New experiments in 2013-2014 have investigated the physics responsible for the decrease in H-mode pedestal confinement observed in the initial phase of JET-ILW operation (2012 Experimental Campaigns). The effects of plasma triangularity, global beta and neutrals on pedestal confinement and stability have been investigated systematically. The stability of JET-ILW pedestals is analysed in the framework of the Peeling-Ballooning model and the model assumptions of the pedestal predictive code EPED. Low D neutrals content in the plasma, achieved either by low D 2 gas injection rates or by divertor configurations with optimum pumping, and high beta are necessary conditions for good pedestal (and core) performance. In such conditions the pedestal stability is consistent with the Peeling-Ballooning paradigm. Moderate to high D 2 gas rates, required for W control and stable H-mode operation with the ILW, lead to increased D neutrals content in the plasma and additional physics in the pedestal models may be required to explain the onset of the ELM instability. The changes in H-mode performance associated with the change in JET wall composition from C to Be/W point to D neutrals and low-Z impurities playing a role in pedestal stability, elements which are not currently included in pedestal models. These aspects need to be addressed in order to progress towards full predictive capability of the pedestal height.