Networks are all over, from natural (river deltas), man-made (traffic maps and power grids), to biological (blood vessels and plant leaves). Intuitively, we expect these networks to be as efficient as possible: the connections are no bigger than they need to be, and there are no redundancies in the form of loops. However, loops are useful, because they allow to take shorter paths sometimes or make detours if there's a break in the network. I modeled and simulated the effects of fluctuations in two systems whose structure is dictated by its flows, both blood vessel networks and river deltas. In blood vessels, I found that realistic fluctuations are enough to make loops like those that are experimentally observed on the surface of the brain. In river deltas, I found that tidal fluctuations accurately control the loopiness that we see in real-world deltas. Understanding how networks make their loops helps us plan for changing flows, like those that result from strokes (for brain blood vessels) or climate change (for river deltas).