Abstract The El Niño‐Southern Oscillation has two warm phase flavors of Eastern Pacific (EP) and Central Pacific (CP) El Niños that exhibit seemingly distinct global teleconnections, but the limited observational sample leaves open whether and where these differences are robust. Coupled climate models are conventionally used to more robustly sample these teleconnections and their differences, yet they have persistent tropical sea surface temperature (SST) biases. To better understand teleconnection differences between EP and CP El Niño, we use atmosphere‐only ensemble simulations with both observed and realistic synthetic SSTs generated from a Linear Inverse Model. With a sufficient number of El Niño events, the perceived teleconnection differences weaken due to broader sampling of event patterns, event strengths, and intrinsic atmospheric variability. However, biases in the atmospheric model’s teleconnection caveat these conclusions. With CP event frequency projected to increase in the future, associated global impacts have the potential to become stronger, too.