What this measures, and what it refuses to do.
The instrument is built on one published result, one well-documented failure mode, and a flat refusal to forecast. All three are on this page, with the citations.
22,500 decisions, decided by a coin.
Between 2013 and 2015 Steven Levitt ran a field experiment at the University of Chicago on people who were genuinely stuck. Participants brought a real, unresolved decision — quit the job, end the relationship, move city, start the business — and agreed in advance to let a coin toss break the tie. Roughly 22,500 coins were flipped. The subjects were surveyed at two months and again at six.
Those whose coin came up for change were about 11 percentage points more likely to have actually made the change, and at the six-month follow-up they reported significantly higher wellbeing than those whose coin said stay. The effect was strongest on the decisions people described as the hardest.
The coin has no information. What it removed was the option of continuing to deliberate — and the measured cost of that option turned out to be higher than the participants had priced it.
Two honest caveats, because the study has them. Participation was self-selected, so this is evidence about people already stuck enough to hand a decision to a coin. And compliance was voluntary: the coin persuaded rather than compelled, which is exactly the mechanism this product uses.
Why the position has to come from outside you.
A stalled decision is not a shortage of information. It is a loop: you run the same two options through the same reasoning, get the same tie, and read the tie as evidence that you need more input. The loop is stable because it feels like diligence. Meanwhile the open goal keeps consuming attention — Masicampo and Baumeister showed that unfulfilled intentions intrude on unrelated thinking until a concrete plan exists, not until the goal is achieved.
An externally-generated position breaks the loop for one reason: you did not author it. You cannot re-derive it, so you have to do something you were not doing before — argue against it. The moment you argue against it you find out, quickly, which option you actually wanted. Kahneman and Lovallo call the general move the outside view; the coin toss is its crudest possible implementation.
ELDR replaces the coin with fourteen answers. Unlike the coin, it then has to show its work: the position, the confidence number, the window, the case against itself, and a log of what ran. A coin cannot be wrong in a way you can inspect. This can.
The counter-case is not a disclaimer. It is a premortem in Gary Klein’s sense, run against our own output before you see it, and it has to name the specific thing that would overturn the position. If a read cannot name that thing, the confidence number is wrong and the read says so.
What each of the 14 is actually for.
Eleven are required and three are optional. Nothing here is a personality quiz: four inputs describe the decision, three calibrate how you decide, two price the stakes, two bound the clock, and three are yours to skip.
| # | Input | What it measures |
|---|---|---|
| 01 | State the decision | The sentence the read has to quote back. If it cannot be said in one line, the decision is two decisions. |
| 02 | Name both options | Forces the alternative to be named. "Do nothing" is a position with a price, not the absence of one. |
| 03 | How long has it been open? | Duration is the strongest single predictor in the set that more information is no longer arriving. |
| 04 | What do you already think the answer is? | The answer you already hold. Weighted, never obeyed — a stated lean plus avoidance is the loudest signal here. |
| # | Input | What it measures |
|---|---|---|
| 05 | How do you usually decide? | Axis 1 of the profile code. Evidence-led and gut-led fail differently, so they are advised differently. |
| 06 | Under pressure, what do you do? | Axis 2. Strike and hold are both defensible; what matters is which one you default to when the clock is real. |
| 07 | Who do you run it past? | Axis 3. A council raises input quality and raises delay. Both are priced into the window. |
| # | Input | What it measures |
|---|---|---|
| 08 | If you get this wrong, what does it cost? | The downside you would actually feel. It sets how hard the counter-case pushes. |
| 09 | And what does waiting cost? | The price of the position you are holding right now. This is the number most people have never named. |
| # | Input | What it measures |
|---|---|---|
| 10 | Is there a hard deadline? | The hard bound on the window. Only externally enforced dates count. |
| 11 | If you decided tonight, how fast could you move? | The lower bound on the window. A position you cannot execute inside is advice, not an instrument. |
| # | Input | What it measures |
|---|---|---|
| 12 | Date of birth | Optional, and reduced to an age band before it reaches the engine. Nothing is cast from it. Skipping it changes nothing. |
| 13 | Anything the questions missed | Free text. The constraint the fixed questions could not ask about — and the phrase most often quoted back. |
| 14 | Where should the read be filed? | Optional. Somewhere to file the read. No account is created by answering it. |
Three binary axes. Eight files. No fourth axis.
