The US political arena is generally a bleak mix of distressing and aggravating, particularly in the current highly polarized climate, but that distress only increases the need to honestly and scientifically understand the forces at play in ideological space, and the polarization offers emergent simplicity and thus a unique opportunity for mathematical modeling. This talk will share some recent and ongoing attempts to visualize and understand the current political landscape, as well as some quantitative patterns in political psychology and the modern information ecosystem. These recently-gathered data inform a mechanistic dynamical model of ideological drift, allowing theory to extrapolate the complex implications of micro-scale data to macro-scale outcomes---while iteratively improving and suggesting further data-gathering efforts to corner remaining uncertainty. The first waves of results from this perspective provide some remarkable insights, providing some hope of understanding and productively informing political messaging and algorithmic design for a more reasonable political future.