NBA MVP Comparisons - Part 3

sports-analytics

python

Author

Kivan Polimis

Published

March 6, 2019

NBA MVP Comparisons

Part 3

6 ^th March 2019

In the last two blog posts we:
1. Gathered the relevant data for all NBA MVPs 2. Evaluated several machine learning models to select the MVP and chose the Latent Discriminant Analysis model as the best predictor.

The goal of this blog post is to gather the data necessary to predict the 2018-2019 MVP

We will look select from the MVP finalists in 2017-2018 * MVP will likely be a finalist from the previous year * Is this a valid assumption?

Update

On May 17^th, the NBA announced the 2018-2019 MVP finalists: * Giannis Antetokounmpo, Milwaukee Bucks * Paul George, Oklahoma City Thunder * James Harden, Houston Rockets * source: https://www.nba.com/article/2019/05/17/finalists-2019-nba-awards

Outline

Import modules
Examine html structure of webpage
Use a function with Beautiful Soup to parse webpages into .csv
Analyze .csv of webpage as a Pandas DataFrame
Process the data

import relevant modules
standard library modules:
- os
- urllib
- webbrowser
- datetime
open source modules:
- pandas
- Beautiful Soup
examine webpage

True

get the html
create the BeautifulSoup object
Extract the necessary values for the column headers from the table and store them as a list

raw column names in finalist table: ['Per Game', 'Shooting', 'Advanced', 'Rank', 'Player', 'Age', 'Tm', 'First', 'Pts Won', 'Pts Max', 'Share', 'G', 'MP', 'PTS', 'TRB', 'AST', 'STL', 'BLK', 'FG%', '3P%', 'FT%', 'WS', 'WS/48', '1', '2', '3', '4']
formatted column names in finalist table: ['Rank', 'Player', 'Age', 'Tm', 'First', 'Pts Won', 'Pts Max', 'Share', 'G', 'MP', 'PTS', 'TRB', 'AST', 'STL', 'BLK', 'FG%', '3P%', 'FT%', 'WS', 'WS/48']
20 columns in finalist table

The data is found within the tr elements of the first tbody element
We want the elements from the 3rd row and on
take a look at the last row to examine data quality

the subset soup object is of type: <class 'bs4.element.ResultSet'>

<tr><th class="right " data-stat="rank" scope="row">13</th><td class="left " csk="Oladipo,Victor" data-append-csv="oladivi01" data-stat="player"><a href="/players/o/oladivi01.html">Victor Oladipo</a></td><td class="right " data-stat="age">25</td><td class="left " data-stat="team_id"><a href="/teams/IND/2018.html">IND</a></td><td class="right " data-stat="votes_first">0.0</td><td class="right " data-stat="points_won">2.0</td><td class="right " data-stat="points_max">1010</td><td class="right " data-stat="award_share">0.002</td><td class="right " data-stat="g">75</td><td class="right " data-stat="mp_per_g">34.0</td><td class="right " data-stat="pts_per_g">23.1</td><td class="right " data-stat="trb_per_g">5.2</td><td class="right " data-stat="ast_per_g">4.3</td><td class="right " data-stat="stl_per_g">2.4</td><td class="right " data-stat="blk_per_g">0.8</td><td class="right " data-stat="fg_pct">.477</td><td class="right " data-stat="fg3_pct">.371</td><td class="right " data-stat="ft_pct">.799</td><td class="right " data-stat="ws">8.2</td><td class="right " data-stat="ws_per_48">.155</td></tr>

to get a player’s 2018-2019 statistical information, we can extract the player’s web page from the BeautifulSoup object

'/players/o/oladivi01.html'

however, the player link is just a stub url and needs a base url for basketball-reference.com appended to it to access the player’s web page

'https://www.basketball-reference.com/players/o/oladivi01.html'

True

create a function to extract MVP finalist data
extract the data we want
and then store it in a DataFrame

