A Beginner’s Data Science Journey
Background
I don’t know where this journey will end. But I just want to document it anyway, in case I can help someone who is on a similar path. I’m new to the field of Data Science — for now it is just a hobby, and don’t even know if I will last the distance. Yet should I happen to make something out of this hobby, I want to do that with documenting my learning throughout. If anything, I hope to use these blogs as ‘repositories’ (new word I learned from Data Science) where I store some of my new-found discoveries.
Who am I?
My name is Lee and I’m a massive sports fan. For years, I have written about sports (mostly on my Facebook), had an irregular blog, tweeted and retweeted about sports and been encouraged by many friends to follow my passion. But I felt the market was saturated, and there was nothing new to be written out there. I’ve always had another passion — numbers! As a 12 year old, I had a small notebook where I kept my ‘data’ from some school football matches. I followed every match and kept details of games played, goals scored, points etc. However, I pursued an Accounting degree and later started working in the non-profit space. But the soccer passion never goes anyway.
The world of Twitter parodies
Fast forward a couple of years and I moved onto Twitter (the whole world did), followed lots of English football pages and discovered a whole new way of communicating about the beautiful game. My favorite accounts were parody accounts like @WengerKnowsBest, @DeludedBrendan & @EvilKagawa but I also enjoyed following @OptaJoe — my first real peek into analytics. One day in 2012, a friend and I thought — what if we were to start our own parody account — to mirror what we’d seen, but based on local football. The account has grown to almost 20K followers today and is one of the most recognized twitter accounts in South African football. Moreover, I was starting to understand the cliche-world of conventional football commentary, and disliked the inconsistencies in commentary and reporting about my favorite club Arsenal, and many other clubs.
From parody to analytics
Having seen the parody account grow, we started an unofficial Opta ‘analytics’ account — again based off the English version Joe, but with a name changed to local name Jabu (named after Jabulani — the 2010 World Cup Soccer ball) and one of my favorite local footballers Jabu Pule (now Mahlangu). Only in February this year did the account come to life, sharing basic trends & anomalies in local football. As the page followers grew, I knew we had to get better at this, so I dug deeper into the world of sports analytics & discovered the world of Data Science — and the journey really started gaining traction:
1. I changed my entire Twitter timeline — unfollowed many random pages and started following industry leaders in sports analytics & data science — it was like joining twitter all over again — a whole new world before my eyes.
2. I subscribed to podcasts (Analytics FC, Learn to Code with Me, Becoming a Data Scientist, Data Crunch, Code Newbie).
3. I read a lot of articles on StatsBomb, Medium, Sloane Sports, Quora and more.
4. I started taking free intro courses on Data Science (DataCamp, KD Nuggets, Udacity, Pluralsight, Cognitive Class).
5. I joined webinars across the world (Metis, Analytics Vidhya, Sloan Sports conferences) just to learn more.
100 days of code
So after a few months of randomly reading, watching and getting very very confused, I was convinced that any type of future in this field would be best served by learning Python or R. I spent a lot of time trying to decide which of the two was best for me, but in the end settled for Python (for many reasons) and so started my #100daysofcode. There are many courses and resources to help you learn out there, but I settled to start with Data Camp courses. I’m sure other courses are just as good (if not better) — but it all depends on your style of learning. In the end, I just had to choose one and run with it. So it’s Data Camp. For now…