One guess a market, changeable until it closes, scored on how confident you were.
The mechanics matter more than they sound, because a badly scored prediction game rewards hedging and a well scored one rewards being right and saying so.
How a market runs.
A market opens
With four options, or two, and a stated settlement source. Never more than four, because five is a shrug.
You pick one
Free, instant, and one per account. The percentages update as everybody else picks.
You can change it
Any number of times until the market closes. The last pick before close is the one that counts.
The market closes
Before the outcome becomes obvious. The combining market closes on 1 July, not on the morning of harvest.
It settles on the published source
The weighbridge ticket, the lambing record, the council register. Never on our opinion.
Points are awarded
More for being right when few people were, less for being right with the crowd. Nothing is deducted.
How the scoring actually works.
| If you picked | And it was | You get |
|---|---|---|
| An option that 15 per cent of people picked | Correct | 100 points |
| An option that 40 per cent of people picked | Correct | 55 points |
| An option that 70 per cent of people picked | Correct | 30 points |
| Any option | Wrong | 0 points. Nothing is ever deducted |
| An option in a market that is voided | Voided | Nothing, for anybody. The market disappears |
| Nothing at all | Anything | Nothing. There is no penalty for sitting one out |
Why the percentages are shown before you pick.
Because hiding them would make this a quiz rather than a market. The whole interest of a prediction market is the aggregate: three thousand people, most of whom know something, arriving at a number that is usually better than any individual estimate.
It does cause herding — people see 38 per cent and go with it — which is exactly why the scoring pays more for a correct unpopular answer. The two effects roughly cancel, and what is left is a number that has genuinely been useful twice.
In 2025 the crowd put 31 per cent on the durum coming in under half its expectation, at a point when the agronomist was still forecasting a normal crop. The crowd was right and it was right early.
What a market will never be about.
Anything to do with a person’s health. Anything to do with a legal proceeding involving an individual. Anything about whether a named person will still be doing a job. Anything about an animal dying.
Those are the four categories that turn a farm prediction game into something unpleasant, and there is no version of the scoring that fixes it. The rule is a list rather than a principle because a principle can be argued with at eleven at night by somebody who really wants to open a market.
The mechanics, questioned.
You can see the percentages, always. You cannot see who picked what, ever.
It is voided and nobody scores. That has happened once, when a planning application was withdrawn and resubmitted with different red lines.
The season leaderboard resets each year. The all-time total does not, because people are attached to it.
Being top of a list. That is the entire prize and it has to stay that way, for reasons the law page sets out at length.
The markets.
Six of them, open now, free to enter.
Elsewhere on the estate.
/estate/the-money/
/estate/recording/
/estate/sitemap/
