While modelling is obviously something that is looked at, the observation of the actual data is I think the best guide. The good news is that the number of deaths has fallen very sharply and continues to fall sharply, and the number of hospitalisations has fallen sharply—not quite as sharply as deaths, but nevertheless sharply. Critically, the link has broken, so the weight that we place on the number of cases as an indicator is not nearly as great as it was before, because it used to be inevitable that that led to hospitalisations and deaths.
The reason for the timings in the road map is so that, after each step, we can see its impact before being able, carefully, to take the next step. That is the core reason for the timings in the road map—four weeks to see the impact of a step, and then one week to give warning of the next step. It is that, rather than the modelling, that I put the stress on for the timings in the road map.