Imagine a steel I-girder lifted by a single wire rope. If the rope is not perfectly centered along the girder length, the girder will rotate about a horizontal latitudinal axis through its centroid. Imagine a girder pair lifted by two wire ropes, with one rope on each side of the girder centroid, each about five percent of the length of the girder away from the centroid. If ropes are not perfectly equidistant from the centroid, the girder will rotate. Imagine a girder pair lifted by two wire ropes, each about five percent of the length of the girder away from an end. If the ropes are not perfectly equidistant from the ends, there will be a net moment / torque about the latitudinal axis, equal to the net moments above, but this situation will be more stable. Why? Additionally, for a rigid girder, cables at the ends would be most stable (stability score of 50), cables at the centroid would be least stable (stability score of -50), and cables each halfway between centroid and end would be stable in an intermediate way (stability score of 0). Why is this a linear relationship?