Thermal barrier coatings (TBCs) could improve the efficiency of internal combustion engines by reducing heat losses. TBC modeling has demonstrated efficiency gains at all loads in compression ignition, including in gasoline compression ignition (GCI). However, after GCI testing of 13 coated pistons of various materials, thicknesses, and surface preparations in this work, a clear trend was observed: TBCs result in an efficiency gain at 6 bar IMEPn (low load) and an efficiency penalty at 15 bar IMEPn (high load). This can be explained by “convective vive”, which was originally presented by Woschni for engines with TBCs. Convection vive refers to a reduction in the quench distance of a reacting flame when the quenching surface temperature is elevated, resulting in an increased heat transfer coefficient and heat transfer rate despite the lower temperature difference between the gas and the wall. Although this work is unable to provide direct experimental evidence for convection vive, several experiments were performed to rule out alternate theories including differences in heat release rate, radiation heat transfer, combustion chamber deposits, and surface catalytic effects. Experiments also showed that at 6 bar IMEPn, injection pressure did not impact the performance of the TBC compared to the metal baseline, but at 15 bar IMEPn, the efficiency penalty increased with rail pressure. These results provide indirect evidence of convection vive: as the surface temperature of the TBC increases with load past a threshold, the quench distance decreases to the point where exothermic reactions can occur in the boundary layer, which raises the convective heat transfer coefficient. This implies that in mixing controlled combustion systems, the TBC should not be present on combustion chamber surfaces that experience high fuel-wall interaction during combustion.