A dart is a projectile, and projectiles are only predictable when the heavy end leads. Put the mass toward the back and the tail overtakes the nose in flight — it tumbles, and it lands on whatever face happens to be pointing forward.

So the mass sits ahead of centre, close to the cup. The tail follows rather than fights, the dart holds its attitude all the way to the wall, and it arrives cup-first. That is the whole reason a beginner can stick one on their first throw.

The first bodies were plain cylinders. They worked, roughly, and they were miserable to throw — a tube has nothing to stop it rolling, so every throw left the hand slightly differently and arrived slightly differently.

Small wings fixed it. They cut drag along the body and give the air something to push against, so the dart settles into its line instead of hunting for one. Same throw, same result — which is what makes the game worth playing.

Everything else is delivery. The cup is the part that has to actually hold — and holding is a sealing problem, not a stickiness one. Too firm and it will not deform enough to close against the surface. Too soft and it collapses, or tears.

It lands at 40mm across, concave, in a silicone with a durometer under 40 — soft enough to flatten and seal on contact, firm enough to spring back and do it again thousands of times. The silicone is recyclable, which matters more the more of these we make.

None of this came from simulation. It came from printing a dart, walking outside, throwing it at things, and printing a slightly different one — twelve-plus prototypes, each one answering a complaint about the one before.

That is a slow way to design a product and an extremely good way to find out what a real throw does to it. Every version that failed failed in someone's hand first, which is the only test that counts.