The observed readout distribution is a checked matrix-vector product. Exact arithmetic here means exact results for the stated model inputs; measured inputs still carry uncertainty and significant-figure limits.

highlighted = computed this step

Calibration changes the visible probabilities

The assignment matrix maps true probabilities to observed probabilities 29/40 and 11/40.

pobs=(2940,1140)p_{\text{obs}}=\left(\frac{29}{40},\frac{11}{40}\right)
Observed probability productThe table recomputes the matrix-vector product.true probabilitieszero 3/4one 1/4assignmentobserved probszero 29/40one 11/40over 40 shots29 zero reads11 one reads

Observed is not the same as true

The true zero probability is 3/4, while the observed zero probability is 29/40.

342940\frac{3}{4}\neq\frac{29}{40}
Calibration shifts the viewThe mismatch is the calibrated readout model, not drift.true probabilitieszero 3/4one 1/4assignmentobserved probszero 29/40one 11/40over 40 shots29 zero reads11 one reads