I don’t know, sorry. You could try figuring out a formula for it, since it has to work with the input values being A=-1 and D=1. The reason why your current implementation sounds like it only works one way is because you probably have highest absolute value as your accumulation behaviour in your input action asset. What this means is that when both input values for A and D received by the input action asset, it will take the answer in absolute value (absolute values are always positive). So A=-1 and D=1 will always equal 1 as the absolute value, meaning one direction will always take higher priority over the other.