These Data Lie

Datinuum Newsletter - December 11th, 2023

Data Unfiltered

Lying with Data

Data are beautiful, but the ugly part is that you can lie with data.

I have always been passionate about numbers and enjoyed math because answers were definitive. 2+2 is always 4.

I fell in love with data for a similar reason, but depending on the person presenting the data, they may tell you that 2+2 is actually 6 because they…

• Only focused on the U.S.

• Excluded active members.

• Selected products with >2 sales.

• Filtered on dates within the last week.

Doing any of this is not inherently wrong, but not providing caveats or explaining how you filtered or transformed data is the equivalent of telling a half-truth.

The first humbling experience I had with lying with data came early in my career. I had to develop reports for our clients and was told to add some custom business logic to modify our KPI for win rate because “that’s how we’ve always tracked it.”

My gut reaction was not to change it, but I went along with the plan because:

1. I believed we tracked it this way over time (which wasn’t true).

2. It was early in my career, and I didn’t know the ramifications.

3. I didn’t feel I had the agency to push back.

Once the KPIs were checked and the win rate was verified, I went on a client trip for the first time a few weeks later and had the opportunity to present the data. We arrived at the slide with the “correct” win rate, and I confidently said our win rate was >70%. The client, who hadn’t interrupted thus far, quickly interjected to state, “Our calculations have your win rate at 16%, but I am sure some different variables are used when comparing our calculations.” I continued the presentation but carefully treaded over the remaining data, worried there would be another 60%+ difference.

That embarrassing moment provided me with an obvious revelation. You, your team, and your company are not the only ones monitoring, analyzing, and interpreting the data.

Given this, when you try to force a narrative or manipulate the data to tell your "story,” you may get away with it a few times, especially with people who don’t understand their data. However, you’ll eventually be caught and lose credibility faster than you obtained it.

This is why it is essential to let the data drive the insights rather than to allow your desired insights to drive the data.

Data in the World

Release of Google Gemini

Google announced the release of ChatGPT competitor Google Gemini. Unfortunately, many reports have highlighted that parts of the demo were utterly fake.

EU AI Regulation

The EU created the first notable and extensive A.I. regulations since the generative AI hype took over this year. Regulation may help safeguard against some issues with AI, but the counter is it will likely stifle innovation and economic growth in participating countries.

Crypto Bull Run

Bitcoin has surged to above $40,000, and Ethereum is back over $2,000, with investors excited about potential ETFs for both coins—more notably for BTC.

Data Career Tips

Avoiding Self-Rejection

When landing any role, especially your first, it is crucial not to self-reject.

I’ve coached hundreds of people through job cycles, and the most prominent self-rejecting I have seen is taking role and responsibility requirements at face value.

Requirement 1: Need to have 2-4 years of experience.

Self-Rejection 1: I am a new grad. I won’t apply.

Requirement 2: Have to have managed teams of 20+ people.

Self-Rejection 2: I’ve only managed 4 people. I am not qualified.

Requirement 3: Need to have experience with both healthcare and data fields.

Self-Rejection 3: I’ve only worked in data but not healthcare. I am not what they’re looking for.

The truth is, no one fits every requirement listed in the roles and responsibilities. When I am hiring, and someone fits every requirement, it means one of three things.

1. The requirements didn’t cover enough.

2. The person is lying about something.

3. We have found a purple squirrel.

You may not need 2-4 years because of a great internship you had that provided you with the right experience.

You may not have needed to manage teams of 20+ because you worked cross-functionally with many departments and had dotted lines of influence rather than direct reports.

You don’t need healthcare experience because you're a rockstar with data and can get support from other team members in the business context.

Self-rejecting results in you having a 0% chance of getting any of those opportunities.

Apply to everything, and you never know who may read your experience and want you on their team.

Data Histories

Pythagoras and “His” Theorem

The Pythagorean theorem is significant in surveying land, construction plans, architectural development, and other geometric applications.

Surprisingly, the Greek philosopher for whom the theorem is named, Pythagoras, did not create or discover the Pythagorean theorem. There is no written record of Pythagoras proving the theorem; Euclid, another Greek philosopher, wrote about the theorem in his book Elements 200 years after Pythagoras lived when discussing a different form of the equation known as the Euclidean distance.

Researchers have found the equation from the Pythagorean theorem used in several cultures long before Pythagoras lived:

• Ancient India: 200 years before Pythagoras.

• Babylonia: 1,000 years before Pythagoras.

• Ancient Egypt: 2,400 years before Pythagoras.

The individual who should likely be credited based on the earliest written record of the theorem is Baudhayan, an Indian mathematician from 800 BCE who wrote about the theorem 200 years before Pythagoras’ time.

Datinuumber of the Week: 388

In November, the ADL reported there had been a 388% increase in antisemitism in the U.S.

Early last week, MIT, UPenn, and Harvard presidents failed to denounce antisemitic language at their respective universities. While all three walked back or clarified their remarks after the hearing, only Liz Magill from UPenn has resigned as of this writing.

Thank You

Thank you for subscribing and reading this week’s newsletter.

If you enjoyed the newsletter, the best way to help is by sharing it with colleagues and friends.

If you prefer to listen to the newsletter, the Datinuum Podletter will be released weekly on Apple, Spotify, or your favorite podcast player.

The video podcast will also be released weekly on YouTube.

Feedback is a gift. Please reach out to [email protected] with any feedback or questions.