Jason B. Scholz

The Effects of Material, Tempo and Search Depth on Win-Loss Ratios in Chess

What exactly is it that gives one player a decisive advantage over the other in an adversarial contest? We have played tens of thousands of simulated chess games to compare how different kinds of advantage contribute to overcoming an opponent. This has allowed us to quantify the effects of search depth, material strength, and tempo on game outcomes. The results indicate that tempo and material are both advantageous, but that the degree of advantage is highly dependent on skill level. Skill level, as reflected by search depth, leverages small inequalities in material or tempo. In addition, searching one move deeper than an otherwise equally matched opponent equates to an advantage of between 85%-98%.

Unravelling Bueno De Mesquita’s group decision model

Political scientist Professor Bruce Bueno De Mesquita, has made significant claims for the predictive accuracy of his computational model of group decision making, receiving much popular press, including newspaper articles, books and a television documentary entitled ‘The New Nostradamus’. Despite these and many journal and conference publications related to the topic, no clear elucidation of the model exists in the open literature or can be found in a single place. We expose and present the model by careful navigation of the literature and illustrate the soundness of our interpretation by replicating De Mesquita’s own results. We also raise issues regarding sensitivity and convergence.

Ubiquitous command and control

We assert that the current conceptualisation of military operations has itself become a legacy system, which fails to take advantage of emerging technology and evolving commercial business practice.In response to this, we present a conceptual framework for the development of potentially extremely robust societies of human decision-makers and automated machine decision systems appropriate for future military conflict, that is commensurate with emerging technology. The framework is based on a context of adaptation of location, through telecommunications and transportation resulting in richer human presence, and its ensuing effect on the function and structure of enterprises. Eight principles or tenets are presented which exploit this context. We present principles of decision devolution, ubiquity, automation, human and system integration, distribution and decentralization to each provide forms of robustness through diversity. Principles of social coordination and management provide unity and bounds on this diversity.

Dynamic Games to Assess Network Value and Performance

This paper looks at the analysis of network effectiveness and vulnerability through dynamic games. Each opposing side consists of a collection of vertices (pieces) connected in a network. Only vertices in the largest sub-graph exhibit mobility. We use the mobility and piece removal rules in the game checkers and modify the evaluation function to include sub-graph size balance. With this modified evaluation function, win-lose results of richly versus sparsely connected and centralised versus decentralized topologies are analysed. The results are compared with the current vulnerability studies in networks of varying topology. Finally we use temporal difference learning to calculate advisor weights for richly or sparsely connected vertices.