Capstone: decoding a headline

Here is a headline, the kind you scroll past a hundred times a week: “Crime surges 50% as new mayor takes office.” It feels solid. It names a number. It sounds like a fact you could repeat. But you have spent twelve lessons learning that a story is a made thing, shaped at every step by choices. So don’t repeat it yet. Run it through the machine you’ve built. By the end of one pass, you’ll know whether to believe it, doubt it, or throw it out — and you’ll know exactly why.

The reader’s checklist

You already own every tool. The capstone is just using them in order, fast, on a single claim. Six questions, drawn from the course:

• Selection (lesson 1) — why am I seeing this and not the ten quieter true things?
• Agenda and sourcing (lessons 2, 3) — who benefits from this being the story, and says who — a checkable document, or “sources say”?
• Framing and the headline (lessons 4, 5) — do the same facts allow an opposite story, and does the headline oversell the article beneath it?
• The numbers (lesson 6) — points or percent, how small is the base, what window was chosen, where’s the denominator?
• Omission (lesson 7) — what’s missing that would change my mind?
• The incentive (lessons 8–11) — is this shaped to win a click or an algorithm, and is it honest spin or deliberate manipulation?

A claim that survives all six is Sound. One with a real grain but missing context or overstated is Shaky. One that’s wrong or badly misleading is Oversold. Most claims are Shaky. Clean lies and clean truths are both rare.

A worked decode

Take the mayor headline through the machine, one question at a time.

The numbers first, because “50%” is where the work hides. Fifty percent of what? You read the article and find the city logged 6 murders last year and 9 this year. That is the 50% — three extra deaths in a city of two million. The percentage is true and the base is tiny. A jump from 6 to 9 swings 50%; a jump from 600 to 900 in a bigger city is the same percentage but a completely different thing. The small base inflated a small change into a scary number (lesson 6).

The window. Why measure from last year? You dig and find that last year was the city’s lowest murder count in two decades — a freak low. Measured against the ten-year average of 11, this year’s 9 is a fall. The reporter picked the one starting point that turns a decline into a surge. That’s a cherry-picked window (lesson 6).

The framing. “Crime surges as new mayor takes office” welds two true facts — the count rose, the mayor is new — into a third claim the facts don’t support: that the mayor caused it. He’s been in office two months; these cases predate him. The same numbers frame just as easily as “Violence stays near a historic low in the mayor’s first months” (lesson 4).

The headline versus the article. The headline says surges. Paragraph nine of the article quietly admits the total is “broadly in line with the long-term average.” The headline did work the article wouldn’t stand behind (lesson 5).

Says who, and who benefits. The 50% came from a rival politician’s press release, repeated without a counter-source. One interested party, no document, no check (lessons 2, 3).

What’s missing. No ten-year trend line. No comparison to similar cities. No mention that “crime” here means only murder, while burglary and assault both fell. The omissions all point one way (lesson 7).

The verdict. Every fact in the story is true. The count did rise 50%. And the picture it leaves — a city sliding into chaos under a new mayor — is false. This is Oversold: not a lie, a construction. You didn’t need a fact-checker. You needed six questions and ten minutes.

What this protects you from

Notice what the decode did not require. You didn’t need inside knowledge or a journalism degree. You didn’t need to assume the reporter was lying — most weren’t; the craft of selection, framing, and number-choosing does the distorting whether or not anyone intends it. You needed only to stop treating the claim as finished and start treating it as built, then ask who built it and how.

This is the difference between the two kinds of reader. One takes the headline as the world and passes it on. The other holds it at arm’s length, runs the questions, and arrives at a verdict that is often more true than the story — because it includes what the story left out. Same words on the page. Opposite relationship to them.

The whole, and you inside it

The goal was never to disbelieve everything. A reader who rejects every claim is as lost as one who swallows every claim — both have stopped weighing. Cynicism is just credulity with the sign flipped; it takes no skill and protects you from nothing. The skill is the weighing itself: this one is Sound, trust it; this one is Shaky, hold it loosely; this one is Oversold, set it down. Calibration, not suspicion.

And the last humbling fact is that you are inside the system you’re now reading. You are the audience the attention business is built to hold, the click the algorithm wants, the mind the agenda is set for. You will never fully step outside that. But you can stop being only its product. The news is a manufactured thing pointed at you, and most of the time no one is even pointing it on purpose — it’s just the machinery running. Knowing that won’t tell you what’s true. It tells you to ask, every time, before the claim settles into what you believe: how was this made, and who would I be if I believed it without checking? That question, asked quietly and often, is the whole course. It’s also the first honest step toward reading the news instead of being read by it.