This is a running list of the projects that I’ve worked on. These correspond to repos that can be viewed on my GitHub that do not have blog posts associated with them. If interested, please see the README files in the respective repos for more detail.

For my own (and the reader’s) convenience, I categorize the projects here at the top level by whether or not I’m actively working on them. Other subcategories are used appropriately.


NBA Decision Analysis

  • Time: May - Jun. 2016
  • Duration: 1 month.
  • Goal(s): Augment previous Twitter sentiment analysis.
    • Apply Decision Analysis (DA) concepts to real-world context.
    • Expand upon existing research related to monetary value of NBA draft picks.
  • Deliverable(s): Old website posts, now converted to a bookdown e-book.

Under Active Development

Presentable (But Not Complete)

Not Presentable (Yet)

General Analysis of the NBA

  • Time: Collecting data manually since 2012. Started R analysis in 2017.
  • Duration: Approx. 3-6 months total
  • Goal(s):
    • Expand upon previous work on predicting team win totals.
    • Apply machine learning (and other techniques) to identify biases in personal prediction behavior.
    • Identify potential “market inefficiencies” (i.e. profitable opportunities) in betting market.
  • Deliverable(s): Ad-hoc visualizations.
  • TODO: Create an e-book?

General Analysis of the NFL

My work/goals with NFL data mirrors what I have done/seek to achieve with NBA data. I have not spent nearly as much time with this project, however. I tend to shift my efforts depending on which sports is in season (i.e. NFL in the fall, NBA in the spring), while slightly favoring the NBA in general.


Automated Email Reports Through Gmail

  • Time: Nov. 2017
  • Duration: 1 - 2 weeks
  • Goal(s): Create an automated weekly/monthly report/email to provide an update regarding the status of personal weekly NFL (and, in the future, NBA) weekly/season/all-time predictions.
  • Deliverable(s): “Proof-of-concept” RMarkdown output.
  • TODO: Complete full implementation.

Predicting NBA Team Win Totals (Early 2016)

  • Time: Jan. 2016
  • Duration: 1 month
  • Goal(s):
    • Learn python.
    • Discover biases/methods to improve personal predictions of NBA team win totals.
  • Deliverable(s): Visualizations of new “back-cast” predictions (as .pdf files).

Predicting NBA Team Win Totals (Mid 2017)

  • Time: Jan. 2017
  • Duration: 1 month
  • Goal(s):
    • Learn R.
    • Expand upon prior work with python on same subject.
    • Explore regression and classification machine learning methods.
  • Deliverable(s): RMarkdown report documenting discoveries, including:
    • Visualizations of training/testing accuracy of models.
    • Error of models.