[Waterstpub87]

For installation of python and selenium, please see this post

https://www.sportsbookreview.com/for...-rotowire.html

The background of this script was my changing from offshore to US books completely for this baseball season. Unfortunately, there is a large difference in the MLB props for offshore sites as compared to onshore sites. Typically, I had bet Hits/Runs/RBIs as a sum, which was the bets offered by the offshores. They offered it as a matchup, typically 2-4 matchups per games, and typically 4 single player over/unders. Similar with total bases. They had hits as well.

Onshore, it seems to be that the offering is individual players. Needless to say, trying to manually price 18 * 4 props (18 players hits,runs,rbis,bases) is time consuming. I wrote this script to pull in the tables I wanted, and configured excel to read and price them.

Draftkings prop page is set up into tables, IE 1 table for home run, 1 table for RBIs, ect.

A note to remember is that python starts at 0 for numbering. So the first table (Home Runs, which would be 0) is ignored.

This script will pull tables 2, 3, 4 ,5, which are Hits, Bases, RBI's and Runs.



Below is example output:

	Player	Category	Line	Over Odds	Under Odds
0	Abraham Toro	Hits	0.5	-180	130
0	Adam Frazier	Hits	1.5	185	-250
0	Alex Bregman	Hits	0.5	-195	145
0	Chas McCormick	Hits	0.5	-165	120

It will produce an excel sheet with this information.

[Waterstpub87]

import pandas as pd

from selenium import webdriver



cols  = ['Player','Category','Line','Over Odds','Under Odds']

props = pd.DataFrame(columns = cols)

driver = webdriver.Chrome('Path to your chrome driver')

driver.get('https://sportsbook.draftkings.com/event/180314630?category=odds&subcategory=batter-props')
tables = driver.find_elements_by_xpath('//div[@class="sportsbook-event-accordion__wrapper expanded"]')

for x in tables[1:5]:
    eventype = x.find_element_by_xpath('.//a[@aria-current="page"]').text
    rows = x.find_elements_by_xpath('.//tr')
    if eventype == 'Home Runs':
        betype = 'HR'
    elif eventype == 'Hits':
        betype = 'Hits'
        
    elif eventype == 'Total Bases':
        betype = 'TB'
        
    elif eventype == 'RBIs':
        betype = 'RBI'
    elif eventype == 'Runs Scored':
        betype = 'Runs'
    else:
        betype = 'Error'
    
    
    for z in rows[1:]:
        currentbet = pd.DataFrame(columns = cols)
        currentbet.at[0,"Player"] = z.find_element_by_xpath('.//span[@class="sportsbook-row-name"]').text
        currentbet.at[0,"Category"] = betype
        currentbet.at[0,"Line"] = z.find_element_by_xpath('.//span[@class="sportsbook-outcome-cell__line"]').text
        betodds = z.find_elements_by_xpath('.//span[@class="sportsbook-odds american default-color"]')
        
        currentbet.at[0,"Over Odds"] = betodds[0].text
        currentbet.at[0,"Under Odds"] = betodds[1].text
        props = props.append(currentbet)
        
props.to_csv('MLBpropscrape.csv')    
driver.close()

[Waterstpub87]

To use this script, you must paste in the url of the game here:
driver.get('https://sportsbook.draftkings.com/event/180314630?category=odds&subcategory=batter-props')

you can go to the draft kings website to find this. Just click on a game on the main page, and you can find the event number.



Additionally, I take only the 4 tables. If you wanted everything ( so what we have and HRs+Singles+Doubles+Triples + Stolen bases,

go here :
for x in tables[1:5]:

and make it look like this :
for x in tables:

I rename some of them, which is the part that looks like this:


elif eventype == 'Total Bases':
betype = 'TB'

if You are pulling in all the tables, the name will become "error" due to the coding. To fix this,

take this part :
betype = 'Error'

and change it to:
betype = eventype

This script will output the table to where ever you have the script saved.

If you want to save it somewhere else, specify it like this:

directory = 'C\\Users\\Documents\\'
It must be formated like that, double slashes within single quotes

change this line:
props.to_csv('MLBpropscrape.csv')
to
props.to_csv(directory + 'MLBpropscrape.csv')

[peacebyinches]

This looks great! Something like this is the kind of tool for streamlining player props (which like you said, can get out of hand fast with the multiple stat categories across multiple players across multiple games) that I could definitely use. Thanks for sharing!