Clark Aldrich is the author of Simulations and the Future of Learning, and Learning By Doing, and a designer of educational simulations. His Blog presents his educational theories about games and simulations. As an introduction to his thinking, he discusses the "Four Sweet Spots" or major cognitive skills, that are inherent to games and simulations: Situational Awareness is the ability to quickly determine the most important details for understanding and controlling a situation; Understanding of Actions is the ability to know what's possible to do, and to weigh the viability and trade-offs of these actions; Awareness of Patterns is the ability to understand how systems work and which small actions will have the most impact; and Conceptual Dead Reckoning is the ability to see the big picture as a kind of map, plot a course to a desired location, and adjust to real world situations to stay on course. Simulations and games help people learn "Big Skills" and "Middle Skills", which aren't usually taught in school. Big Skills, often called "soft skills", are things like "leadership, nurturing/stewardship, communication, and relationship management." Developing them requires practice, going through cycles of "frustration and resolution", and applying "Middle Skills", such as managing people, adapting to new situations, and creating functional processes. The Blog is structured like a book, with labels or keywords organizing blog posts into chapters and organizing definitions of related terms into a glossary. Below are summaries of a selection of posts relevant to using simulations and games for learning.

Simulation Elements: why libraries are obsolete, self-help books don't work, and Google won't help you learn to ride a bike or be a better leader
Simulations enable users to determine what actions are available, how actions affect systems, and how those systems operate to produce results, and they are driven by questions and learning goals.

• Actions: Actions are what people do at the most basic level, and have two components: timing and magnitude. The scope of the simulation determines the granularity of the actions. "The educational challenge of actions is building a new awareness of students’ real-world options."
• System Content: "Systems are the layer between what we do (Action) and what we get (Results)." Aldrich proposes 7 types of systems in simulations: Pure Mathematical System, Units on Maps as System, State Based System, Artificial Intelligence as System, Work Process as System, Middle and Big Skills as a System, Community as a System.
• Process, Work: Processes are about transforming an idea, thing, or person, delivering a service or message, branching to other processes, or triggering an action. Usually someone owns a process and takes responsibility for managing it. Processes range in complexity and usually need to be connected to other processes to bring about results.
• Unit: A discrete object or person. Units have attributes and exist in context with other units and processes. They can be purchased, discovered, built, placed, upgraded, and destroyed. Units play a role in an ecosystem, they transform and consume resources, and they often produce emergent behavior when many act at the same time. Units wit artificial intelligence are usually called agents, and units representing players are usually called avatars.
• Maps: Most games and sims use a map of some kind, whether it be spatial, conceptual, abstract or geographical. A map focuses the activities of the players, and one's location on a map is usually essential to the context of the simulation. Locations on maps have value, and taking control of locations is often a crucial aspect of one's strategy. The use of Maps should be carefully considered: Anytime you draw a map you are drawing something that has a distance function, and anyone looking at your map will see the distance function. If you're trying to visualize something that doesn't really have an analog of "distance", your map may include more relationship than what you're trying to show. And that can be very misleading!
• State Based Systems and Models: State Based Systems and Models are special types of maps, with discrete, organized elements, such as rooms or levels. Often different states are organized hierarchically, and moving between them requires unlocking a "gate" by achieving a goal. Educational simulations and games often involve a branching story, where the player  chooses a path through different states, or alternative branches of the storyline.
• Desired Results/Mission: Types of missions include: managing a process, defeating a competitor, developing a functioning system, discovering/implementing a new process, or creating an outcome that is measured against a balanced scorecard, and developing Big Skills either in individuals or in organizations.

Game Elements: A spoonful of sugar if you can avoid hypoglycemic shock
Game elements motivate people to engage with a simulation. These include elements such as fantasy, competition, and story. Without them simulations can become dry, and this transfers to the learner's opinion of the content as well. Too many game elements, however,  can distract from the content of the simulation, or make it seem silly or frivolous. "As a rule, the more one cares about content, the more one is intolerant of game elements."

Pedagogical Elements: Learn faster and better
Pedagogical elements in sims facilitate the participant's learning. Elements such as in-game tips and directions, graphs, breadcrumbs, and after action reviews help learners know what to do, how to use the interface, avoid developing superstitious behavior, see relationships faster, work through frustration-resolution pairs, try different approaches, and apply lessons learned in the simulation to the real world.

Aldrich applies his analysis of what makes for a good educational simulation to second life in
Commentary: Top Ten Missing Features of Second Life as an Educational Simulation Platform