Cooperative Games and the Ecological Data Economy

Kyle Birchard

Part 1: Metaphors and market design
The advent of token economies has given us a profusion of metaphors to describe how these economies work. A non-exhaustive list of token economy metaphors would include digital gold (assets), digital cash (currencies), and arcade (utility/access) tokens. This is a necessary step in any emerging field, and as time goes by, that language will emerge. Indeed, we are already seeing this today, with talk of cryptoeconomic primitives and other building blocks.

When electronic instruments such as the theremin and the analog synthesizer were invented, early players often adapted existing musical repertoire. The results could be stunning, like Clara Rockmore’s rendition of Saint-Saëns’ “The Swan,” but, because the instruments were so new, there wasn’t a native language available to capture the full expressive range and tonal color made available by these new inventions. The same can be said for many other technologies, from photography to cinema to computer animation: it takes time for new modes of interacting with these technologies to emerge.

We are in a similar state with crypto assets today. In attempting to find our way in this space, we adapt familiar notions, such as the token economy concepts listed above. These metaphors are useful cognitive shortcuts: they facilitate a common understanding of complex phenomena, which is a good thing in itself, but this hopefully allows us to invent the new vocabularies we will need to evolve beyond where we are today.

This series begins with a treatment of a commonly-used metaphor/toolkit in cryptoeconomics, game theory, and ends with another model used in the field of ecological economics. In between, we will discuss value creation in the ecological data economy, how that value is distributed among the economy’s members, and different approaches to implementing these concepts in the real world, which is what we’re trying to change, after all.

The very term game theory is itself a cognitive shortcut that permits people in fields as disparate as politics, mathematics, computer science, economics, and behavioral psychology to collaborate on some of the interdisciplinary problems of our time. The applications of game theory and market design in token economies have been covered in a number of places (see here and here, for example) and will not be covered here in too much detail. Rather, we will use a specific area of study, cooperative games, as a jumping off point for our discussion of ecological and data-driven economic systems, with a particular focus on the developments emerging from Regen Network.

The Network is the Value
IBM-owned Weather Underground reports weather data obtained from roughly 280,000 connected devices in the United States. While the data from any individual weather station is worth little by itself, transformative value emerges when those data are combined with information and analysis from other sources.

The value of weather information, if fully monetized, has been estimated at roughly $14 billion per year in the United States. This amounts to an average value of $50,000 per station for the network. Even at current levels of weather data monetization, roughly $2 billion per year in the U.S., over $7,000 per year in revenues can be attributed to an “average” weather station (yes, I’m aware of the limitations of that definition) — but only if the network is considered as a whole.

This phenomenon was recognized by some of the pioneers in game theory, namely Lloyd Shapley and Martin Shubik who, in a remarkable string of articles dating from the early 1950s, helped develop a theory of cooperative games that, in contrast to zero-sum or adversarial contests such as the Prisoner’s Dilemma, shed light on ways economic surplus arises from people working together within organizations, and how seemingly insignificant participants in a system can organize in ways that make an outsized impact.

Theory of cooperative games
In the sense used here, a cooperative game is comprised of coalitions of individuals working toward a common goal. While those early papers often considered business decision makers, in the present day, we might also model these actors as households, firms, sensors, data owners, personal AI agents, DAOs, and other entities that haven’t yet been invented.

The key feature of coalitional forms is that working together creates a larger surplus than the sum of any of the individual actors. In one sense, this is almost too obvious to mention, but one of the important contributions of game theory is a formal representation of “cooperation” that gives us ways to measure this surplus in quantitative terms and to embed rules in code that can be deployed, for example, in Ecological Contracts executed over the Ecological State Protocols envisioned by Regen Network.

Other properties we are interested in with cooperative game theory are functions that are superadditive (the whole is greater than the sum of its parts) and supermodular (joining a coalition yields increasing returns as the coalition size gets larger, i.e., increasing returns).

This is a good time to introduce the Shapley Value, a measure of the contributions made by the individuals that comprise a coalition. Introduced by Lloyd Shapley in a seminal 1953 paper, this value presents a dynamic allocation mechanism that reflects the value contributed by the entities participating in a coalition at any given time.

Until recently, the Shapley Value was largely of theoretical interest because it is so computationally expensive: the number of computations grows as a factorial of the number of agents in the coalition. So, for example, if you have 20 members of a coalition, you would need to compute 20! combinations for each of the 20 members (or 2,432,902,008,176,640,000 computations each). Improvements in processor power and new approximation methods have made it possible to compute values for larger groups, and the Shapley Value has recently been used to study coalition formation in automated negotiation systems, the economics of peer-to-peer networks, and even the contributions of individual players to an English soccer club. Given some of its interesting properties, the Shapley Value and its variants could see important application to data valuation in cryptoeconomic settings.

There are other ways of allocating the surplus created by cooperation over economic networks, which we will discuss in a later article. These include specific rules, such as proportional and equal division schemes, which are not always optimal or fair, as we will see. They also include interactive systems such as continuous auctions and negotiation platforms. Our goal will be to illustrate how specific allocation methods can be implemented in the context of an Ecological State Protocol.

Coevolutionary approaches to token economies
Did you see what I did just there? I just switched metaphors from cooperative games to networks. This is another major challenge in the early stages of a field — finding the metaphor that works best may evolve over time as the problems are better understood.

There are other useful ways to depict organizations, coalitions, and economic production models. Network science and information theory both have important applications in economic settings, as well as other ecological and social disciplines. Together, these contributions at the intersection of social, computer, and hard science raise the possibility of a data economics and governance that is native to the world of sensors, distributed systems, and diverse cultures.

And while it may take time and some artistic refiguring, eventually we will find the right language for the ecological economy…even though it will always be changing.

Future articles in this series will include the following topics:

1. How networks allow us to generate surplus value

2. Allocating the surplus generated by cooperation

3. The role of accounting models in the ecological data economy

4. Using economic elements in an Ecological State Protocol / Ecological Contract framework

5. “Eye-opening metaphors” in ecological economics: beyond stocks and flows

Follow Regen Network
Website: http://regen.network
Facebook: https://facebook.com/weareregennetwork
Twitter: http://twitter.com/regen_network
Medium: https://medium.com/regen-network
Reddit: https://reddit.com/r/regen_network/
LinkedIn: https://www.linkedin.com/company/regen-network/
Telegram: http://t.me/regen_network

You can find the original article on Medium

Images: