A 300 mile-range automotive battery pack is comprised of many individual cells connected in series/parallel to make up the required voltage, energy, and power. The cell groupings can take the form of parallel strings of series cell groups (S-P), series string of parallel cell groups (P-S), or a hybrid of the two. Though the different battery configurations deliver identical output voltage and energy, they exhibit varying cell level behaviors due to differing electrical structure, particularly when cell imbalance occurs. In this work, we explore the relative merits of various cell grouping configurations using a model-based approach. The emphasis of the study is to evaluate the impact of electrical variation between cell-to-cell, originating from cell manufacturing process variation, battery assembly (laser tab bonding) process variation or from normal operation, on the performance of the battery pack. A first-order equivalent circuit model is used to represent a lithium-ion cell. A model of the battery system with Nm modules of Ns cells in series and Np cells in parallel is developed and used to analyze the voltage and state-of-charge behavior of the system with various cell grouping configurations for a specified duty cycle. The imbalances between cell-to-cell voltage and currents are used as metrics for comparing different battery architectures. Using a low voltage battery system (<50V, 5 kWh nominal) as a case study, the simulation results indicate that the S-P cell grouping architecture is significantly more robust to cell-to-cell variations than P-S cell grouping architecture. Further, when paired with typical battery management system, the ability to diagnose a single cell malfunction (for example, faulty cell, poor tab weld, inadequate thermal management) is demonstrably higher for the S-P cell grouping architecture than the corresponding P-S system.