This video outlines a three-hour workshop led by James Payne, System Dynamics instructor at MIT OpenCourseWare. System dynamics is a discipline that understands and solves complex problems through the concepts and methodologies of system thinking. This summary covers the essentials of what system dynamics is, how it works, and how it can be applied to a variety of real-world problems. The instructor explains the history, key concepts, modeling tools, and practical applications of system dynamics in a friendly and easy-to-understand manner, and guides participants to experience systems thinking through simulation.

1. Introduction to system dynamics and instructor introduction 👨‍🏫

The video begins with a greeting from instructor James Payne and an introduction to a 3-hour system dynamics overview session held by the MIT System Dynamics Group. The instructor stated that this session would highlight the nuances between Systems Thinking and System Dynamics and explain how these two concepts interact in a complex world.

Highlights from today's session include:

• Systems Thinking Overview: Covers the conceptual differences and overlap between systems dynamics methodology and systems thinking.
• Executive Flight Simulator: A hands-on session that allows participants to experience system dynamics first-hand. (Not included in the actual video due to copyright issues)
• Post-simulation discussion: Lessons learned from the simulation are discussed and linked to system concepts.
• Introduction to additional materials: Guide to system dynamics and related materials from MIT.

The instructor introduced himself as a second-year PhD student in MIT's System Dynamics group and shared his 10 years of experience working in various industries, including chemical engineers, industrial engineers, and product managers. In particular, he said that the System Dynamics Group understood and accepted his diverse career through his experience working in a variety of fields from nuclear power plants to bras. "When I told them that I had worked in everything from nuclear power plants to bras, their reaction was 'Oh, I see.' It was." He said. 💡

The instructor explained that his main research interest is Behavioral Operations Management, which involves analyzing the human decision-making process within a supply chain and studying how to reflect this in system design. "In behavioral operations management, the presence of people is a feature, not a bug. How can we plan for it and leverage it?" He emphasized the importance of human behavior.

2. MIT System Dynamics Group and History 📜

The instructor introduced key faculty members of MIT's System Dynamics group, particularly emphasizing that Professor John Sterman is a master in this field. He explained that Professor Sterman's book, Business Dynamics, has now become the de facto standard textbook in the field of system dynamics. He recommended that reading just the first two chapters of this book will greatly help you understand the fundamental concepts of system dynamics and systems thinking.

System dynamics is a field that originated at MIT, and started from the idea of ​​Jay Forrester. Forrester formalized this concept by publishing Industrial Dynamics in 1958. From his perspective, the instructor defined system dynamics as "the application of control theory to social systems." 🧑‍🏫

Forrester was the inventor of RAM (Random Access Memory) and a man with an electrical engineering background who helped develop MIT's first general-purpose digital computer. He sought to apply his engineering background to business systems in an attempt to solve their complex social, bureaucratic and structural problems. Through his early projects with General Electric, he developed a system dynamics model analyzing the problem of three-year employment instability. Afterwards, he was invited by the Club of Rome to develop the World Two simulation, which applied this model to a larger social system, which led to the controversial book 『The Limits to Growth』. The instructor explained that this book was intended to deliver a social structural message about resource depletion and the limits of regeneration, rather than a specific future prediction.

Doctoral students in MIT's System Dynamics group come from a wide variety of backgrounds, including engineering, aerospace, electrical, biology, geophysics, and applied mathematics. The instructor explained that this diversity is due to "a shared goal of recognizing that problems exist within systems through a systemic approach and applying this to a broader area."

The instructor introduced two cases he is currently researching.

1. Nonprofit Organization Decision Making: This is a study that models the unique pressures and performance environments that nonprofit organizations face and analyzes cases where achieving specific goals is structurally impossible.
2. Multi-tier supply chain and inventory fluctuation: A study based on the 'beer game', analyzing the impact of human ordering behavior on the supply chain and exploring algorithmic control mechanisms within supply chains with delays. "We're talking about an algorithmic approach to the control mechanisms that exist within a supply chain with delays, taking the idea of ​​people being human and letting their ordering behavior flow," the instructor explains, noting that in addition to traditional system dynamics tools, they are working on optimization using TensorFlow and Python.

3. Key concepts of systems thinking: feedback loops and structures 🔄

To illustrate the core concepts of systems thinking, the instructor pointed out that the typical problem-solving process (identifying the problem, collecting data, evaluating alternatives, selecting and implementing a solution) is an incomplete model. He emphasized that every decision we make does not exist in isolation, but is embedded and interactive in a larger system.

"No decision you make exists in isolation. No process you influence exists in isolation. The only difference is the boundaries of your system. In fact, I would go so far as to say that there are no such things as side effects. There are just effects that you haven't thought of yet."

