I think the trick is to let the data speak first, then fit a story around it that makes sense, not the other way around. We gather fragments, analyze them, and only after the evidence is clear do we decide what narrative to weave. If we start with a story, we’ll skew the interpretation like a biased compass. Respect the past by treating each piece like a clue, not a prop. And if a piece doesn’t fit our hypothesis, that’s not a failure—it’s another line in the puzzle waiting to be solved. The line is thin, but staying on the evidence keeps us from turning archaeology into a vanity project.

I keep the data and the story in separate folders, literally. First I clean, catalogue, and analyse with raw statistics, no assumptions. Then I let the numbers suggest patterns, not dictate meaning. After that I draft hypotheses, and finally I write the narrative, always circling back to the raw evidence to confirm every claim. Peer review is a must—fresh eyes catch when a narrative is sliding in. And I keep a running log of every decision, so I can see if I slipped from data‑driven to story‑driven. That’s how I keep the gentle pull at bay.

Sure, the pull shows up when you’re staring at a blank page and the past starts whispering “this is what you need.” That’s when the log feels more like a safety net than a compass. I log every assumption, every question that pops up, and let the numbers answer before I write the story. Still, a good review from a skeptic—someone who can spot a bias I missed—is the real check that keeps the narrative from creeping in.