Inputs five, six and seven produce a two-letter code and a digit: evidence-led or gut-led, strikes or holds under pressure, decides alone or by council. That is the whole taxonomy. It describes how you have answered, not who you are, and it changes when your answers change.
| Code | File | Reads as |
|---|---|---|
| ES-1 | OPERATOR | evidence-led · strikes under pressure · decides alone |
| ES-2 | STAFF OFFICER | evidence-led · strikes under pressure · decides by council |
| EH-1 | ANALYST | evidence-led · holds under pressure · decides alone |
| EH-2 | COMMITTEE | evidence-led · holds under pressure · decides by council |
| GS-1 | RAIDER | gut-led · strikes under pressure · decides alone |
| GS-2 | FRONTMAN | gut-led · strikes under pressure · decides by council |
| GH-1 | HOLDOUT | gut-led · holds under pressure · decides alone |
| GH-2 | DELIBERATOR | gut-led · holds under pressure · decides by council |
// two further numbers come off the same answers: conviction, 0-100, and decision drag in days.
// both are descriptions of your inputs. neither is a prediction.
Forecasts issued: 0
The category next door sells certainty about the future for $45 a month. This is the list of things this product will not do, and it is not a marketing position — it is enforced inside the prompt that generates every read.
| Will not | Why | |
|---|---|---|
| Predict what will happen | No instrument can. Forecasts issued by this product: zero, permanently. | |
| Read a birth chart | Date of birth is optional and is reduced to an age band. Nothing is derived from the date itself. | |
| Agree with you | Every read carries the strongest case against its own position, and names what would overturn it. | |
| Hide the price | Both doors are priced on the landing page, before the intake. No trial, free or for a dollar. | |
| Replace a professional | Not financial, legal, medical or mental-health advice, and it never pretends otherwise. | |
| Keep you here | The product is finished when the window closes and the outcome is logged. That is the whole loop. |
Nine carry meaning. The rest are texture.
| tiwaz | the decision | the wordmark |
| ansuz | the message | heads a read, and the log that built it |
| raidho | move | a position of MOVE, and every call to act |
| isa | hold, ice | a position of HOLD |
| jera | the cycle | windows, deadlines, anything on a clock |
| thurisaz | the thorn | the case against your own read |
| sowilo | the sun | confidence, and anything measured |
| fehu | what it costs | price, and the terms attached to it |
| perthro | the lot | the coin. Marked amber: cited, never used |
// the strips at the head and foot print the full elder futhark, all 24.
// nine carry meaning. the rest are texture, and are marked as such.
// perthro is the lot. it is cited on this page and used nowhere in the product.
Everything above, sourced.
Levitt, S. D. (2021). Heads or Tails: The Impact of a Coin Toss on Major Life Decisions and Subsequent Happiness. The Review of Economic Studies, 88(1), 56–87. Earlier as NBER Working Paper 22487 (2016).
The 22,500-subject field experiment. Source of the +11pp action effect and the six-month wellbeing result.
Klein, G. (2007). Performing a Project Premortem. Harvard Business Review, 85(9), 18–19.
The counter-case block is a premortem run against our own position, before it is shown to you.
Kahneman, D., & Lovallo, D. (1993). Timid Choices and Bold Forecasts: A Cognitive Perspective on Risk Taking. Management Science, 39(1), 17–31.
Inside view versus outside view. The reason an external position is worth anything at all.
Masicampo, E. J., & Baumeister, R. F. (2011). Consider It Done! Plan Making Can Eliminate the Cognitive Effects of Unfulfilled Goals. Journal of Personality and Social Psychology, 101(4), 667–683.
Open loops consume attention until a plan exists. This is the cost the intake asks you to name.
Gilbert, D. T., & Wilson, T. D. (2007). Prospection: Experiencing the Future. Science, 317(5843), 1351–1354.
People are poor at predicting how they will feel about outcomes. One of the reasons this product does not try.
Tetlock, P. E. (2005). Expert Political Judgment: How Good Is It? How Can We Know? Princeton University Press.
The evidence base for treating confident long-range forecasting as a product defect rather than a feature.