training ground guru | Mikhail Zhilkin: How to hire your first data scientist

Written by Mikhail Zhilkin – November 2, 2021

INDUSTRIES such as IT and finance are years ahead of football in using data to gain an edge over the competition.

Take Candy Crush, a popular free mobile game franchise that I worked on for four years before joining Arsenal in 2018. It’s run by a double-digit number of data scientists and engineers.

With tens of millions of daily users, it is a logistical and technological challenge to ensure that user activity data is collected, stored and fed into numerous reports and dashboards. The amount and depth of data collected enables sophisticated analysis of player behavior and automated content generation.

Surrounded by experienced colleagues, with the best practices and processes in place, an incompetent person can do just as much damage. In contrast, when a football club hires its first “data manager,” their incompetence can slow data analytics uptake for months or even years.

A training ground is full of people who can test your knowledge of the game, sports medicine, exercise science, etc. There are, however, few people willing to debate database architecture, statistical methods, or writing code.

It’s easy for a budding data science team to get lost and become any of the following:

  1. A toy. In the days of the Internet bubble, everyone wanted a website. Whether the website had done something for the company was irrelevant. Data Science is the new dot-com website: it has the potential to be useful, but it’s not enough to have it – it must serve a business purpose. In football, it’s winning matches.
  2. A marketing tool. ‘Big data’, ‘machine learning’ and ‘AI’ look great in PowerPoint presentations, but what is often missing is how one of them actually changes something. In this situation, data science is just shiny wrapper filled with hot air.
  3. A decision-maker. It is the most difficult to call. It may sound perfectly legitimate: data, analytics, reports and dashboards. Everything is there, and everything is reviewed by the decision makers. But when a decision is backed up by data, people don’t necessarily stop and ask, “Would we have acted differently if the numbers had been different?” If the answer is no, then data science was only a ritual to justify the decision that had already been made. To quote my recently published book, Data Science Without Makeup, “The end goal of data science is to change opinions.”

When it comes to data analysis, the first hire is as decisive as it gets. So how do you find and hire the best candidate?

A common misconception is that to work in soccer you need soccer related experience and skills. This may be largely true for coaches and analysts, but less so for data people.

A data scientist working in a football club can quickly understand what he needs to know about the rules of the game, tactics and physical aspects. On the other hand, a person with a lot of football experience but insufficient data skills will struggle.

If your job posting makes it clear that data skills are paramount and knowledge of football is just a bonus point, you won’t risk turning away perfectly suited candidates.

Beyond basic selection, a CV is primarily an indication of the candidate’s ability to write a CV. I used to be a part of the recruiting process at King, the makers of Candy Crush, and my main takeaway was that CVs are a poor predictor of how an interview will unfold.

A candidate with a poor CV might turn out to be very competent and passionate and we would have an engaging conversation, while a person with an impressive CV full of accomplishments might make me think, “If I can barely stand a 45 minute conversation. with this person I can’t imagine it being good to have him as a full time colleague.

Technical skills are always best judged in action. You probably already have data from the games or the training process (otherwise there are publicly available data sets) and questions you would like to answer using data (if not why are you looking for a data scientist? ?) offer candidates a take-away test.

It doesn’t have to be difficult. It can be as simple as calculating the total distance traveled by each player from the raw tracking data and plotting it as a bar graph.

The best recruiting book I’ve read is “Smart and Gets Things Done,” and the title alone sums it up well. The purpose of the home test is exactly that – to check whether the candidate is smart enough to solve a typical problem, and that he or she is getting things done.

Talking to the candidate about their experience of working on concrete projects is another opportunity to check that their work translates into a tangible result. A fraction of the applicants will return the take-out test and you will hopefully end up with a handful of applicants that you can invite for an in-person interview.

There are two other things you want to focus on when interviewing for a position in football:

  1. Does the candidate understand what the job will involve and agree with it? Football is an unusual industry, and people can have all kinds of ideas. If you’re looking for someone who can take over spreadsheets and replace them with automated reports, you need someone who’s up for it and doesn’t expect to be in charge of all eleven. departure.
  2. Can the candidate explain complex things in a simple way? A data scientist in a football club will mainly address people without a STEM degree. If he can’t make his work transparent to end users, that will limit their impact even if they have the best of intentions. And if they don’t have the best intentions, complexity is the best way to cover up the bullshit.

If you find these pointers useful or even intellectually stimulating, you might be interested in reading my book, Data Science Without Makeup, in which data science recruiting and other topics are explored in detail. Most of the examples used come from my three years (and more) as a data scientist at Arsenal.

Sean N. Ayres