This thesis presents an experimental investigation of the plasma formation in the TCV tokamak. The primary goal of this work was to program a reliable and smooth plasma formation at several positions within the TCV vessel and then use the gained understanding to revisit the creation of doublet plasma formation. The first part of this thesis is dedicated to understanding and improving the single-axis plasma formation scenario in TCV. A database for the single-axis TCV plasma formation scenario was created for discharges spanning several years of operation to understand the physics of the plasma formation dynamics. The database shows that most of the failed plasma formation in TCV were during the burn-through and ramp-up phase with only 0.5% of the discharges failing at breakdown. The failed plasma breakdowns are mainly attributed to technical issues, such as no injection of neutral gas into the vacuum vessel, absence of the toroidal field or the Ohmic coil current, and issues with the plasma control system. The improvement of the single-axis plasma formation was separated into two parts: improvement of the breakdown scenario and improvement of the plasma burn-through and ramp-up scenario. During the breakdown phase, a large mismatch was exposed between the intended and experimentally obtained vertical breakdown position, for both Z=0.05m and Z=+0.23m standard vertical breakdown positions and for both I_P and B_phi directions. This mismatch was caused by an additional poloidal field mainly due to errors in the back-off of the stray field generated by vessel currents. The use of a vessel resistivity assuming axisymmetry in the TCV discharge preparation procedure to model the vessel currents was identified as the main reason for the mismatch in the breakdown positions. The analysis of the plasma formation database revealed that most of the failed plasma formation during the burn-through and ramp-up phase occurred due to insufficient, albeit often temporary, Ohmic heating to sustain the plasma. The insufficient Ohmic heating was either due an insufficient initial I_P ramp rate, or a combined effect of strong I_P and/or radial position oscillations caused by the feedback control system due to too high initial I_P ramp rate. A bump-less transfer control technique was implemented to improve the reliability of plasma formation by avoiding the strong oscillations in I_P and radial position that resulted in reliable and sufficient Ohmic heating. The second part of this thesis focuses on developing a doublet shaped plasma configuration, which is a highly unconventional plasma configuration, and TCV's modern and unique shaping capabilities warrant an effort to revisit the configuration. Successful simultaneous breakdown at two locations in TCV was achieved by using the improved inductive breakdown scenario. The similar magnetic properties of the two magnetic null points ensured that the plasma current ramp rate in the two droplets were close and the plasma current in both droplets was ramped up to 50 kA each with Ohmic heating alone. A highly reproducible doublet formation scenario was finally achieved and was verified by several diagnostics. A highly reproducible doublet formation scenario was finally achieved and was verified by several diagnostics. A highest plasma current of 130 kA was achieved in each droplet, with a core electron temperature at 1.3keV, core electron density at 1.3e19m^-3, and 30ms duration with ECRH heating.