The Data Analytics Game

Patrick Marchand, Brian van der Voort // September 1, 2021


What Do Your Data Analysts Need to Be Game Ready?

It's a storybook ending to the championship game. The scores are tied and time is running out. Your team has one last chance to score. For the final play, who on your team do you rely on to take that last shot? 

On a professional sports team, there is usually that go-to player. They are the ones who are confident under pressure and have consistently demonstrated their ability to perform.

These are the players who may have obvious natural talent. But more so, they're the ones who have put in the most time to perfect their game. They have the fundamentals burned into muscle memory. It takes no effort to do the easy things and that gives them the advantage to get creative when time (or defense) is tight. 

Getting to that level takes work, dedication, and the resources to learn the skills, to be the best at their position. In the business arena, the go-to player for scrutinizing your data is the data analyst. 


Photo by Jose Francisco Morales on Unsplash

In a high-performing business environment, data analysts are key players within any technical team. They are responsible for distilling data and coming up with the stories, insights, and messages that drive business decisions. 

These analysts need to work together with their team, take advantage of other members’ strengths, and come up with creative options that aim to provide the business with insights that drive informed decisions. Like any team member, a data analyst will need to build up their capabilities by gaining new knowledge, practicing skills, and preparing for game time. It all starts with building a solid foundation. 

In any sport, you need to develop fundamentals; learn how to dribble, swing the club, throw the ball. In the world of data analysis, it starts with acquisition, exploration, and transformation. 

Data Analysis Fundamentals

1. Data Acquisition

Data acquisition is much more than just downloading a file or connecting to a relational database. It’s understanding the structure of the game, what the rules are, and preparing for your offense. In data speak, this is about the metadata that tells you the descriptors, the type of data, expected data density, and other key details. It may seem trivial, but interpreting metadata is like reading the defense and allowing you to develop a strategy for your next play.

2. Data Exploration

Receiving a new data set can sometimes be like practicing with a new teammate. You need to understand their strengths, their shortcomings, what’s possible to improve, and really whether you’ll be able to work together for the required analysis.

Data exploration is simply the process of understanding what data is good for, what kinds of analysis it can support, and what fields you may just want to ignore. It involves a good mixture of subject matter expertise and curiosity to develop confidence in what truths the data will tell.

Some good drills to work on for data exploration include generating high-level statistics, visualizing data distributions, checking for consistent values, and making notes of analysis-limiting values. This last point may reveal other data analysis opportunities that can be achieved with the help of data transformations, where we can change the game without breaking any (data integrity) rules.

3. Data Transformation

Datasets need to adhere to some rules, otherwise, we’d never know how to play the game. However, sometimes the data may be following a relaxed set of rules making it difficult for proper analysis. For this, a good data analyst will have a set of data transformation plays to run. 

If data is stuck together, then split it into meaningful fields. If data is not in the desire units or data types, then convert them. If values are not displayed appropriately or consistently, data cleansing may be required.  

Data transformations may also be required when combining entirely separate data sets or you need to shape the data to make it easier to do analysis. If you need to connect datasets, join based on common keys. If measures are not in the shape you need, consider pivoting or un-pivoting the data. If you need summarized data, then look at grouping or aggregating the data. 

Dealing with extremely large datasets, then considering sampling the data to get a smaller representative set of data. These are some of the data transformations techniques that can take a limited data set and open it up to more scoring opportunities.  

Multilingual Data Analysis Skills


If you watch any good sports highlight reel, you know that players who create the opportunities tend to know their options. One way data analysts can create options for themselves is by knowing more than one programming language.

Every data analyst will know at least one language suitable for analytics: SQL, R, Python, Java, MATLAB, DAX, SAS, Scala, JavaScript, C/C++, Julia, and more. Different languages have different capabilities and one may be better depending on the situation. Being familiar with multiple data analytics and reporting tools is also essential.

This goes beyond knowing how to create charts or pivot tables in Excel. You should have experience with modern data analysis tools such as Power BI and Tableau, which are industry leaders in this space. They provide advanced data preparation, data modeling, and data presentation capabilities.

Knowing different languages and data analytic tools is like being able to switch from right to left-handed: when one approach is looking like it’s about to get blocked, switch hands and find another opening.

Data Analyst Core Competencies

Data acquisition, exploration, and transformation combined with the languages to do these tasks will certainly get things moving, but it’s game over without knowing how to get answers.

The analysis is where the action is: where data becomes information, insights are born; and questions get traded for answers. There are a variety of techniques to work the data from basic statistics and visualizations to more advanced statistical methods, specialized charts, predictive analysis, and machine learning (though the latter may belong in a “data scientist” league.)

The goal here is not to beat the data but rather to refine it into meaningful messages for the business. To capture this action, it often helps to take a picture.

Visual Data Imagery

There’s a lot going on in sports. Things happen so fast that we don’t appreciate the moment that just flew by. A picture puts a pause on the action and freezes the moment. Pictures pack in a lot of information and are also great for showing what’s going on in business such as displaying trends, comparing performers, and monitoring distribution flow.

The standard set of bar charts, line charts, and scatterplots will be in every analyst's visual toolbox but it’s important to keep current with the ever-changing repertoire of chart types (e.g., Sankey, Gantt, Waterfall, Heatmap, Network graph, etc.) that can not only provide variety to reports but can be specialized for specific situations. Using a variety of visualizations is often essential to uncovering the hidden stories and value in the date. 

Professional Presentation of Data 


Photo by Path Digital on Unsplash

A player that gets the attention of the fans has not only mastered the technical skills, but they look professional on and off the field. The same can be said for a good analyst. They make the data look good and can present separate results together in an engaging presentation.

Anyone can create a chart. But, when presenting multiple charts together, there are design considerations that help present components in a unified way. This will make it more appealing, easier to understand, and interesting for the audience.

Consistent color schemes, font choices, and visual alignment are just some of the basic elements that need to be on every analyst’s mind. However, what really matters is to present what is important and compelling to the reader.

Know the Subject Matter

A good data analyst can go through the motions and make the easy shots, but a more effective analyst will understand the subtleties, nuances, and adjustments necessary for a given subject area. This can be achieved using training materials, standards documents, and/or business process diagrams.

It can mean teaming up with the business users and learning from their playoff experiences. Getting out on the floor to see the action is a better experience than viewing operations from data points. Learn by example using existing queries, code, reports, and notebooks.

Get a good coach. Like in anything else, reading a book is no substitute for someone who knows their stuff and can instill good practices.

Cover the Story

The game looked over. The score was reported 19-7 for the home team. However, the score doesn’t report the whole story. Perhaps their best player was put on the disabled list, or a rookie scored 7 of the 19 runs. Simply reporting on numbers does not tell the story that could develop into bigger business insights. 

The point of analysis is to understand what happened, what is happening, why it happened, and maybe even point out what will happen. It is more than just distilling a dataset into statistics. There needs to be enough information to be able to understand that current conditions in one area of the business may have future consequences in others. The data analyst is responsible for telling a good story. 


Photo by Chris Chow on Unsplash

Becoming an effective data analyst, much like professional athletes, requires years of working on perfecting skills on and off the field. Do you or your organization need to up your data analytics game or improve your bench strength? Improving has the training, mentoring, and coaching programs that can help establish effective data analytics fundamentals and practices. Reach out to learn more!

Most Recent Thoughts

How can we help on your next project?

Let's Talk

Like what you see?

Join Us