These data are from Laboratory Earthquake Experiments from the Cornell 0.76 m apparatus in support of the following research: Natural faults likely include both Velocity-Weakening (VW) and Velocity-Strengthening (VS) sections. We developed a laboratory method to replicate this frictional heterogeneity using a 760 mm long Polymethyl methacrylate (PMMA) block with eleven VW patches (bare PMMA) separated by VS barriers (Teflon tape). We compared the behavior of this Multiple Patches (MP) arrangement to those from One single VW Patch (OP) with the same total VW fault area. Seismic events that occurred in clusters with foreshocks and aftershocks were observed only in the MP tests, and total slip, maximum slip rate, seismic moment, and recurrence time of the largest event (termed the mainshock) in a slip cycle, were an order of magnitude smaller in the MP tests compared to the OP tests. Varying the loading rates, we found that the mainshock magnitude in the OP tests increased with recurrence time as expected due to fault healing. In contrast, the mainshock magnitude in the MP tests decreased with increasing recurrence time due to the increased effectiveness of VS barriers at slower loading rates. In some MP tests, foreshock-like events migrated at ~0.7 m/s, followed by faster reverse migration at ~7 m/s, resembling Rapid Tremor Reversal (RTR) in subduction zones. We used a numerical simulation to quantitatively reproduce the RTR-like behavior, help explain its mechanics, and constrain the friction properties of the laboratory system. Overall, our findings highlight how identical structural features on heterogeneous faults can behave differently under different loading conditions due to the velocity dependence of VS barriers.