Ditto

Ditto grew up beside a fast river with her family of otters, who fished the same three bends every day. Young Ditto once caught a fish at the shallow bend and decided, proudly, that the shallow bend was the best bend. Her grandfather just smiled. "One fish, one day," he said. "Try it tomorrow. And the day after. Then tell me which bend is best."

So she did. And it turned out the shallow bend had been luck — a fish that happened to be passing. The deep bend caught more, over and over, day after day. Ditto learned the thing that would become her whole craft: a single result can fool you completely. You only find out what's really true by doing it again, the same way, and seeing if it holds.

When she was grown, Ditto came to the CuriosityQuest academy. Lumen, the head of the academy, asked her, "What is replication?"

Ditto rolled her marble down the ramp, then scooped it up and rolled it again — same ramp, same spot, same push. "It is doing the test again the same way, to check the result holds," she said. "Once could be luck. A weird fluke. A thing that only happened because something else you didn't notice was going on. But twice, and three times, the same way? Now it starts to be true. Replication is how you tell a fluke from a finding." Lumen nodded. "You are appointed," she said.

In her classroom, Ditto has students drop a ball and time its bounce — once. "Write it down," she says. "Now, is that the real answer?" The students say yes. Then she has them do it nine more times, exactly the same way, and the numbers wobble around — and suddenly they see it: no single drop was the whole truth. The answer lives in the repeats.

She teaches her students a few habits about repeating: Do it again, the same way. Replication only works if you keep everything the same and just repeat. Change something, and you're doing a different test. Once is a maybe. A single result is a hint, not an answer. Be excited, then check it. Watch the wobble. When you repeat, the results won't be identical — they'll wobble a little. That wobble tells you how much to trust any one try. A fluke doesn't repeat. If a result was luck, doing it again usually shows it. If it keeps happening, it's probably real. * Repeating isn't doubting yourself. It's how everyone checks — the best scientists in the world repeat their tests. Nobody's first try is automatically the truth.

Ditto tells her students, "My very first result is wrong all the time. That is not a failure. That is just what a single try is like. So I do it again, and again, and let the repeats tell me what's real."