Using modeling and simulation tools, we have demonstrated that the basic predictor module, which utilizes linear-dynamic predicting algorithms, has the potential to offer defense against bias injection type of cyber-attacks targeted at closed-loop-controlled systems, such as Kalman filter-driven brushless DC (BLDC) motors. Specifically verified is scenario of introduced earlier current drainage attack with full transparency for the defender, whose activity is shown to be able to counteract the manipulation game from the attacker. The attacker manipulation with distortion coefficient modifying angular speed estimates is shown to trigger significant growth of operating motor currents with accumulating effect on increasing speed of the rotor. Moreover, an abrupt termination of the attack is shown to result in even stronger currents increase in the forced braking process, creating potentially more damaging circuits overloads (to be called "attack withdrawal syndrome"). The defender is shown to be able to defeat the attacker within an asymmetric Stackelberg game relying on linear-dynamic predictor to preemptively compensate the distortion of the angular speed estimates. Demonstrated is potential for the defender to effectively negate the current drainage attack itself as well as its immediate consequence, such as attack withdrawal syndrome. Out of the reach for the defender remains the attack initiation impulse, whose damaging impact needs to be limited via dedicated algorithm with separate anomaly detector.