Richard’s Rules for Data Scientists

Published

December 9, 2025

Abstract

One person’s thoughts on becoming a responsible data scientist.

The following themes occur throughout my courses. In the spirit of the Jack Sparrow movies, consider them as

“more what you’d call ‘guidelines’ than actual rules.” Hector Barbossa

Rules for Data Scientists

1. Be Responsible

  • Consider laws, professional guidelines, and ethics
  • Think about how your work affects others
  • Act with intention and integrity
  • Own your mistakes

2. Follow a Life Cycle

  • Use a consistent analytical framework
  • Understand the question before solving it
  • Revisit earlier steps when necessary
  • Check your work after it is deployed

3. Respect the Data

  • Understand the source, context, and measurements
  • Protect and preserve the data
  • Check for errors, bias, gaps, and limitations
  • Treat extreme values as information, not “outliers”

4. Think Before You Code

  • Create a plan or strategy
  • Balance creativity (art) and rigor (science)
  • Consider alternative approaches
  • Allow room for discovery and iteration

5. Write It Down

  • Use literate programming, e.g., Quarto
  • Record assumptions and decisions
  • Document and comment code thoughtfully
  • Use version control

6. Make Your Work Reproducible

  • Use relative paths
  • Use environments (renv, uv) to manage change
  • Organize projects with consistent structure
  • Design for modularity

7. Build a Little, Test a Little

  • Work in small chunks (with pipes)
  • Project and check intermediate results
  • Follow a style guide and avoid “hard coding”
  • Test forwards and backwards

8. Communicate Effectively

  • Listen. Ask clarifying questions
  • Start with the bottom line
  • Tailor presentations to the audience
  • Focus on outcomes and impacts

9. Ask for and Offer Help

  • Use help documents and other sources
  • Solicit and support peer reviews
  • Engage with technical communities
  • Share what you learn

10. Plan, Project, Monitor, Update

  • Follow a project management life cycle
  • Draw a simple “Killer Graphic” that shows your workflow
  • Use a WBS to break the elephant into bite-size chunks
  • Estimate cost, performance, and schedule early
  • Monitor progress and risks at least weekly
  • Collaborate with stakeholders frequently
  • Update the plan as life happens

11. Keep Learning

  • Stay curious
  • Request and reflect on feedback
  • Follow developments in tools and methods

12. Support Your Team

  • Contribute as a member
  • Lead with empathy and respect
  • Build trusting relationships
  • Share credit freely

13. Strengthen Your Network

  • Connect with peers and mentors
  • Build your professional brand
  • Maintain strong and weak ties across time

14. Embrace Uncertainty

  • Accept that projects and data begin in ambiguity
  • Meet the blank page with curiosity, not fear
  • Treat errors as information, not failure
  • Iterate, revise, and adjust course as you learn

15. Find Joy

  • Notice what energizes you
  • Celebrate progress. Hunt the “good stuff” every day.
  • Create space for play, wonder, and curiosity