate joined you for a match over time')
	ax.xaxis.set_major_formatter(mdates.DateFormatter('%b %Y'))

	fig.autofmt_xdate()
	plt.show()
	analysing_players_cumsum()


	def analysing_results():
	c = Counter()

	for item in matches:
	c[item["result"]] += 1

	## for item in matches:
	## if item["world"] == "Nuke": # Check for specific map
	## c[item["result"]] += 1 # (comment out above for loop)

	c = c.most_common()

	SORT_ORDER = {'16 : 0': 0, '16 : 1': 1, '16 : 2': 2, '16 : 3': 3,\
	'16 : 4': 4, '16 : 5': 5, '16 : 6': 6, '16 : 7': 7,\
	'16 : 8': 8, '16 : 9': 9, '16 : 10': 10, '16 : 11': 11,\
	'16 : 12': 12, '16 : 13': 13, '16 : 14': 14, '15 : 15': 15}

	c.sort(key=lambda val: SORT_ORDER[val[0]])
	pprint.pprint(c)

	y = [count for tag, count in c]
	x = [tag for tag, count in c]

	fig, ax = plt.subplots()
	width = 0.85
	ind = np.arange(len(y))
	ax.barh(ind, y, width)
	ax.set_yticks(ind)
	ax.set_yticklabels(x)
	plt.tight_layout()
	for i, v in enumerate(y):
	plt.text(v, i, " "+str(v), color='blue', va='center')
	plt.subplots_adjust(top=0.9, right=0.9)
	plt.title('Number of times a specific result has occured')
	plt.show()
	analysing_results()


	def analysing_maps():
	c = Counter()
	for item in matches:
	c[item["world"]] += 1
	pprint.pprint(c.most_common())

	y = [count for tag, count in c.most_common()]
	x = [tag for tag, count in c.most_common()]

	fig, ax = plt.subplots()
	width = 0.85
	ind = np.arange(len(y))
	ax.barh(ind, y, width)
	ax.set_yticks(ind)
	ax.set_yticklabels(x)
	plt.tight_layout()
	for i, v in enumerate(y):
	plt.text(v, i, " "+str(v), color='blue', va='center')
	plt.subplots_adjust(top=0.9, right=0.9)
	plt.title('Number of times a map has been played')
	plt.show()
	analysing_maps()