the MVP finalist dataframe has 13 rows (player-year observations) and 21 columns

	Rank	Player	Age	Tm	Pts Won	Pts Max	Share	G	MP	...	TRB	AST	STL	BLK	FG%	3P%	FT%	WS	WS/48	player_link
7	8	DeMar DeRozan	28	TOR	32.0	1010	0.032	80	33.9	...	3.9	5.2	1.1	0.3	.456	.310	.825	9.6	.170	https://www.basketball-reference.com/players/d...
8	9	LaMarcus Aldridge	32	SAS	6.0	1010	0.006	75	33.5	...	8.5	2.0	0.6	1.2	.510	.293	.837	10.9	.209	https://www.basketball-reference.com/players/a...
9	10T	Jimmy Butler	28	MIN	5.0	1010	0.005	59	36.7	...	5.3	4.9	2.0	0.4	.474	.350	.854	8.9	.198	https://www.basketball-reference.com/players/b...
10	10T	Stephen Curry	29	GSW	5.0	1010	0.005	51	32.0	...	5.1	6.1	1.6	0.2	.495	.423	.921	9.1	.267	https://www.basketball-reference.com/players/c...
11	12	Joel Embiid	23	PHI	4.0	1010	0.004	63	30.3	...	11.0	3.2	0.6	1.8	.483	.308	.769	6.2	.155	https://www.basketball-reference.com/players/e...
12	13	Victor Oladipo	25	IND	2.0	1010	0.002	75	34.0	...	5.2	4.3	2.4	0.8	.477	.371	.799	8.2	.155	https://www.basketball-reference.com/players/o...

6 rows × 21 columns

Extract the necessary values for the column headers from the table and store them as a list
create the BeautifulSoup object

the columns in the career data tables are: 
 ['Age', 'Tm', 'Lg', 'Pos', 'G', 'GS', 'MP', 'FG', 'FGA', 'FG%', '3P', '3PA', '3P%', '2P', '2PA', '2P%', 'eFG%', 'FT', 'FTA', 'FT%', 'ORB', 'DRB', 'TRB', 'AST', 'STL', 'BLK', 'TOV', 'PF', 'PTS'] 

there are 29 total columns in the career table

let’s examine the data to make sure it conforms with our expectations

player name: James Harden
player years active: ['2009-10', '2010-11', '2011-12', '2012-13', '2013-14', '2014-15', '2015-16', '2016-17', '2017-18', '2018-19']
player career data: ['20', 'OKC', 'NBA', 'SG', '76', '0', '22.9', '3.1', '7.6', '.403', '1.2', '3.3', '.375', '1.8', '4.3', '.424', '.484', '2.6', '3.2', '.808', '0.6', '2.6', '3.2', '1.8', '1.1', '0.3', '1.4', '2.6', '9.9', '21', 'OKC', 'NBA', 'SG', '82', '5', '26.7', '3.6', '8.3', '.436', '1.4', '4.0', '.349', '2.3', '4.4', '.514', '.518', '3.5', '4.2', '.843', '0.5', '2.6', '3.1', '2.1', '1.1', '0.3', '1.3', '2.5', '12.2', '22', 'OKC', 'NBA', 'SG', '62', '2', '31.4', '5.0', '10.1', '.491', '1.8', '4.7', '.390', '3.1', '5.4', '.579', '.582', '5.0', '6.0', '.846', '0.5', '3.6', '4.1', '3.7', '1.0', '0.2', '2.2', '2.4', '16.8', '23', 'HOU', 'NBA', 'SG', '78', '78', '38.3', '7.5', '17.1', '.438', '2.3', '6.2', '.368', '5.2', '10.9', '.477', '.504', '8.6', '10.2', '.851', '0.8', '4.1', '4.9', '5.8', '1.8', '0.5', '3.8', '2.3', '25.9', '24', 'HOU', 'NBA', 'SG', '73', '73', '38.0', '7.5', '16.5', '.456', '2.4', '6.6', '.366', '5.1', '9.9', '.515', '.529', '7.9', '9.1', '.866', '0.8', '3.9', '4.7', '6.1', '1.6', '0.4', '3.6', '2.4', '25.4', '25', 'HOU', 'NBA', 'SG', '81', '81', '36.8', '8.0', '18.1', '.440', '2.6', '6.9', '.375', '5.4', '11.3', '.480', '.511', '8.8', '10.2', '.868', '0.9', '4.7', '5.7', '7.0', '1.9', '0.7', '4.0', '2.6', '27.4', '26', 'HOU', 'NBA', 'SG', '82', '82', '38.1', '8.7', '19.7', '.439', '2.9', '8.0', '.359', '5.8', '11.7', '.494', '.512', '8.8', '10.2', '.860', '0.8', '5.3', '6.1', '7.5', '1.7', '0.6', '4.6', '2.8', '29.0', '27', 'HOU', 'NBA', 'PG', '81', '81', '36.4', '8.3', '18.9', '.440', '3.2', '9.3', '.347', '5.1', '9.6', '.530', '.525', '9.2', '10.9', '.847', '1.2', '7.0', '8.1', '11.2', '1.5', '0.5', '5.7', '2.7', '29.1', '28', 'HOU', 'NBA', 'SG', '72', '72', '35.4', '9.0', '20.1', '.449', '3.7', '10.0', '.367', '5.4', '10.1', '.531', '.541', '8.7', '10.1', '.858', '0.6', '4.8', '5.4', '8.8', '1.8', '0.7', '4.4', '2.3', '30.4', '29', 'HOU', 'NBA', 'PG', '78', '78', '36.8', '10.8', '24.5', '.442', '4.8', '13.2', '.368', '6.0', '11.3', '.528', '.541', '9.7', '11.0', '.879', '0.8', '5.8', '6.6', '7.5', '2.0', '0.7', '5.0', '3.1', '36.1', '', '', 'NBA', '', '765', '552', '34.1', '7.2', '16.2', '.443', '2.6', '7.3', '.365', '4.5', '9.0', '.506', '.525', '7.3', '8.5', '.857', '0.8', '4.5', '5.2', '6.2', '1.6', '0.5', '3.6', '2.6', '24.3', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', 'HOU', 'NBA', '', '545', '545', '37.1', '8.5', '19.3', '.443', '3.1', '8.6', '.364', '5.4', '10.7', '.506', '.524', '8.8', '10.2', '.861', '0.8', '5.1', '6.0', '7.7', '1.8', '0.6', '4.4', '2.6', '29.0', '', 'OKC', 'NBA', '', '220', '7', '26.7', '3.8', '8.6', '.444', '1.5', '3.9', '.370', '2.4', '4.7', '.506', '.529', '3.6', '4.3', '.835', '0.5', '2.9', '3.4', '2.5', '1.1', '0.3', '1.6', '2.5', '12.7']

