Two kinds of kinetic energy from one fall
Drop an ordinary ball and every joule of its gravitational potential energy becomes translational kinetic energy, 1/2 m v^2. A yo-yo unwinding on its string does something different: as it falls a height h, the released energy mgh has to supply both the translational kinetic energy of its center of mass falling, 1/2 m v^2, and the rotational kinetic energy of its spin about its own axle, 1/2 I w^2. That split, and only that split, is why a yo-yo descends noticeably slower than free fall even though gravity is the only external force doing net work on it.
Rolling without slipping on the string: v = wr
The string does not unwind from the yo-yo's outer rim — it unwinds from the narrow axle at radius r, which is much smaller than the disk's outer radius R. Because the string does not slip against the axle, the linear speed of the falling center of mass and the angular speed of the spin are locked together by v = w r, exactly analogous to a wheel rolling without slipping. This constraint is what couples the two energy terms into a single, solvable equation of motion instead of two independent ones.
Moment of inertia sets the acceleration
Modelling the yo-yo as a uniform disk gives I = 1/2 m R^2 about its own axis; a real yo-yo, with more mass concentrated near the rim, behaves like I = k m R^2 for some shape factor k somewhat above 1/2. Applying Newton's second law separately to the translational and rotational motion and eliminating the string tension gives a downward acceleration far smaller than g whenever the axle radius r is small compared with R, since a small r makes the moment-of-inertia term dominate the denominator.
translation: mg - T = m a rotation: T * r = I * alpha, with a = alpha * r solving: a = m g r^2 / (I + m r^2) = g / (1 + I / (m r^2)) small axle radius r -> I/(m r^2) large -> a much less than g
The 'sleeping' bottom and the snap-up
When the string finally runs out, the yo-yo is spinning fast, but the friction between string and axle is far too weak to stop that spin instantly — instantly halting the rotation would require a huge string tension the friction simply cannot supply. So the string goes slack and the yo-yo keeps spinning freely on the axle, or "sleeps." A small upward tug on the string re-engages the axle's grip (the deliberate "bind" technique used with modern unresponsive yo-yos), catching the string again and letting the stored angular momentum pull the yo-yo back up the string — though friction losses mean it never quite reaches its starting height, and every up-down cycle loses a little more energy.
Why a thinner axle makes a 'better' yo-yo
A narrower axle radius r increases the ratio I / (m r^2) in the acceleration formula above, pushing more of the released potential energy into spin rather than downward speed, and stretching out the sleep time at the bottom. Modern competitive yo-yos push this further with ball-bearing axles, which cut friction dramatically and let the yo-yo sleep for tens of seconds — at the cost of needing that deliberate bind to return it to the hand, since the near-frictionless axle no longer grips the string on its own the way a classic fixed axle did.
Frequently asked questions
Why does a yo-yo fall slower than a dropped ball?
Because gravitational potential energy has to supply both translational kinetic energy AND rotational kinetic energy as the string unwinds, whereas a dropped ball converts all of its potential energy into translational motion alone. With a thin axle, most of the released energy goes into spin rather than downward speed, so the descent acceleration is a fraction of g rather than g itself.
What makes a yo-yo 'sleep' at the bottom of the string?
Once the string is fully unwound, the string tension needed to stop the yo-yo's spin instantly is much larger than the friction at the axle, so the string goes slack and the yo-yo keeps spinning freely on its axle instead of climbing back up. A small upward jerk on the string re-engages the axle's grip on the string (a 'bind' on unresponsive yo-yos), catching it and converting the stored angular momentum back into an upward pull.
Why do competitive yo-yos use a thin or ball-bearing axle?
A thin axle radius increases the ratio of moment of inertia to axle radius squared in the acceleration formula, which shifts more of the falling energy into spin and away from linear descent speed, letting the yo-yo sleep (spin freely at the bottom) for much longer. A ball-bearing axle reduces friction further still, extending sleep time even more, at the cost of needing a deliberate 'bind' technique to return the yo-yo to hand.
Try it live
Everything above runs in your browser — open Yo-Yo Physics and change the parameters while it is running. Nothing is installed, nothing is uploaded, the whole model lives in one tab.
▶ Open Yo-Yo Physics simulation