Expected behavior of a combination of Model 1 and Model 2, the DDM with both threshold and drift adaptation (drift = d · (R1 – R2), θi = ξ/Ri, with d = 0.4, and θi = 0.5). Note the rotation once again of the plot in panel B relative to panel A and C, and its truncation to 5 sec. As with Model 1 alone, the absolute magnitude of obtained reward rates has little effect on the quality of the approximation to strict matching.