Use when concluding an ML experiment - produce a sober conclusion that separates confirmed results from exploratory observations and states the highest verified rung.
