There have been few attempts in the past to understand the collision of individual cases of interacting Coronal Mass Ejections (CMEs). We selected 8 cases of interacting CMEs and estimated their propagation and expansion speeds, direction of impact and masses exploiting coronagraphic and heliospheric imaging observations. Using these estimates with ignoring the errors therein, we find that the nature of collision is perfectly inelastic for 2 cases (e.g., 2012 March and November), inelastic for 2 cases (e.g., 2012 June and 2011 August), elastic for 1 case (e.g., 2013 October) and super-elastic for 3 cases (e.g., 2011 February, 2010 May and 2012 September). Admitting large uncertainties in the estimated directions, angular widths and pre-collision speeds; the probability of perfectly inelastic collision for 2012 March and November cases diverge from 98%-60% and 100%-40%, respectively, reserving some probability for other nature of collision. Similarly, the probability of inelastic collision diverge from 95%-50% for 2012 June case, 85%-50% for 2011 August case, and 75%-15% for 2013 October case. We note that probability of super-elastic collision for 2011 February, 2010 May and 2012 September CMEs diverge from 90%-75%, 60%-45% and 90%-50%, respectively. Although the sample size is small, we find a good dependence of nature of collision on CMEs parameters. The crucial pre-collision parameters of the CMEs responsible for increasing the probability of super-elastic collision, in descending order of priority, are their lower approaching speed, higher expansion speed of the following CME over the preceding one, and longer duration of collision phase.