The instructor explained that decisions to solve a problem change the state of the system, which in turn affects our goals and decisions, and can lead to unexpected side effects. These sequential interactions and delays ultimately determine the behavior of the system, and this is the fundamental concept of system-level thinking.

A system is "a set of interdependent parts with a common purpose," and socioeconomic systems are characterized as being dynamic, tightly coupled, governed by feedback, and non-linear. Systems thinking is a framework for understanding and closing these feedback loops.

In system dynamics, it was explained that it is important to derive mental models. It's about understanding why individuals make rational choices based on specific information, and modeling that process to see how it interacts in the context of a larger system. He also said that it is much more efficient to predict and analyze the impact of change through simulation rather than directly making changes to the social system.

"The purpose of simulation is not to be right. The purpose of simulation is to be useful."

He emphasized that rather than predicting the exact future, simulations are useful for identifying high-leverage policy elements of the system to influence desired outcomes.

The three key points of systems thinking and system dynamics are:

1. Structure generates behavior: This means that people's behavior is determined by the structure and environment of the system in which they exist.
2. Importance of mental models: Not only the physical system structure, but also people's thinking and decision-making processes are important structural elements of the system.
3. Fundamental Attribution Error: We often blame people for problems, but in reality, structural flaws in the system are often the cause. The instructor explained this error with the example of when another car cuts you off while driving, you think, "What the heck!", but when you are in a hurry, you rationalize cutting into the other car. He added that even within the MIT group, to overcome this error, we share the basic assumption "We believe that everyone in this community is intelligent and capable, strives to do their best, acts with integrity, and wants to learn."

The instructor explained the concept that 'structure creates behavior' through the hierarchical relationship of events, patterns of behavior, and structure. Using oil price fluctuation articles as an example, he explained that individual events (rising and falling oil prices) come together to form behavioral patterns such as 'boom and bust' patterns, and that these patterns are ultimately caused by the fundamental structure of the system, such as physical structure, information availability, and actors' mental models.

They also noted that it is very difficult to learn and change in dynamically complex environments, citing limited information and time delays** as the main reasons for these difficulties. System dynamics is this **applied systems thinking** that focuses on generating research results to improve complex social systems.

4. System dynamics modeling tools 🛠️

The instructor explained that system dynamics modeling is an iterative and spiral approach. This is not a process of creating a perfect model from scratch, but rather a process of gradually improving the model.

One of the commonly used tools in system dynamics is the stock and flow diagram or compartmental model.

4.1. Causal Loop Diagrams

• Causal Links: Represents the basic principle that "when one thing changes, the other changes." For example, when production increases, inventory increases (positive link), and when shipments increase, inventory decreases (negative link).
• Difference between causality and correlation: The example of ice sales and murder rates increasing together emphasizes that simply because there is a high correlation does not mean there is causation. In reality, the hidden variable of rising average temperatures is the common cause of both phenomena. The instructor said that "causality is the domain of mathematicians and physicists," and that social system modeling takes consistency as the standard for causal relationships. "When one thing moves in one direction, the other always moves in the same direction."

"Causal relationships are the domain of mathematicians and physicists. True causal relationships are unresolved. Therefore, when one thing happens, we have to ask whether something else follows, and whether they are truly connected in space and time."

There are two types of causal loops:

1. Reinforcing Loop (R): A positive feedback loop that keeps the system moving in one direction. For example, as employee skills improve, customer satisfaction increases and complaints decrease, which allows managers to spend more time coaching employees, which in turn improves employee skills – a virtuous cycle.
2. Balancing Loop (B): A negative feedback loop that tends to return the system to a steady state. For example, as the attractiveness of a market increases, the number of competitors increases, which leads to lower product prices and lower profits, which reduces the attractiveness of the market. The goal-seeking loop is also an example of a balance loop.

4.2. Stocks and Flows

Stock is an element with memory within the system and is a concept in which changes accumulate over time. It can be compared to a bucket filled with water. Flow is what changes the amount of this inventory.

• Inventory example: Number of employees, inventory units, assets (balance sheet), wealth, carbon dioxide atmospheric concentration
• Flow example: adoption rate, production, sales, income and expenses, carbon dioxide emissions.

The instructor explained that 'interest rate' is easy to misunderstand as flow because of the word 'rate', but it actually represents the price of money and thus corresponds to inventory. It was emphasized that inventory and flow can vary depending on how variables are defined.

5. Simulation practice and its meaning 🕹️

The instructor explained the Executive Flight Simulator, one of the important teaching methods of the MIT System Dynamics Group. This helps participants experience the problem firsthand and build mental models.

