This paper compares two questionnaire surveys conducted in the Karoo to investigate the claim that in human-wildlife conflicts farmers systematically inflate predation reports to score political points. Although predation rates and updated predation values for the Karoo are presented, the main contribution is not “a” number but rather an analysis of what affects the magnitude of predation self-reports. The two surveys produced quite different figures, which were due to methodological choices rather than anything farmers said. With the methods standardized, the figures converged to within 3% of each other, which either means that farmers never lied at the height of the Karoo’s gin trap wars or that they are still lying about their losses despite the trust we think we have in Koup. In this work, an important decision is whether to assign all, some or no perinatal losses to predation. In the Karoo, it is concluded that counting none towards predation is most prudent as lambs are born on the open range and few producers’ document ewe conception with ultrasound scanning. Determining the lamb inventory at tagging for the first time inflates predation rates because it divided by a smaller number while dividing lamb losses by an overall inventory (as we, unfortunately, must in this country due to data deficiencies) reduces losses because it divided by a larger number. Standardised methods are therefore essential and begs the question of how we explanation spatial and temporal variations in these standardized data. Some preliminary modelling will be presented.