The authors propose a modification to the spin transfer torque term in the Landau-Lifshitz-Gilbert (LLG) equation to explain the motion of magnetic domain walls (DWs) in nanowires under current. They show that this modification can reasonably explain the measured DW velocities and the variation of DW propagation fields under current. The introduction of coercivity by considering rough wires leads to a finite DW propagation field and threshold current for DW propagation, suggesting that threshold currents are extrinsic. The modified LLG equation includes a second term describing the current torque, which competes with the damping term. Micromagnetic simulations confirm that this term resolves the discrepancy between theoretical predictions and experimental results, particularly for DW motion in perfect wires. The authors also discuss possible models that support the new term, including the effects of spin accumulation and momentum transfer. The magnitude of the new term, denoted as \(\beta\), is estimated to be around 0.04 for permalloy, and its material dependence is discussed.The authors propose a modification to the spin transfer torque term in the Landau-Lifshitz-Gilbert (LLG) equation to explain the motion of magnetic domain walls (DWs) in nanowires under current. They show that this modification can reasonably explain the measured DW velocities and the variation of DW propagation fields under current. The introduction of coercivity by considering rough wires leads to a finite DW propagation field and threshold current for DW propagation, suggesting that threshold currents are extrinsic. The modified LLG equation includes a second term describing the current torque, which competes with the damping term. Micromagnetic simulations confirm that this term resolves the discrepancy between theoretical predictions and experimental results, particularly for DW motion in perfect wires. The authors also discuss possible models that support the new term, including the effects of spin accumulation and momentum transfer. The magnitude of the new term, denoted as \(\beta\), is estimated to be around 0.04 for permalloy, and its material dependence is discussed.