What do professional sports bettors do? Episode 2

At this point, you understand the concept behind what a professional sports bettor does to capitalize on a perceived edge in the market. How they do it is a whole nother story involving the wearing of multiple metaphoric hats.

As they say, it’s best to start at the beginning.

Read More

First Sweat

My journey into data science began as every data science student’s journey tends to begin: by placing a bet.

Putting your hard earned money down on a sporting event is an experience like none-other on this planet. It is quite difficult to describe, to a non-sports fan, the experience of “sweating” a game. The fate of your funds lying in the hands of athletes that will never know your name, never think about your investment, and never care.

One simple little wager would set off a chain of events that eventually brings me here, to your computer screen. The details of the wager are seemingly insignificant at this point in time, but if you stick around, I might share the details with you.

What that little wager ultimately signified was the beginning of my data science career.

After collecting the winnings on my first wager ever placed, I thought I was a natural. All I needed to do was comb social media, read some articles about how Team A was going to beat Team B, place a bet, and walk away with a winning ticket. Well… unfortunately, turns out sports betting is much more complicated. A few weeks later, as I was checking my account balance at the online casino I had deposited all of $100 at. I noticed the account was strangely low on funds. I guess I was not doing enough research into the teams. Better start listening to podcasts. I was driving 6-8 hours a day for work, so I had plenty of time to absorb some sports information though my earballs. The analysts on the radio would list statistic after statistic about whatever sporting event was happening the next day. When I got home, I would open Microsoft Excel and start copying box scores from ESPN, Football-Reference, and any other website that publicly compiled sports statistics. I had no idea what I was doing, but I knew I needed to do SOMETHING. The best way to describe my process is to say I was mashing numbers together and hoping they would spit out an actionable piece of information I could use to make a bet.

My process was not rocket science or even mathematically sound. But for some reason, I started to lose less. A lot less. One might even dare to say, I started to win. Well, lets not go that far… I started to lose less!!


Over time, my Excel files started to grow. I figured out macros. Now I was scraping data. My Excel files were starting to get big. Excel has graphs!? Oh boy. My Excel files were getting huge. Guess I need to split some of this up. A tangled web of Microsoft Excel goodness was taking form on my PC. I learned about linear regression and some other basic data analysis techniques. Index & Match functions by the dozens were littered across many of my documents. I was actually developing a process. A standard operating procedure. And I was winning! What an incredible feeling. Dreams of quitting my day job were almost coming into focus. I needed to get better!

More data. More index & match functions. A little more data! More. More… MORE!

Oddly enough, Microsoft Excel started to slow down. I would go to change a cell and Excel would lock up for 2-3 seconds. Not the end of the world, but far from ideal. A few seasons went by and my datasets were getting larger and more complex. Now I would go to change a cell and Excel would lock up for 10-15 seconds. There had to be a better way.

I had been hearing about this popular and nimble little programming language called Python. Apparently it was really good for working with numbers for the purpose of making predictions. I downloaded Python and decided to give it a try. Quickly I realized I was in over my head. Coincidentally, a YouTube advertisement for a Python class on Udemy popped up as I was researching some sports match ups. (Weird, how did YouTube/Google know I was exploring the realm of coding… So strange…)

I signed up for a $7 Python crash course and starting learning the basics. The term, “machine learning” had begun being tossed around some sports betting circles I ran in, so I typed “machine learning” into Udemy’s catalog, found a course that I liked, and bought that one too. The course was taught in Python and R simultaneously. For me, R was so much easier to use, so I stopped taking coding lessons in Python and completely switched to R.

Machine learning basics were starting to make sense to me. Now I was building…mostly… mathematically sound models. My win percentage on wagers started to go WAY up. Instead of making hundreds in a season, I was making thousands. I was well on my way to sports betting success.

Python eventually found its way back onto my computer screen. Now I’m coding in two languages. My scripts in R are talking to my scripts in Python. I’m scraping significantly more data. My processing speeds are increasing. Most importantly, my win percentage is continually on the rise. But, I still need to get better! I know there is so much that I don’t know.

Enter General Assembly and the Data Science Immersive. I enroll in my first formal class for Data Science. I need to take my skills to the next level.