The information based on the tripartite (truth, indeterminacy, and falsity) can be described by single-valued neutrosophic numbers, but the linguistic information based on these tripartite cannot be represented by single-valued neutrosophic numbers. To overcome this obstacle, linguistic single-valued neutrosophic numbers are characterized independently by the truth, indeterminacy, and falsity linguistic variables. In this paper, some new operations are introduced to increase the richness of linguistic fusion based on the linguistic single-valued neutrosophic approach. The Hamming distance-based formula is proposed to measure the distance between two linguistic single-valued neutrosophic numbers. Afterwards, we initiate the theory of linguistic single-valued neutrosophic soft (LSVNS) set by combining linguistic single-valued neutrosophic set and soft set. Meanwhile, the different computational functions and desirable properties of LSVNS sets are studied. Furthermore, we construct the framework of LSVNS technique for order preference by similarity to ideal solution (TOPSIS) and then present a game theory model based on this framework. Finally, we give an outstanding example to validate the practicality and effectiveness of the proposed LSVNS TOPSIS-based game theory model.