Why does walking to the center of a merry-go-round feel so hard? Coriolis stabilization and the metabolic cost of staying on track

arXiv:2606.27400v1 Announce Type: new Abstract: A standard undergraduate problem has a student walk radially inward on a rotating, frictionless merry-go-round. The textbook analysis -- angular momentum is conserved, so the kinetic energy rises and the student does work -- is exactly correct for a point mass. Yet anyone who has tried it knows the effort is dominated by something the point-mass model never mentions: the muscular work of \emph{not} being thrown sideways. We make that effort quantitative with a deliberately minimal model. Treating the student as an active controller that cancels the Coriolis force at a metabolic cost $P \propto F_C^{\,n}$, we show that the cost scales as $r^{-2n}$ as the axis is approached. The widely used quadratic cost ($n=2$) gives a steep $r^{-4}$ rise; a linear cost ($n=1$) gives $r^{-2}$. We argue that this \emph{sensitivity of the prediction to the cost model} is the most useful thing in the problem: it forces students to see how a modeling assumption, not just an algebraic step, drives a physical conclusion. We give an order-of-magnitude estimate (explicitly flagged as such), a one-line entropy-production / efficiency argument that connects the exercise to non-equilibrium thermodynamics, a feedback (PD-controller) reformulation that reproduces the same scaling, and a back-of-the-envelope experiment students can do on a playground with a phone and a heart-rate strap. The material is aimed at an intermediate-mechanics or biophysics elective and is designed to teach model validity, assumption sensitivity, and the idea that staying alive and on-course has a thermodynamic price.