Rant about “Data Scientist”

The following is in response to what was originally published by Aureus Analytics, titled "Demystifying a data scientist", some time ago.



The reason why I have chosen this old article to express my opinion  is because this article is very precise articulation of what is still commonly understood behind term “Data Scientist”.  So, in this article red is presented as myth, green letters are “Aureus” view of what they think is  factually correct, and blue is my humble opinion.Data Scientist

I think their blog is really great, I certainly recommend it – so this is not to knock them down in any way. It is just an attempt to be somewhat “contrarian” and it give different angle when approaching term “data scientist”.  I just wish readers NOT to take this "rant" overly seriously.


I could and I will write (on a good day) other articles that are completely pro "Data Scientist" term. Today was not the day. And lastly  – I do consider myself data scientist for time-being, at least… until Harvard Business Review invents another term ...




Define “Data Scientist”, Geeks with gigantic glasses and round bellies.

Data scientists are warm, pleasant individuals like any of us; who are very adept at analyzing data, seeing non-obvious patterns in large volumes of data, and contributing significantly to creating a competitive edge for the business.

This “fact” is somewhat distorted. No human can see “non-obvious” patterns in data just by looking at it! Technology is needed to extract the patterns, but then how do you separate useless, trivial from actionable and unknown patterns that may solve business problems? And proposed answer is – just employ data scientist! Hmmm..


Want to be a Data Scientist? Fancy tool are all you needed!

Tools are just tip of the iceberg. All the tool in the world can’t make you a Data Scientist if you don’t have sound knowledge of mathematics, programming and business.

Think about this – “sound knowledge mathematics, programming and business”! I presume statistics, computer science as well as different methodologies are all part of it!? Do we mean – well, data scientist should at least know some facets of it. Problem this is so “loosey –goosey” that anything goes under it! And when you throw all these knowledge disciplines together, question becomes  not where is the weak link in this knowledge conglomerate, but how many of them are?  


Data Scientists serve no purpose and are a noveau fad that will soon fade.

Machines will solve complex equations, sort humongous volumes of data but you always need human expertise to figure out what to do with it. Be rest assured; Data Scientist are here to stay and robots aren’t taking over anytime soon.

The term Data Scientist homogenize an inherently heterogeneous array of technologies, skills and purposes, so If I tell you that I am data scientist – since there is no science without data – I might as well drop the term data and call myself a scientist and you will be equally confused of what I  am able to do.


 Data Scientists put data into a magic hat, wave a wand over it, and voila! Game changing insights magically appear!

Data Scientist aren’t magicians. They are just normal folks with sound sense of logic and incredible capabilities. A keen eye for non-obvious patterns in obvious data is their super power. 

When you see this “incredible capabilities” description you wonder of who benefit of sky-high expectancy that is been created! To bridge the abyss between business question and business solution – there is no a superhero to fly you from one side to the other. It takes organizational support, it takes team work comprising disparate skills, it takes proven methodologies and best practices, it takes technology and when all this is deployed correctly on right data– I call it successful project, others can call it Incredible Data Scientist.


Data Scientists ate statisticians and physicists who failed to make the cut.

While a lot of Data Scientists have an academic background in statistics or physics, many of them have consciously chose to solve business problems. And a lot of them are experts in the business domain they choose to operate in.

I don’t know where physics comes in into the mix? Would Einstein, Hawking, Heisenberg, Feynman see themselves as data scientists? And if they were - would you confuse them with someone who use programming languages to show you pie chart? Unless, that is mischaracterization of what Data Scientist does.


Data Scientists are BI experts with fancier titles.

BI Specialists are kind of Data Scientist, but the role of Data Scientists goes much beyond and include predictive modeling, graph analysis, and more.

So, Data Scientist he can do much more than create pie chart. It looks like there are many "flavors" of data scientists. Imagine you go camping and you scream “animal in my tent!!” Is it the ant, mouse, spider or poisonous snake? It makes "big-data" difference in terms of the response - knowing what's in tent.


Only large enterprises need Data Scientists to keep their ships floating.

Any company, no matter how big or small needs an expert who can analyze and search for patterns that can help determine future risks and loopholes as well as scope for improvement.

Term expert is usually used for someone who uses specific set of skills and technologies to provide results better than, shall we say "non expert". So, if you going to call me Data Scientist - your expectancy of what I can do for you  is wholly determined by your own understanding of this horribly loose term. And because it means so many different things to so many people - it confuses rather than enlightens.


1 Response

  1. Great artcile, thank you again for writing.

Leave a comment