The capacity of dowel-type timber connections is generally governed by brittle failure modes, such as splitting and row shear failures. In design standards, this is indirectly addressed through the effective number of fasteners concept. This parameter accounts for both the non-uniform load distribution and the interactions among fasteners, which can induce brittle failure before the fasteners' capacity is fully mobilised. Although the current expressions for the effective number of fasteners in European design codes are based on tests on timber-to-timber connections, they are also applied to other configurations, such as steel-to-timber connections. The present study investigates, through numerical simulation, the validity of existing empirical expressions under various loading conditions and connection arrangements. The numerical model adopted was previously proposed and validated by the authors and is grounded in linear elastic fracture mechanics. The model validation performed herein on the original experimental dataset used to derive the empirical expressions for the effective number of fasteners further strengthens its validity. The parameters influencing the connection response across different configurations are further evaluated, including the fastener slenderness ratio and the spacing between fasteners. The findings reveal that both connection arrangement and loading direction (compression vs tension) substantially affect failure assessment and, by extension, the effective number of fasteners. Specifically, steel-to-timber and tension-loaded connections tend to fail earlier than timber-to-timber connections under compression, thereby leading to a lower effective number of fasteners. These critical factors are not adequately captured in current design equations, underscoring the need for revised approaches, supported by further experimental validation and probabilistic assessment, to improve the reliability of design expressions for the effective number of fasteners.