Malaysia’s previously hegemonic Barisan Nasional (BN) government was unexpectedly defeated in the 2018 general election despite a fragmented opposition and widespread three-corner fights that theory states should inhibit turnover. Why? We argue that the opposition-split hypothesis rests on three core assumptions: third parties split only the anti-incumbent vote; coalition/party support is relatively uniform across the country; and opposition parties are not “elite splits” in disguise. The Malaysian context challenges all three of these assumptions. Counterfactual election simulations ultimately suggest that the opposition split neither dramatically helped nor hurt the BN. While this does not upend conventional wisdom on opposition coordination, it does demonstrate that the theory manifests only when its assumptions accord with local realities. More substantively, our analysis also provides insights into why the new opposition will likely seek to increase the salience of ethno-religious issues in a bid to recapture electoral ground.