the player career data is a list that includes every year the player was active
- this list is not broken up by year
to create yearly player data from the player career data list, we need a function to seperate the list
we can use the following function is from StackOverflow
every 29 items in the list represents a year of player data
- we will split the player_career_data object every 29 steps (using the language of the slice_per function) to create a year of data
then we add the column headers to the dataframe as a sanity check that the information was extracted correctly

<class 'list'>

	Age	Tm	Lg	Pos	G	GS	MP	FG	FGA	FG%	...	FT%	ORB	DRB	TRB	AST	STL	BLK	TOV	PF	PTS
0	20	OKC	NBA	SG	76	0	22.9	3.1	7.6	.403	...	.808	0.6	2.6	3.2	1.8	1.1	0.3	1.4	2.6	9.9
1	21	OKC	NBA	SG	82	5	26.7	3.6	8.3	.436	...	.843	0.5	2.6	3.1	2.1	1.1	0.3	1.3	2.5	12.2
2	22	OKC	NBA	SG	62	2	31.4	5.0	10.1	.491	...	.846	0.5	3.6	4.1	3.7	1.0	0.2	2.2	2.4	16.8
3	23	HOU	NBA	SG	78	78	38.3	7.5	17.1	.438	...	.851	0.8	4.1	4.9	5.8	1.8	0.5	3.8	2.3	25.9
4	24	HOU	NBA	SG	73	73	38.0	7.5	16.5	.456	...	.866	0.8	3.9	4.7	6.1	1.6	0.4	3.6	2.4	25.4
5	25	HOU	NBA	SG	81	81	36.8	8.0	18.1	.440	...	.868	0.9	4.7	5.7	7.0	1.9	0.7	4.0	2.6	27.4
6	26	HOU	NBA	SG	82	82	38.1	8.7	19.7	.439	...	.860	0.8	5.3	6.1	7.5	1.7	0.6	4.6	2.8	29.0
7	27	HOU	NBA	PG	81	81	36.4	8.3	18.9	.440	...	.847	1.2	7.0	8.1	11.2	1.5	0.5	5.7	2.7	29.1
8	28	HOU	NBA	SG	72	72	35.4	9.0	20.1	.449	...	.858	0.6	4.8	5.4	8.8	1.8	0.7	4.4	2.3	30.4
9	29	HOU	NBA	PG	78	78	36.8	10.8	24.5	.442	...	.879	0.8	5.8	6.6	7.5	2.0	0.7	5.0	3.1	36.1
10			NBA		765	552	34.1	7.2	16.2	.443	...	.857	0.8	4.5	5.2	6.2	1.6	0.5	3.6	2.6	24.3
11											...
12		HOU	NBA		545	545	37.1	8.5	19.3	.443	...	.861	0.8	5.1	6.0	7.7	1.8	0.6	4.4	2.6	29.0
13		OKC	NBA		220	7	26.7	3.8	8.6	.444	...	.835	0.5	2.9	3.4	2.5	1.1	0.3	1.6	2.5	12.7

