『dylan』

the feedback every interaction designer gets in review: "can it feel more snappy?" or the sibling version, "can it feel more bouncy?" and you nod and go tweak something and bring it back and the pm still says "idk something feels off." this dynamic goes on for two weeks cause the stakeholder doesn't have the language for motion, and you haven't given it to them. here's what's actually happening. a spring has three variables — stiffness, damping, mass — and almost every complaint about "feel" maps cleanly onto one of them. "too slow" is almost always stiffness. "too floaty" or "too bouncy" is damping. "too lethargic" or "too heavy" is mass. when you walk into review with a slider for each and the pm can watch the press animation update from stiffness 180 / damping 16 / mass 1 to stiffness 400 / damping 32 / mass 1 in real time, the conversation stops being about vibes and starts being about the physical property they actually want adjusted. the review closes in ten minutes instead of two more rounds. but stiffness and damping aren't universal. they're a function of what the movement represents. a swipe card retains the momentum and angle it was thrown at, so stiffness 250 / damping 10 reads as physical. the same config on a press-scale or hover-scale feels jarring because a press is mechanical, not fluid — there's no throw, no momentum to preserve. press interactions want stiffness 250 / damping 25. satisfying, no overshoot. mass is the one most people reach for to "make it feel heavier" and then regret, because increasing mass makes things lethargic, not weighty. if you want weight, you want damping. mass only earns its keep on pendulum-like motion where you want a bit of residual wobble after the object comes to rest. edge case: this breaks for destructive actions. you don't want a nice spring there. you want friction. hold-to-confirm, a progress ring that fills over 1.5s, a commit that requires sustained intent — the whole point is that deleting should feel heavy to perform, not pleasant. so conceptual weight isn't only a spring problem. sometimes the right answer is "make the gesture harder," not "tune the spring harder." counter-argument i get a lot: "users don't care about damping, they care about how it feels." correct. which is exactly why you tune stiffness and damping — those are the variables that control how it feels. saying "users don't care about the numbers" to defend not learning the numbers is like saying "users don't care about javascript" to defend not knowing how state updates batch. the user experiences the output. the numbers are how you produce the output consistently, component after component, across a whole system, without re-litigating every animation in every review for the rest of your career. one more thing people get wrong. css linear() is technically a spring, but the values look like linear(0 0%, 0.005927 1%, 0.022466 2%) and the real problem isn't the ugly output — it's that linear() is not fluidly interruptible. if a user grabs the element mid-animation and the target changes, the motion snaps instead of smoothly arcing to the new target. for anything gesture-driven — drags, swipes, magnetic snaps, rubber banding — you want a real physics engine like motion/react so the spring retains velocity through the retarget. css springs are fine for decorative flourishes. they fall over the second a human interrupts the animation. if you can't state a spring as three numbers tied to what the motion represents, you're going to lose every motion review forever.