Some theories are alleged to be implausible because they are committed to systematic ‘overdetermination’. In response, some authors defend ‘compatibilism’: the view that the putative overdetermination is benign, like other unproblematic cases of a single effect having many sufficient causes. The literature has tended to focus on the following question: which relations between sufficient causes of a single effect ensure that problematic overdetermination is avoided? This paper argues that several widely endorsed answers to this question are subject to counterexample. It (...) then proposes a diagnosis of this failure: the standard answers neglect what really matters––how the causes are connected to their shared effect. In particular, overdetermination is avoided when there are no independent causal connections. (shrink)