English: "1,000 different planets were generated randomly. a Results when each planet was run one time: those planets which remained habitable are shown as green circles and those which did not as black circles; b results when the same 1000 planets were run a second time (same planets but different starting temperatures and perturbations compared to the first run). In this particular example, 15 planets stayed habitable on the first occasion, 10 on the second, with 6 remaining habitable on both occasions (additional analysis shows that on average there is a 39% overlap, or in other words there is a 39% chance of a planet being habitable in the repeat set if it was habitable in the original set)."