The pit wall
The room every lab and the race run in, and why it is the same for everyone
The Faculty’s pit wall is one room with forty seats, and every Lab runs in it. So does the race.
It is the same room in the same state for everyone: the same screens in the same layout, the same seat, the same setup on the machine in front of it. Nothing about the environment is a dial a student can turn.
Why it is held constant#
This course marks calls rather than results, and two calls compare only if they were made from the same place. The race already holds luck constant — the cohort’s stochastic events are drawn once, before anybody races — and the room holds the second thing steady. What is left to vary between two students is what they put on the screens.
A student who found a number faster because their monitor was wider would have demonstrated the monitor rather than a call, and calls are what the marking is about. It is also why week 11 can measure a display at all: between two measurements, the display is the only thing that differs.
Where the fixed part stops#
The interface between a model and a wall is fixed and available from the start of teaching. It accepts a race state. It returns legal candidate strategies with distributions over finishing positions and the derived situation quantities named in week 11, followed by the recommendation, its confidence and the stated objective. The instrument publishes the same output schema week 11’s harness used, so that lab’s wall configuration runs against the student’s final model unchanged. Same screens, same schema, different wall.
The race log preserves each call, its timestamp and the recommendation standing at that timestamp. A departure also preserves the predeclared actionable-limit citation and the evidence available at the call showing that its clear, observable condition applied. General limitations remain part of an honest domain even when they cannot support a live departure. Missing either departure record can cost race conduct marks; an honest analysis may still earn defence marks, but cannot revise the log or justify the departure after the fact.
Two things about the room are stated elsewhere, because they are rules rather than explanations. The room is shared equipment, and what that means for leaving a machine as you found it is in the code of conduct. An adjustment arranged through the University’s process is a question about access to the room rather than a variation in what is assessed; policies draws that distinction.
Both are described rather than linked: this is a prototype of the course site, and neither the room nor its environment is part of it.