We present herein a diagnostic method which provides insights into the interactions between parasitic reactions at battery electrodes and their consequences for battery performance degradation. We also provide a cautionary tale about misinterpreting or misrepresenting the significance of test data, as is sometimes found in the peer-reviewed literature or in developers' claims. Reversible cycling of the LiNic(0.5)Mn(1.5)O(4) positive electrode in a full cell with an electrolyte solution containing no additives may appear achievable through tailoring of the operating potential window of the cell. Self-discharging of the negative stems from parasitic oxidation products formed on the positive. We show that either excess negative electrode capacity over the positive or initial pre-lithiation of the negative suppresses their detrimental effect on capacity retention. Simultaneous monitoring the potentials of the two electrodes vs. Li/Li+ during galvanostatic cycling of a full cell shows, however, that self-discharging of the negative still takes place. The latter process was tracked by the drift of the average potential of the cell towards higher values and leads to two characteristic patterns in the failure of full cells during their long-term cycling, depending on whether a cut-off voltage or a capacity limit is used as the control criterion during cycling. (C) 2014 Published by Elsevier B.V.