A friend recently asked me what I thought about parlay betting. I figured this would make a good case study on practical data science. I’ll go over this like what would normally happen in a consulting engagement.

Let’s get into it.

The Question/Problem

Friend thought parlay dumbing was a dumb bet and that sports books penalize the payouts based on the UNLV Published Gaming Reports.

Looking at this data it’s striking how much parlays result in wins for sports books. How would you explain this? What about the other sports? What does that show?

Load the Data

First, let’s load and plot the data. We will slice it by sport and aggregate it across all sports. If you had a lot of data we could use PDF extraction. But that’s unnecessary for this, we can just copy, paste, and parse the resulting formatted string.

From here go to the Jupyter Notebook and watch my Youtube video.