There is an aspiration and challenge in live coding to control many aspects of the music during the performance as possible. Therefore, an important aspect of developing the practice is the search for efficient and expressive musical processes and their computational representations. In this talk I will outline some of my ongoing research into psychology-inspired algorithmic. The techniques build on a Gestalt psychology heritage. In keeping with this, the use of these processes in combination amounts to more than their sum. In implementing these techniques I have built on previous work on live coding methods and utilised libraries and design patterns from the Impromptu environment.The techniques are focused on symbolic music making in a tonal and metric setting. They are based on several principles. 1) Proximity: Implemented as a Gaussian walk through musical space: pitch, dynamic, duration, etc. 2) Range constraint: That creates a kind of ‘elastic band’ force that increases with the distance away from the mean. 3) Structural tones: Defining and probabilistically moving toward important points in the harmonic and/or melodic space, that can provide points of climax, variety and resolution. 4) Process: A tendency to repeat musical elements including pitch intervals, note durations. Hierarchically applied to sequencing musical segments. 5) Closure: Identifying degrees of stability that can indicate opportunities for changes (in contour, harmonic context, texture, and so on) and points of rest or ending.