Many studies have highlighted the advantages of expanding the input space of mobile devices by utilizing the back of the device (BOD). One of our previous studies conducted an elicitation study to explore BOD gestures and determined several user-defined gestures. We extended this work by investigating the recognition of these gestures using scratch input (as described by Harrison and Hudson in 2008) for an Introduction to Machine Learning Course. To amplify the sound of scratches performed on the back of a mobile device, I designed and manufactured a stainless steel adhesive scratchpad using a Silhouette Cameo cutting machine. This scratchpad succeeded in making BOD gestures more audible and therefore more viable for recognition.