Comment by vasco

Ok that is adding an interesting variable of perceived interest and then it's not about # of people anymore, but rather time plus # of _interested_ people. I guess that has to be the reasoning because it does make sense if that's the optimization goal.

Though I'd imagine mostly its going to be a random sample of people that happen to be there that day. I imagine there's likely under 1% of museum visitors actually chasing single works and planning trips like that. So most people that would see it are still just random "museum people" (which under the "interest" metric is still better than purely random people).