Skip to main content

Week 1: What 99% Promises

The one analytic move this course runs for twelve weeks — claim, source event, verdict — and why a local simulator is an instrument rather than a convenience.

23 February 2027

Open the slides

Slides

This course has four lectures. This is the first, and it is the only one that introduces something you will use every single week.

What this lecture carries

The move. Three terms, defined once and used for twelve weeks:

Why the simulator is not a shortcut. Every transfer in this course is local and fake. That is not because real networks are inconvenient; it is because a claim about a failure you cannot reproduce is not a claim anybody can check. The simulator is what makes the rest of the semester falsifiable.

The course’s one rule, stated in week 1 so it can be held against you later: nothing on your screen is allowed to assert more than your evidence supports. This applies to your interface, and it applies to the prose you write about your interface.

What it sets up

Where it landsWhat it carries there
Week 1 studio, two days laterruns the move on someone else’s interface
Week 2“complete” is a claim; a state is a candidate source event
Week 3a denominator is a source event you can count
Week 10the same move, turned on your own code, with one verdict taken away

What it assumes

Nothing. If you can write a function that takes a callback, you have the prerequisites.

A worked example

Take one claim from the case this lecture opens with — a bar that reaches 99% in nine seconds — and run it through the three terms rather than the table.

The claim is not “99%”. It is the rate: a bar that covers 99 points in nine seconds tells a reader the last point takes about a tenth of a second. Nobody wrote that sentence; the movement asserts it.

The source event would have to be something that reports how long the remaining work takes. No such event exists — the process reports bytes, and the last point is not bytes.

So the verdict is unsourceable, and notice what that does not mean. It does not mean the designer lied. It means the screen carried a claim the system had no way of supporting, which is a different and much more common thing.

What this lecture cannot settle

This move tells you what a claim rests on. It does not tell you whether the interface is any good.

An unsourceable claim can be the right design decision — taken knowingly, for a reason, with a cost. Week 10 gives that case its own verdict. Until then, resist converting an analysis into a judgement: you are building the instrument, not using it to condemn anything.

If you miss the hour

The slides carry the argument in the order it was made, and this week’s reading is listed on the studio page. The studio on Thursday assumes the vocabulary above rather than re-teaching it, so read the deck before you arrive rather than after.

Teaching team