The incentives in science don’t always encourage openness—but being wrong might just be the key to getting it right.