Possible the adjacent possible over the TeamScience graph

Two things the graph holds, drawn for a reason: each pair fires a named combinability signal (a shared keyphrase across fields, a method with a track record where it has never been used, a contested claim that a method could settle, an old paper still cited that a new problem re-asks, external demand meeting a tractable shape, or the same problem on two fields' lists). The reason, the numbers behind it, an opening question and the grounding papers travel with the pair. Refresh for another (highest scores first, with some chance); lock one side to walk its neighbours. Descended from possible.vision, made to earn each draw. An answer is a hypothesis with a falsification and a cheapest test. Changelog · all pairs · all open problems