14 rows × 29 columns

now we need a function to extract the player career data using the link to the player’s profile as the input
use list comprehension to extract career data for each player link in mvp_finalist_2018_data
use Pandas concat function to store all finalist data in one DataFrame
store all finalist data in one DataFrame

the MVP finalist dataframe has 13 rows (player-year observations) and 21 columns

	Player	Year	0	1	2	3	4	5	6	7	...	19	20	21	22	23	24	25	26	27	28
0	James Harden	2009-10	20	OKC	NBA	SG	76	0	22.9	3.1	...	.808	0.6	2.6	3.2	1.8	1.1	0.3	1.4	2.6	9.9
1	James Harden	2010-11	21	OKC	NBA	SG	82	5	26.7	3.6	...	.843	0.5	2.6	3.1	2.1	1.1	0.3	1.3	2.5	12.2
2	James Harden	2011-12	22	OKC	NBA	SG	62	2	31.4	5.0	...	.846	0.5	3.6	4.1	3.7	1.0	0.2	2.2	2.4	16.8
3	James Harden	2012-13	23	HOU	NBA	SG	78	78	38.3	7.5	...	.851	0.8	4.1	4.9	5.8	1.8	0.5	3.8	2.3	25.9
4	James Harden	2013-14	24	HOU	NBA	SG	73	73	38.0	7.5	...	.866	0.8	3.9	4.7	6.1	1.6	0.4	3.6	2.4	25.4

5 rows × 31 columns

rename columns by concatentating column_headers_player with the two new columns we added in our extract_career_data function

Index(['Player', 'Year', 'Age', 'Tm', 'Lg', 'Pos', 'G', 'GS', 'MP', 'FG',
       'FGA', 'FG%', '3P', '3PA', '3P%', '2P', '2PA', '2P%', 'eFG%', 'FT',
       'FTA', 'FT%', 'ORB', 'DRB', 'TRB', 'AST', 'STL', 'BLK', 'TOV', 'PF',
       'PTS'],
      dtype='object')

examine the mvp_finalist_2019_career_data data

	Player	Year	Age	Tm	Lg	Pos	G	GS	MP	FG	...	FT%	ORB	DRB	TRB	AST	STL	BLK	TOV	PF	PTS
1	Victor Oladipo	2014-15	22	ORL	NBA	SG	72	71	35.7	6.6	...	.819	0.7	3.5	4.2	4.1	1.7	0.3	2.8	2.6	17.9
2	Victor Oladipo	2015-16	23	ORL	NBA	SG	72	52	33.0	5.9	...	.830	0.7	4.1	4.8	3.9	1.6	0.8	2.1	2.4	16.0
3	Victor Oladipo	2016-17	24	OKC	NBA	SG	67	67	33.2	6.1	...	.753	0.6	3.8	4.3	2.6	1.2	0.3	1.8	2.3	15.9
4	Victor Oladipo	2017-18	25	IND	NBA	SG	75	75	34.0	8.5	...	.799	0.6	4.6	5.2	4.3	2.4	0.8	2.9	2.3	23.1
5	Victor Oladipo	2018-19	26	IND	NBA	SG	36	36	31.9	6.9	...	.730	0.6	5.0	5.6	5.2	1.7	0.3	2.3	2.0	18.8

5 rows × 31 columns

Now that we fixed up the necessary columns, let’s write out the raw data to a CSV file.

Write out the career data for 2018 MVP finalists to the raw_data folder in the data folder

Cleaning the Data

Now that we have the raw MVP data, we need to clean it up a bit for data exploration

