Left: toroidal field component squared amplitude as a function of the Rossby number, Ro_f, probed near the surface. At low Ro_f, the agreement between the simulations (purple line) and observational (dashed black line) fits is qualitatively good, with both simulations and observations showing a decreasing trend, although the field amplitude is too low near the surface (left panel). This is likely due to our choice of top magnetic boundary conditions. Right: same quantity probed in the middle of the convection zone, where the agreement is quantitatively good, with both fits being close in terms of the power law index. A possible inverse trend is indicated by the simulations at higher Rossby numbers, Ro_f > 1 (not covered by the observational database), and hence suggests a minimum in the toroidal field strength near Ro_f ∼ 1, which could have interesting consequences for stellar spin-down via wind braking.