To maximize on the growing reliance on predictive analytics
and sports data manipulation, RotoRithms has developed a business model to
offer the raw data and dynamic predictive analytic models that forecast
individual player value based on current player performance. This report will
prove just how that data was made possible post web scraping processing, but
pre-analysis to be delivered globally on the website through VBA scripts.
Objectives
To get reliable data up on the new website available for
purchase by February 25, 2015 for historical data of the NBA, NFL, MLB, NCAAB,
College Football, and NHL. Then devise the automated system to upload new and
current data for sports during their seasons to the website by 8 AM Eastern
Time for on-site purchases.
Highlights
·
Data was sorted by sport into appropriate
folders prior to SAS processing
·
Cleaned data with missing data values
·
User forms to call specific Mozenda API’s run
web scraping scripts
Keys to Success
·
VBA Instruction from BYU’s Dr. Allen
·
Mozenda Data Scraping Software
·
EmailItIn FTP Server
·
WordPress
Summary of Benefits
·
VBA scripts will save numerous hours for
RotoRithms of having to pull up the google drive folder and finding the files
and then placing those files into the appropriate folders to be able to start
analyzing. This automated macro is saved to the ribbon and can be easily accessed
from any open Word document as requested by RotoRithm.
·
All data documents are properly cleaned to now
include key date fields to run predictive analytics on
·
On a semi-regular basis, a macro I wrote will reverse
compile a pdf document that tried to compile into a Word document but the formatting
is off, and then macro will save over 2 hours each time the Word document needs
to be reformatted correctly in a Word environment.
·
To minimize number of times to run the
algorithms to determine most profitable players to pick up for fantasy leagues,
my Macro will be able to compile all the disparate documents into a single
file.
·
For any rescrapes of particular days for a
particular sport, I created a customizable user form that enables the user to dynamically call Mozenda APIs to run the
Mozenda Script that will pull in the information directly to the Google Drive folder.
No comments:
Post a Comment