The three year principle

By Gordon Rugg

If a problem’s still a problem three years after its discovery, then the answer probably isn’t in the obvious place.

When you’re a new researcher tacklling a hot, newly-identified problem, there’s a strong temptation either to follow a hunch (a really bad idea, since the odds are heavily stacked against you) or to do much the same as everyone else is doing, so you can hide in the crowd. If the research community is making steady progress on a problem, then hiding in the crowd is a pretty safe strategy for new researchers, while they learn their skills. At worst, they’ll end up with some dull results that they’ll manage to publish somewhere; at best, they might happen to get lucky, and find something interesting.

However, if the research community still isn’t making progress on a problem after three years, then it’s worth considering a sideways move to a different problem, or a change in how you’re tackling the problem.

Phrased that way, it looks self-evident, which it is. But why three years, rather than two, or four? Here’s the reason.

Continue reading