To identify critical technologies and make improved investment decisions, the technique of tech- nology road-mapping is commonly employed. However, most methods used today rely on a qual- itative approach, failing to capture the game-theoretic nature between competing actors and the stochasticity of technology development. This raises an important question: How can game theory be applied to model technological development quantitatively? The objective of this thesis is to develop a quantitative, game-theoretic framework that simulates technology investments while con- sidering the stochasticity of scientific impacts. A minimax optimization model using LSS provided by Mazumdar et al. (2019), has been implemented to approximate equilibrium points between the actors. The estimation of the probability distribution describing the likelihood of a specific scien- tific impact has been performed by utilizing citation data from academic papers related to relevant technologies. The data were fitted to a log-normal curve using MLE. 

The framework establishes a continuous two-player zero-sum game in which players compete to increase their probability of winning simulated military combat. Each player is provided with a technology portfolio consisting of various technologies available for investment. Investing in a technology leads to stochastic progress in that specific technology. The level of advancement in a player’s technology portfolio determines their military capabilities, which are tested against each other on a simulated battlefield. This incentivizes players to make investments that increase (or avoid to decrease) their probability of winning. Upon completion of the game, various scenarios of accumulated technology investments will be available for analysis. 

The results show that (i) the proposed method successfully uses game-theoretic concepts to model decisions of technology investments (ii) the estimated probability distributions follow the findings done by Radicchi et al. (2008), that citation data can be well-fitted to a log-normal PDF at a signif- icance level of 5% (iii) the LSS algorithm developed by Mazumdar et al. (2019) show successful convergence towards (local) Nash equilibrium in the implemented game. From the results of the technology road map, it can be concluded that the potential of a quantitative framework, such as the one proposed in this study, is of great significance as an aid to qualitative methods that are used in the field today. With further work, the framework can serve as a valuable tool for incorporating a quantitative perspective into the decision-making process of allocating technology resources.