Sensitive information used online has resulted in the emergence of many security systems to be protected by secret keys. One of these security systems is the so-called secret sharing, in which certain key portions are distributed to allow sharing access by a specified number of participants. The sharing system requires that several users enter their shares, at the same time, to rebuild the main target key (TK) to access sensitive data within the system. Normally, the number of shares is limited by the number of zeros within TK as of our focus on the recent counting-based secret sharing strategy. If the number of participants is more than TK zeros, the complete system cannot work needing to reconsider regenerating TK again. In this work, we improved the original system of counting-based secret sharing to accommodate more participants via longer trustworthy TK selection. We proposed several possible models to enlarge the number of participants for applicable share’s keys generated from TK. The trustworthy security randomness of the new shares and adjusted target keys were all measured to select the best choice based on standard reliability randomness test. The study experimented the proposal on typical simulations applied on different target keys applicable sizes. The results presented that the final target key TK created as having the highest ratio of randomness to be preferred choice proven secure and reliable to protect information within the system.