Why the course is built this way, what it deliberately avoids, and what the "855" case-file numbers on lecture pages are.

This course studies things that get taken more seriously than the evidence for them warrants — omens, astrology, tarot, numerology, the sense that a recommendation algorithm “gets” you. Teaching that material carries an obvious risk: build the wrong website around it, and the site itself becomes another thing that gets taken more seriously than the evidence warrants. This page says plainly what the course does to avoid that, so the choice is visible rather than assumed.

What the site will not do#

No page on this site claims that a horoscope, a tarot spread, an I Ching casting or a numerology reading has a confirmed predictive mechanism, because none of them do. Where a lecture covers a divination system, it says what the system actually does — draws a symbol, folds a coin, reduces a birthdate to a digit — and separates that from what it is often taken to do. Week 6 says this most directly: controlled studies asking astrologers to match horoscopes to strangers using birth data alone have not found accuracy above chance, and the lecture does not soften that finding to make the topic more entertaining.

The site also avoids the interaction patterns that make an actual horoscope site sticky — no daily push notification, no streak counter, no invented urgency, no result gated behind sharing it first. Every activity on this site can be closed and forgotten with nothing lost.

Every activity discloses its own mechanism#

Ten different interactive components appear across the lecture pages — two or three per week, deliberately never the same pair twice in a row — but every one of them follows a single rule: never simulate a power the course doesn’t have. Where a component produces something that could be mistaken for a reading, a verdict or a working prediction, it ships with a visible “How this works” disclosure in the same place you see the output, not tucked away on a separate page.

• Perception checks ask you to notice something about your own perception (a shape, a pattern, a coincidence) and then show everyone the same explanation, regardless of what you picked. There is no correct answer to score.
• Interpretation choices show you several genuinely different readings of one ambiguous case and let you open any or all of them, in any order — because the lecture’s point is usually that more than one reading survives contact with the evidence.
• The oracle generator, used where a lecture covers a working divination method, always names the exact mechanism (a random pick from a fixed list, most often) next to the reading itself. Build an Oracle asks you to hold your own project to the same rule.
• Probability labs run a small simulation — repeated coin tosses, a shared-birthday count — in the browser, in front of you, so the surprising number is something you watch get generated rather than a claim you have to take on faith.
• Branching scenarios let you make a sequence of small calls inside one case and see where each path actually leads, rather than being told the single “correct” reaction after the fact.
• Compare-and-reveal panels put two or more explanatory lenses on the same event side by side, so the point is the comparison, not picking a winner.
• Case analyses ask you to weigh a real case against a named set of criteria yourself before the discussion underneath shows how the course itself weighs it.
• Archive viewers let you open several real artefacts of the same kind — four different interpretive techniques applied to superficially similar material, say — and see the technique named next to each one.
• Classification games ask you to sort a handful of examples into two named categories, then check each answer against an explanation grounded in that week’s material, not an arbitrary score.
• Reflection tools are a plain, ungraded text box for a prompt that only works if you actually do it — nothing typed into one is saved, sent anywhere, or read by anyone.

None of these components remembers you between visits, competes for your attention, or holds back its own explanation as a reward for finishing it.

The case-file numbering#

Lecture pages carry a small stamp reading “CASE FILE 855-NN” next to the date. It borrows the visual language of an archive — a place where things are filed, dated and open to being reopened — because that is closer to how this course wants a coincidence treated than either “sacred” or “silly.” The number is nothing more than the course’s own numeric suffix and the week; there is no hidden pattern in it to go looking for, which is itself a small, deliberate joke about the whole subject of the course.

Why the visual design looks the way it does#

The site is dressed as an archive — a low dot-grid and hairline coordinate lines behind every page, a serif for headings, a recurring ◇ mark at dividers and section breaks, dates set in a monospace catalogue hand. That is a deliberate choice to look like the kind of place where a pattern gets filed and dated rather than one where it gets revered, and it is worth being honest about why that isn’t a contradiction of everything above.

A course arguing that meaning gets projected onto plain material could easily undercut its own argument by projecting exactly that onto its own pages — dressing the site in the unexplained-symbol, half-lit aesthetic of the very things it studies, and hoping nobody notices the site is doing the thing the lectures describe. The difference this site relies on is disclosure: the oracle generator earns the right to look like a working instrument only because it also shows you, in the same place, exactly what it is. The same rule applies here, at the level of the whole site rather than one component. The ◇ is not explained on first appearance — Week 12 names it and says where it has been hiding — but it is named, on this site, by the people who put it there, rather than left for a visitor to half-notice and wonder about indefinitely. An unexplained pattern that never gets explained is the thing the course is built to be suspicious of; a pattern flagged as deliberate, on the page that put it there, is just a running joke about archives, played by people who also teach why archives feel the way they do.

Where the course still leaves room for the feeling#

None of this is an argument that the feeling of a meaningful coincidence is fake or unworthy of attention — see the homepage and Week 12, which both turn this same scrutiny on the website you are reading right now. The course’s position, stated plainly, is that the feeling is real and worth taking seriously as a feeling, which is a different claim from the pattern itself carrying a hidden cause. Holding both of those at once, rather than collapsing into either “it’s all meaningless” or “it must mean something,” is the actual skill the final project asks you to demonstrate.