Totally agree Marco, there is so much confusion over this and I had the chance to have a peak to one of the ‘professional’ model of peak projections and they are all… how to say it, ‘tentative’… The problem of peak is a significant one as, since this is a very long disease (permanence of 2–3 weeks in the hospital), when active cases start to fall the dynamics of the infection might have abated well before. Because of the impact on the economy of every additional day of containment, it is important to have early peak indicators. I am still searching, but for now a good one seems to be the ratio between the daily changes in recoveries and the daily changes of death. In China and in Korea, a three days smoothed version of this indicator started growing consistently 14 and 10 days, respectively. The gap between the two countries is also consistent with different testing strategies. I am currently having a look at it, to see if it works for other countries as well. Unfortunately, in none of the other countries this indicator started rising so far.