Judgment instrument · choice blindness · Johansson et al. 2005

Pick. Justify. Then defend the other option.

Reasons come after the choice, not before. Across 3 rounds you pick, you write why — while the table has already swapped your card: you defend the option you rejected, in your own words quoted back. The unmasking is complete, per round and at the end.

  • Local
  • Full unmasking
  • 3 rounds

This is a demonstration trick: you are shown the rejected option as “your pick”. Nothing leaves your browser. Cost: about 3 minutes. Skip straight to the explanation.

Start — 3 minutes

Play: 3 rounds

Each round you pick between two cards, say how sure you are, and write your reason. Then the cards get shuffled.

Dealing the cards…

Why it happens

The mind does not read its reasons from a ledger — it writes them on the spot, after the fact. When the card on the table is no longer the one you picked, the story gets rewritten with it: same pen, different pick, same “reasons”. It is called confabulation, and it is not lying: it is the sincere filling of a gap.

Proof it is not carelessness: some people catch the swap on the spot. If detection is possible, non-detection is an effect — not fate.

Where to watch out

Ballots with the “right” order, checkouts that remember your “preference”, recommenders that tell you what you want. The defending question: is that my reason — or minutes written afterwards?

Limit next to the claim: the page plays 3 rounds of humour and civic comfort, not your vote or your money. The score is illustrative, not a measure of rationality.

Method, sources, limits

How it plays: card order is drawn at random per round (the ?o= address parameter, one bit per round). You pick, give confidence (50–100%), write your reason. You are shown the rejected option as “your pick”: you confirm it or catch the swap. If you confirm, you defend it once more — then comes the unmasking with quote.

The score: rounds caught (you pressed “I picked the other one”) vs rounds defended without detection (you confirmed + wrote a reason for the rejected option). The domain counter shows humour / comfort / city.

The paradigm: Johansson, Hall, Sikström & Olsson (2005), Failure to detect mismatches between intention and outcome in a simple decision task, Science (doi). The lab is mechanism context; your quotes are your proof.

Limits: 1) 3 rounds do not measure how rational you are. 2) The bank is invented for play (fictional headlines, illustrative priorities). 3) Remembering the bank on replay inflates detection. 4) The lab figure (13% concurrent detection) is a literature group average, not your norm.

Reuse: CC BY 4.0. Cite “Marius Comper — The Swap (freeze 3 Sept 2026)”.