Recent scholarship on far-right (FR) parties has focused on the consequences of government prosecution of FR politicians, with some studies finding that such efforts can be counterproductive and even lead to increased voter support. However, in other cases, such as that of the Greek Golden Dawn (GD) party, it’s clear that harsh government measures can be effective and lead to party collapse. Embedded within this puzzle is our limited knowledge on how targeted FR parties react to such measures. One effective narrative at their disposal is that of “playing the victim” against an “undemocratic” attack from the government. Such rhetorical devices are particularly dangerous, as they can increase hostility towards outgroups and whip-up support for political violence. We argue that victimhood claims also help increase FR party cohesion and voter loyalty and thus may be used strategically by a targeted FR party. This paper focuses on the case of the GD party and tests whether its official communication adopted more victimhood narrative claims following its sanctioning in 2017, while using its Cypriot “sister-party” ELAM—which did not experience a government crackdown—as a comparison case. We gather thousands of official party news articles and combine supervised machine-learning approaches to detect and classify the use of victimhood narratives by both parties and assess how their usage evolves over time. We expect that as a government cracks down on a party there will be an increase in party victimhood claims that will also correlate with a rise in general claims of national victimhood.