As an example, we introduced the "Fishbanks" simulation game. In this game, participants take on the role of fishermen and must catch fish and earn money. The instructor explained that this simulation was developed as a board game in the past and was converted into an interactive simulation. (The simulation play scene was deleted from the video due to copyright issues)

During the post-simulation Q&A, one student asked, "I think the game broke so quickly because I didn't understand the system or equations behind the model. Was the information hidden intentionally?" The instructor responded as follows:

"To a certain extent. One of the interesting ways to run this simulation is to run it all with you again, which means everyone in the room plays the game again, but with slightly different parameters. So you can't set up what you want right from the start. The argument here is that information is imperfect, but it still exists, and some information is better than no information."

The instructor once again emphasized the concept that 'structure creates behavior' and explained that even if you know the rules or equations of the game, it is difficult to change people's behavior patterns if the fundamental structure does not change. He used the example of Beer Game to emphasize the powerful influence of structure, saying that even professors who teach the game have not been able to escape the game's trap.

6. Additional Resources and Tools 📚

The instructor recommended various resources for those who want to learn more about system dynamics.

6.1. MIT System Dynamics Lecture

• 15.871 Introduction to System Dynamics: Focuses on inculcating system thinking concepts. Learn how to move from a linear mindset to a loop mindset by drawing causal loop diagrams.
• 15.872 System Dynamics II: Focuses on applying the thinking processes learned in 871 to real-world scenarios and becoming familiar with modeling tools.
• 15.873 System Dynamics for Business and Policy: Moderate difficulty between 871 and 872, with emphasis on business and policy cases. The instructor emphasized the importance of unintended consequences by giving the example of the "Vasa", a ship built with the best technology but sank due to last-minute design changes. "There are no side effects. There are just effects you didn't plan for."

6.2. recommended books

• John Sterman's 『Business Dynamics』: Considered the bible of system dynamics, the first two chapters cover key concepts of systems thinking.
• Peter Senge's The Fifth Discipline - Fieldbook: Part 2, the 'Systems Thinking' section, provides a good summary of the work on system dynamics.
• 『The Limits to Growth』: This is a book that helps you understand the historical importance and controversy of system dynamics. The instructor explained that rather than predicting the future, this book is a discussion of the system structure that causes patterns of excess and collapse.

6.3. Recommended papers

• 『System Dynamics 660, the path forward』: Covers the current and future direction of system dynamics, emphasizing that system dynamics is not simply limited to a compartmental model but a way of thinking.
• 『Making the Numbers』: This is an interesting paper that applies the classic operations management concept, capability trap, to stock valuation.

6.4. Other websites and communities

• Creative Learning Exchange: A website focused on K-12 education that explains system dynamics concepts visually and interactively.
• Tom Fiddaman's MetaSD: You can see examples of Dr. Tom Fiddaman analyzing various interesting phenomena using system dynamics models. (e.g. the self-reinforcing nature model of UFO sightings)
• Self-Study Website: Free study material that allows you to systematically study from the basics of system dynamics to advanced modeling.
• System Dynamics Society: A system dynamics-related society that provides a variety of information and community activities.

6.5. software tools

• Vensim: Provides a free academic license, and the full commercial version is available for free to users with MIT email addresses.
• Stella Architect: Storytelling mode and visualization features are its strengths.
• NetLogo: An agent-based modeling tool that is useful for modeling the behavior of individual entities to understand the behavior of the system as a whole.

The instructor emphasized that the important thing in system dynamics is way of thinking, so you can use whatever tool is right for the job. He even suggested that you can start modeling the system with just a whiteboard, suggesting writing a specific phenomenon (e.g. carbon dioxide in the atmosphere, the number of deaths in car accidents, or the number of employees leaving a non-profit organization) in the center of the whiteboard and drawing a loop by connecting the factors that increase or decrease it with arrows.

Lastly, the instructor introduced a climate policy simulator called En-ROADS. The tool draws on complex climate models to show the impact of different policy choices on future climate change, and is a great example of how system dynamics can be used to drive real policy change.

Conclusion ✨

Instructor James Payne once again emphasized that system dynamics is a way of thinking that goes beyond simply creating complex models, to understanding the interconnectedness of decisions in a complex world, to insight into the fundamental principle that structure creates behavior, and to solving problems from a systemic perspective beyond the fundamental attribution error. He concluded his lecture by encouraging people to apply this way of thinking to real-world problems through hands-on experience through simulations and using a variety of tools and learning materials. Rather than accurately predicting the future, system dynamics provides a powerful framework for understanding system behavior patterns, finding high leverage points, and driving effective policy change.