	Player	Year	Age	Tm	Lg	Pos	G	GS	MP	FG	...	FT%	ORB	DRB	TRB	AST	STL	BLK	TOV	PF	PTS
0	James Harden	2009-10	20	OKC	NBA	SG	76	0	22.9	3.1	...	0.808	0.6	2.6	3.2	1.8	1.1	0.3	1.4	2.6	9.9
1	James Harden	2010-11	21	OKC	NBA	SG	82	5	26.7	3.6	...	0.843	0.5	2.6	3.1	2.1	1.1	0.3	1.3	2.5	12.2
2	James Harden	2011-12	22	OKC	NBA	SG	62	2	31.4	5.0	...	0.846	0.5	3.6	4.1	3.7	1.0	0.2	2.2	2.4	16.8
3	James Harden	2012-13	23	HOU	NBA	SG	78	78	38.3	7.5	...	0.851	0.8	4.1	4.9	5.8	1.8	0.5	3.8	2.3	25.9
4	James Harden	2013-14	24	HOU	NBA	SG	73	73	38.0	7.5	...	0.866	0.8	3.9	4.7	6.1	1.6	0.4	3.6	2.4	25.4

5 rows × 31 columns

Index(['Player', 'Year', 'Age', 'Tm', 'Lg', 'Pos', 'G', 'GS', 'MP', 'FG',
       'FGA', 'FG%', '3P', '3PA', '3P%', '2P', '2PA', '2P%', 'eFG%', 'FT',
       'FTA', 'FT%', 'ORB', 'DRB', 'TRB', 'AST', 'STL', 'BLK', 'TOV', 'PF',
       'PTS'],
      dtype='object')

create dictionaries for renaming columns
rename all columns with dictionaries

	player	year	age	team	league	Pos	games_played	games_started	avg_minutes	field_goals_made_per_game	...	free_throw_pct	offensive_rebounds_per_game	defensive_rebounds_per_game	total_rebounds_per_game	assists_per_game	steals_per_game	blocks_per_game	turnovers_per_game	fouls_committed_per_game	points_per_game
0	James Harden	2009-10	20	OKC	NBA	SG	76	0	22.9	3.1	...	0.808	0.6	2.6	3.2	1.8	1.1	0.3	1.4	2.6	9.9
1	James Harden	2010-11	21	OKC	NBA	SG	82	5	26.7	3.6	...	0.843	0.5	2.6	3.1	2.1	1.1	0.3	1.3	2.5	12.2
2	James Harden	2011-12	22	OKC	NBA	SG	62	2	31.4	5.0	...	0.846	0.5	3.6	4.1	3.7	1.0	0.2	2.2	2.4	16.8
3	James Harden	2012-13	23	HOU	NBA	SG	78	78	38.3	7.5	...	0.851	0.8	4.1	4.9	5.8	1.8	0.5	3.8	2.3	25.9
4	James Harden	2013-14	24	HOU	NBA	SG	73	73	38.0	7.5	...	0.866	0.8	3.9	4.7	6.1	1.6	0.4	3.6	2.4	25.4

5 rows × 31 columns

Index(['player', 'year', 'age', 'team', 'league', 'Pos', 'games_played',
       'games_started', 'avg_minutes', 'field_goals_made_per_game',
       'field_goals_attempted_per_game', 'field_goal_pct',
       'three_pt_fg_made_per_game', 'three_pt_fg_attempted_per_game',
       'three_pt_pct', 'two_pt_fg_made_per_game',
       'two_pt_fg_attempted_per_game', 'two_pt_fg_pct', 'effective_fg_pct',
       'free_throws_made_per_game', 'free_throws_attempted_per_game',
       'free_throw_pct', 'offensive_rebounds_per_game',
       'defensive_rebounds_per_game', 'total_rebounds_per_game',
       'assists_per_game', 'steals_per_game', 'blocks_per_game',
       'turnovers_per_game', 'fouls_committed_per_game', 'points_per_game'],
      dtype='object')

Cleaning Up the Rest of the Data

convert the data to proper numeric types

<class 'pandas.core.frame.DataFrame'>
Index: 121 entries, 0 to 120
Data columns (total 31 columns):
player                            121 non-null object
year                              121 non-null object
age                               121 non-null int64
team                              121 non-null object
league                            121 non-null object
Pos                               121 non-null object
games_played                      121 non-null int64
games_started                     121 non-null int64
avg_minutes                       121 non-null float64
field_goals_made_per_game         121 non-null float64
field_goals_attempted_per_game    121 non-null float64
field_goal_pct                    121 non-null float64
three_pt_fg_made_per_game         121 non-null float64
three_pt_fg_attempted_per_game    121 non-null float64
three_pt_pct                      121 non-null float64
two_pt_fg_made_per_game           121 non-null float64
two_pt_fg_attempted_per_game      121 non-null float64
two_pt_fg_pct                     121 non-null float64
effective_fg_pct                  121 non-null float64
free_throws_made_per_game         121 non-null float64
free_throws_attempted_per_game    121 non-null float64
free_throw_pct                    121 non-null float64
offensive_rebounds_per_game       121 non-null float64
defensive_rebounds_per_game       121 non-null float64
total_rebounds_per_game           121 non-null float64
assists_per_game                  121 non-null float64
steals_per_game                   121 non-null float64
blocks_per_game                   121 non-null float64
turnovers_per_game                121 non-null float64
fouls_committed_per_game          121 non-null float64
points_per_game                   121 non-null float64
dtypes: float64(23), int64(3), object(5)
memory usage: 30.2+ KB

