M. Edward (Ed) Borasky

is creating tools for daily fantasy sports analytics

Select a membership level

Become A Patron
$1
per month
Unlock exclusive content and join my community of patrons. Thank you!
  • Access to patron-only content
  • Connect with other patrons
  • Plus all previous rewards
Annotated Bibliography
$5
per month
I do a tremendous amount of research on daily fantasy sports analytics. You'll get my complete bibliography in human- and machine-readable form.
Weekly NHL Newsletter
$10
per month

I'm starting a weekly NHL newsletter for the 2019-2020 season. It will come out every Monday and feature a recap of the previous week's games and projections for the coming week.

0

patrons

$0

per month

About

I'm a data scientist who discovered daily fantasy sports about five years ago. I'm currently playing on a very small scale, mostly using optimizers available on the web. I'm breaking even ... but I want to scale up.

And I've found that I'm limited by the tools available. For openers, they're expensive. I don't mind paying for tidy data, or even raw data via an API, but paying a monthly fee for analysis tools I can build myself seems counter-productive.

Moreover, the currently-available optimizers and their underlying projection systems are mostly black boxes. And there's no quality assurance / testing readily available, even though there are well-known methods for doing that. Both projection and optimization are core technologies of operations research - management science and have been since World War II!

So I'm building my own tools and publishing the software as open source. The math is challenging and one of my goals is to teach others how to do it. The technologies I use are cutting-edge: R, RStudio, the Tidyverse, Keras and PlaidML. No more black boxes!

You'll be able to use the software without spending a dime above and beyond what you've invested in your laptop or workstation. Because it's open source, you'll be able to run it on Windows, MacOS or Linux, or even in the cloud. And I'll be working with data from providers who are also here on Patreon. Watch this space!
Goals
$0 of $100 per month
  1. Tidy API data acquisition: I'm working with the MySportsFeeds.com API. They are here on Patreon. The data acquistion / tidying for this API is done and is in the package at https://github.com/znmeb/dfstools.
  2. Build a model using mvglmmRank to predict game scores for the current MLB seasons. I have some code to do this already, but it only works for Kaggle NCAA basketball ("March Machine Learning Madness") data.
  3. Expand my "Archetypal Ballers and Ternary Plots" lightning talk (http://rpubs.com/znmeb/pdxdataviz20170209) to a longer presentation / article / notebook. I'm in the final stages of getting this working for WNBA; watch this space!
  4. Build an R notebook for analyzing DraftKings and FanDuel contest histories. This is absolutely core to my process; without knowing how your lineups compare with the competition you're just gambling.
1 of 3
By becoming a patron, you'll instantly unlock access to 1 exclusive post
1
Writing
By becoming a patron, you'll instantly unlock access to 1 exclusive post
1
Writing

Recent posts by M. Edward (Ed) Borasky

How it works

Get started in 2 minutes

Choose a membership
Sign up
Add a payment method
Get benefits