Video uses a property of the error function for a Taylor series expansion from the previous video to bound the (n+1)th derivative of the error function. Result is the Lagrange error bound. [15:08]