The public record

Every figure on this page is read live from the same database the product runs on. Nothing is held back for looking bad.

What was read

candidates assessed since launch

cleared

disqualified

Roughly one candidate in clears.

Today

read

cleared

disqualified

Where the candidates went

The three most recent disqualifications, whichever they happen to be.

Today, by reason

What has happened to the cleared signals

cleared so far, on two axes that move independently.

Where they sit in time

upcoming

live

closed

What has happened to the thesis

unresolved

confirming

catalyst fired

invalidated

lapsed

signals have not reached their window yet. The median opens in days.

A signal read before its window opens is read as early as it can be. The development is already public. The period its catalyst is expected in has not started.

Every one of them carries a stated catalyst and a stated way to be proven wrong, and both were recorded on the day it was scored.

How long a signal gets

Timeframe

Window

Grace

Total

days

Windows are sized to the implication timeframe, with a grace period on top.

A signal whose only test is spent lapses early. A signal given a replacement catalyst is re-sized from its discovery date, which caps how far it can ever run.

signals have run out of time without a verdict. That is the entire lapsed record, and at that sample size no rate is worth computing, so none is given.

Where five months comes from

Five months, on the landing page, is the median near-term window: days. Medium-term theses run about months.

What gets read

Signals cleared per category since launch.

A signal can carry tags belonging to more than one category, so these counts add up to more than the number of signals. sit in two or more, and in none.

The distribution is a readout of where research activity is concentrated. When a field starts producing more that clears the bar, its count rises here.

How the scores fall

cleared signals, by total score.

9

8

7

6

5

4

3

2

1

0

Nothing has ever scored 9 out of 9. signals have scored . The mean is .

Every one of these is published, at every score. The signals that scored carry the same reasoning, the same sources and the same verified market exposure as the that scored .

When things enter

Nothing enters a run more than hours after it was published.

Across the cleared signals that carry both a publication and a discovery timestamp: violations, longest gap , median .

Everything scored so far is there the moment you sign in. Every run since launch, every signal at every score, with its rejection log.

Start 14-day free trial

No credit card required. Cancel anytime.

How the reading is done →
PUMPKN.AI logo
Follow𝕏
FAQDisclaimerTermsPrivacyCookieRefundLegal