Virus self-assembly is a critical step in the virus lifecycle. Understanding how viruses assemble and disassemble provides needed insight into developing antiviral pharmaceuticals. Few tools offer sufficient resolution to study assembly intermediates that differ in size by a few dimers. Our goal is to improve resistive-pulse sensing on nanofluidic devices to offer better particle-size and temporal resolution to study intermediates and capsids generated along the assembly pathway. To increase the particle-size resolution of the resistive-pulse technique, we measured the same, single virus particles up to a thousand times, cycling them back and forth across a series of nanopores by switching the polarity of the applied potential, i.e., virus ping-pong. Multiple pores in series provide a unique multipulse signature during each cycle that improves particle tracking and, therefore, identification of a single particle and reduces the number of cycles needed to make the requisite number of measurements. With T = 3 and T = 4 hepatitis B virus (HBV) capsids, we showed the standard deviation of the particle-size distribution decreased with the square root of the number of measurements and approached discriminating particles differing in size by single dimers. We then studied in vitro assembly of HBV capsids and observed that the ensemble of intermediates shift to larger sizes over 2 days of annealing. On the contrary, assembly reactions diluted to lower dimer concentrations an hour after initiation had fewer intermediates that persisted after the 2 day incubation and had a higher ratio of T = 4 to T = 3 capsids. These reactions indicate that labile T = 4 intermediates are formed rapidly, and dependent on conditions, intermediates may be trapped as metastable species or progress to yield complete capsids.