Get the column names for the numeric columns
Replace all NaNs with 0

<class 'pandas.core.frame.DataFrame'>
Index: 121 entries, 0 to 120
Data columns (total 31 columns):
player                            121 non-null object
year                              121 non-null object
age                               121 non-null int64
team                              121 non-null object
league                            121 non-null object
Pos                               121 non-null object
games_played                      121 non-null int64
games_started                     121 non-null int64
avg_minutes                       121 non-null float64
field_goals_made_per_game         121 non-null float64
field_goals_attempted_per_game    121 non-null float64
field_goal_pct                    121 non-null float64
three_pt_fg_made_per_game         121 non-null float64
three_pt_fg_attempted_per_game    121 non-null float64
three_pt_pct                      121 non-null float64
two_pt_fg_made_per_game           121 non-null float64
two_pt_fg_attempted_per_game      121 non-null float64
two_pt_fg_pct                     121 non-null float64
effective_fg_pct                  121 non-null float64
free_throws_made_per_game         121 non-null float64
free_throws_attempted_per_game    121 non-null float64
free_throw_pct                    121 non-null float64
offensive_rebounds_per_game       121 non-null float64
defensive_rebounds_per_game       121 non-null float64
total_rebounds_per_game           121 non-null float64
assists_per_game                  121 non-null float64
steals_per_game                   121 non-null float64
blocks_per_game                   121 non-null float64
turnovers_per_game                121 non-null float64
fouls_committed_per_game          121 non-null float64
points_per_game                   121 non-null float64
dtypes: float64(23), int64(3), object(5)
memory usage: 30.2+ KB

We are finally done cleaning the data and now we can save it to a CSV file.

 the dimensions for the final data are: (121, 31) (rows, columns)

	player	year	age	team	league	Pos	games_played	games_started	avg_minutes	field_goals_made_per_game	...	free_throw_pct	offensive_rebounds_per_game	defensive_rebounds_per_game	total_rebounds_per_game	assists_per_game	steals_per_game	blocks_per_game	turnovers_per_game	fouls_committed_per_game	points_per_game
120	Victor Oladipo	2018-19	26	IND	NBA	SG	36	36	31.9	6.9	...	0.730	0.6	5.0	5.6	5.2	1.7	0.3	2.3	2.0	18.8
25	LeBron James	2018-19	34	LAL	NBA	SF	55	55	35.2	10.1	...	0.665	1.0	7.4	8.5	8.3	1.3	0.6	3.6	1.7	27.4
91	LaMarcus Aldridge	2018-19	33	SAS	NBA	C	81	81	33.2	8.4	...	0.847	3.1	6.1	9.2	2.4	0.5	1.3	1.8	2.2	21.3
39	Damian Lillard	2018-19	28	POR	NBA	PG	80	80	35.5	8.5	...	0.912	0.9	3.8	4.6	6.9	1.1	0.4	2.7	1.9	25.8
32	Anthony Davis	2018-19	25	NOP	NBA	C	56	56	33.0	9.5	...	0.794	3.1	8.9	12.0	3.9	1.6	2.4	2.0	2.4	25.9

5 rows × 31 columns

Review

In this tutorial, we learned how to:
- examine the html structure of webpage
- use functions based on the Beautiful Soup module to parse tables on multiple webpage into .csv
- analyzed a .csv file using the Pandas module

Download this notebook or see a static view here

last updated: 2019-07-01 09:13 

System and module version information: 

Python version: sys.version_info(major=3, minor=7, micro=1, releaselevel='final', serial=0)
urllib.request version: 3.7
pandas version: 0.23.4
Beautiful Soup version: 4.6.3

Source: NBA MVP Comparisons

Notebooks

NBA MVP Comparisons

Part 3

The goal of this blog post is to gather the data necessary to predict the 2018-2019 MVP

Update

Outline

Cleaning the Data

Cleaning Up the Rest of the Data

Review