I’m on record that solving for pattern is a better approach to knowledge work than connecting the dots. Connecting the dots is pretty common usage. I don’t think it needs much elaboration. Except to point out that there is a presumption that someone else has already drawn the picture you are now expected to reveal.
It’s about locus of control. Connecting the dots grants you only the task of discerning what someone else has already done.
Solving for pattern puts the responsibility where it belongs; on you. I encountered the term in Wendell Berry’s essay of the same name (add link).
A few days back I picked on the notion that if something works for you 10,000 times you have license to assert you have proven something in a substantive way. Implicitly I was asserting that there is only one path to reliable knowledge—experimental and replicable science. Pretty arrogant for a former case writer.
What I was arguing, perhaps badly, was that you need to be careful about the claims you can make about what you think you know. Solving for pattern is a shorthand for the process of building explanations from the experience (first and second hand) that you accumulate and models of the underlying systems you are interacting with along the way.
Solving for pattern is an active process that never ends. Each new cycle through the pattern is an opportunity to try a new variation or to investigate how changes in the external environment tweak the outcomes you expect. You’re free to try any “proven system” you’d like. You’ve also learned to be appropriately skeptical about other people’s proof.