DataMMA in a nutshell
Getting started in Data Science is no easy feet: DataMMA is your one-stop site for becoming a full-stack data scientist and continue to improve your skill set. Data Mixed Martial Arts focuses on both interesting applications and strong fundamentals. We do so with a deep respect for the art and craft of building real-world data products, while keeping a casual, non-scientific tone. DataMMA’s content is and will always be free, but you can apply for master classes to get an edge over the competition. All posts and courses are graded by belt color, as is usual in most martial arts. You start out with the basics (white) and iteratively add more techniques and practices to your training routine until you reach mastery (black). Once you’re ready, you go out and practice your craft in interesting projects at your own pace.
What’s the problem?
Despite being around for more than 10 years now, Data Science is still not a very clearly defined field. You can find a lot of misconceptions around the topic, and in particular about what Data Scientists supposedly do, or should do. Naturally, beginners who want to enter the field don’t quite know what to study, what to expect on the job market and how to prepare for it. Even worse, many companies don’t really know what they want from their Data Science department either. For those reasons, you see many (Junior) Data Scientists in the industry having a tough time getting started and don’t get their ideas shipped to production.
Consider how hard it is to take good first steps and progress quickly in a domain that’s new to you, especially if the field isn’t established yet and has no classic text books everyone is referring to. It’s not easy to find good, comprehensive resources online. Just because the internet creates more and more noise every day, that doesn’t mean it’s becoming easier to find your way around a complex topic like Data Science. While there are, for instance, many good blogs out there on the topic, these blogs tend to jump on the latest and hottest topics, but usually don’t guide beginners on their journey. That tendency can lead to a complete neglect of fundamentals, which will show later on in your work. Without a solid foundation it’s difficult not end up with imposter syndrome and it’s difficult to play catch-up in a fast-moving environment, especially on the job.
If we define Data Science as the art and craft of gaining actionable insights from data and building smart products for real world problems, the term looks like a misnomer. To start with, it only really makes sense outside of academia, so much for the “Science” part. It’s also very limited in scope, somewhat suggesting that practioners are concerned with answering scientific questions using some form of data analysis. In turn, it’s no wonder you hear statements like “back in the day we just called it statistics”. Properly defined, it’s much more than this: it’s Data Mixed Martial Arts. Here’s how we see it:
- It’s a complex cocktail of disciplines: Just like MMA is the fruitful interplay of many martial art forms, DataMMA is rooted in the complex art and practice of various techniques and crafts, applied in the real world. If you focus on just one aspect of the mix, you will likely fail at some point. Focus on strengths, but seek to eliminate weaknesses along the way. Seek solid foundations and pick specialisations that suit you.
- It’s crafty: Working with data requires skillful handling of many aspects and should be approached from a craftsmanship perspective: solid fundamentals, a deep care for the work and attention to detail.
- It’s an art form: Every problem can be approached in a million different ways and everyone brings their unique perspective and skill set to the table. There is no single “answer” to a problem, as if studied in complete isolation, but many different ways to build and form a data product. It has a philosophy: The ethical aspect of working with data, in particular privacy concerns, biases of machine learning models and many others require that every Data Martial Artist has to have a code of conduct, if not a philosophy. Just as no martial artist is trained to attack the innocent, your skills come with a responsibility and have to be applied with care and due diligence. It takes years to master: Sorry to bring it to you, but building smart data products is hard and requires deliberate practice for many years. Don’t expect miracles. This is not to discourage you, every martial art is like that as well. Just keep an open mind, keep learning, and stay humble while doing so. You’ll be a black belt in multiple disciplines sooner than you think. It’s counterculture: Compared to, say, statisticians in an academic context, Data Mixed Martial Artists have a certain coolness to them that comes from building real-world applications. If you need to solve problems for a customer, you have to combine different techniques and can’t be picky about it. You might even have to throw in a cheap trick or two. It’s messy, unpredictable & brutally competitive: Even after more than 10 years, you don’t know what to expect out there, what new trends will emerge and where the market will go. So you need to come prepared. Getting a job is one thing, but striking a top data science job is not easy. It’s a tad bit overhyped and crowded with eccentric superstars: Let’s face it, we’re probably collectively overpromising what can be done. At the same time the field is filled with public figures that work hard to establish themselves as the Conor McGregor of Data Science (yes, I’m looking at you, you know who you are).
What to expect?
Datamma.com is a highly structured, holistic data science blog. Most material is broken down by two dimensions: topic (data martial arts) and difficulty (belt color). This allows beginners to browse easy, entry-level content, while offering complex deep dives for advanced readers. Also, this naturally allows you to bundle graded courses by aggregating by belt color. We use the following building blocks:
- Fundamentals: Working on your fundamentals is crucial in any profession (and martial art), in particular Data Science. Fundamentals are centered around one specific aspect that can be drilled into. Think of these posts as deep dive into a topic. There might in fact be a lot of sessions on the same topic, graded by difficulty. Masters and beginners should practice fundamentals regularly. “I fear not the man who has practiced 10,000 kicks once, but I fear the man who has practiced one kick 10,000 times.” – Bruce Lee
- Toolbox (Weaponry): In depth study of tools, libraries and products from the vast data science tooling industry.
- Data Katas (Forms): A good way to practice your fundamentals, working on these “forms” will improve your memory and technique and solidify your style. Like http://codekata.com , but specifically for data science. Data Sparrings (Fights): Practical and useful applications of your fundamental techniques to data problems out on the streets. This is where things go crazy and many aspects of DataMMA come together so we can build awesome applications together.
- Graded courses : Structured collections of fundamentals and sparrings form the basis for “graduation”. Since we deal with a great many different disciplines to tackle to become a true DataMMA master, there are many graded courses to go through.
- Philosophy: Aspects that underpin and influence your daily work, but aren’t considered part of your core mathematical and technical work. That includes ethical considerations, mental and physical health aspects, how to approach problems and deal with resistance etc. Those aspects are important and can be found in any dojo.
- Learn from the pros: Interviews with interesting guests from the data science ecosphere, written from the perspective of a professional athlete. What are your habits? What makes you unique in what you do? What are your biggest strengths? What do you always struggle with in practice? What made you stumble, but not fall, over the course of your exceptional career?