Skill is decided not by 'exactly' but by 'order of magnitude first'

A good price reader's weapon isn't 'always landing near the answer' — that's nearly impossible. The real weapon is calibration: a state where your felt confidence and your actual hit rate line up. Political scientist Philip Tetlock tracked tens of thousands of expert forecasts over two decades (Expert Political Judgment) and found that flashy, certainty-spouting experts were actually less accurate than those who hedged and spoke in ranges. The trait his later 'superforecasters' (Good Judgment Project) shared wasn't more knowledge — it was the habit of speaking in ranges, not points, and in probabilities, not verdicts. Price estimation is the same. '$370K exactly' is almost always a worse answer than '$300K–$500K, most likely $400K.'

Why magnitude comes first — 1 trillion and 10 trillion aren't 'both big companies'

The bigger a number gets, the duller our read on the same difference, so cramming two assets of different orders of magnitude into one mental slot is a routine error. Loading the data made this vivid. NVIDIA's market cap sits around $2.9 trillion; Hyundai Motor's around $52 billion. Both live under the single phrase 'famous listed company,' yet the gap is roughly 56x — an entire order of magnitude apart. The same trap is wired into asset prices:

• F-35 fighter jet ~$80M vs Boeing 737 MAX ~$125M — the narrative 'a fighter jet is obviously pricier' overwrites the magnitude. The airliner actually costs about 1.5x more.
• Gold 1kg ~$99,000 vs one Bitcoin ~$87,200 — same order of magnitude, only ~13% apart. A point guess here is nearly a coin flip, but a range answer safely catches both inside the $80K–$100K band.
• Mona Lisa insurance estimate ~$870M vs one Mars mission ~$2.7B — both feel like 'priceless, astronomical' sums, yet one is about 3x the other.

Get the magnitude right and your score holds. Get it wrong — pricing the iPhone 16 Pro's component cost (about $568) at $5,000 — and there's no recovery. Precision is a luxury; the order of magnitude is the lifeline.

Stereotypes wreck your calibration

The root of overconfidence Tetlock flagged is 'the plausible story that pops to mind first' — the availability heuristic. Lump salaries together as 'all about the same' and a Korean corporate entry-level wage (~$36K) and a US software engineer (~$180K) end up in one 'wage earner' slot, evaporating a roughly 5x gap. Calibration is the work of dragging that narrative back to data: swapping 'both are expensive' for the habit of nailing it as a ratio — 'this one is about 1.5x the other.'

✍️ Operator's note — The person who got the jet-vs-airliner question squarely wrong while testing this mode? Me — dragged along by the 'weapons are expensive' image. That's when I learned I almost always overcommit when I guess a single point. So now I jot a one-line range in my head before guessing; even a loose 'this is in the tens-of-millions range' sharply cuts the urge to insist afterward, 'oh, I totally had that.'

How to build calibration with this game

This mode scores your guess next to a precomputed estimate on the same asset, so how many times off you were stays right there on the chart. To use it well, set a range before you guess instead of a single point — something like '850–1,100, most likely 950.' Then, separately from pass/fail, log whether the answer fell inside your range. If you're often outside it, you're overconfident, so widen it; if you always land comfortably, tighten it. Note which categories (stocks, crypto, commodities) keep tripping you, and within a few rounds you'll have a map of your own weak spots. To be clear again: none of the figures here are investment advice, and the results are not user statistics — just a record of pure price-sense practice.