News, research and discussion on virtual goods, currencies and economies globally.

Gambling and econometrics in the new issue of Journal of Virtual Worlds Research

Journal of Virtual Worlds Research, an online outlet for virtual worlds -related "think pieces" and scholarly papers, has published its sixth issue, titled Technology, Economy and Standards. Below is my selection of economy related papers from the issue, with comments.

Payback of Mining Activities Within Entropia Universe

Markus Falk, Daniel M Besemann, James Michael Bosson

In subscription-based virtual worlds the fee a user pays for participation is clear. However, in worlds in which the provider's revenue is generated via transactions made by users using a real in-world currency, lack of transparency in the underlying mechanisms can make it difficult for a user to gauge the service fee being paid. This paper studies the payback of mining activities within the virtual world Entropia Universe, with an aim to determine the cost of participation in this activity for the user. Entropia Universe developer MindArk estimates the normal service fee for an active user averages at $1 per hour. We compare our findings to this figure and observe how transactions between players can affect a user's returns. A statistical approach is employed, based on a large number of data points acquired by two Entropia Universe avatars. A theoretical mining-returns model is also created, consistent with our data sets, and is used to make predictions about general cost and profitability for Entropia Universe miners. These methods could be used to analyse the economy of other activities in Entropia Universe, and possibly activities other virtual worlds.

My comments:

This paper nicely summarises the Entropia Universe economy, which is described as a game of chance played against the house. A level of skill is required to maximise one's chances of winning, but in the long run, the house always wins. Since Entropia's virtual currency is redeemable for real money, it is essentially gambling. Blackjack seems to be the the closest comparison that comes to my mind.

The empirical part of the paper is a curious effort of reverse engineering the loot probability distribution set by the game's programmers. It represents a sort of natural science of virtual spaces: meticulously executed, but of dubious scientific value, because no new knowledge is created (the programmers know the loot distribution, even if they won't tell). The paper would have been tremendously improved if the authors could have shown what useful or interesting implications flow from this loot distribution table. For example, the results might have been linked to behavioural psychology, gambling studies or business model research. I hope the "Journal of Virtual Worlds Research" does not stand for "journal of research on how the laws of nature work in virtual spaces".

Still, this is one of the better papers of the issue. For me, the most interesting parts are the (unsourced) descriptions of Entropia Universe's players' own conceptualisations of the activity they are engaging in, and of the approximately $1 per hour which they are paying for it.

Measuring Aggregate Production in a Virtual Economy Using Log Data

Tuukka Lehtiniemi

Virtual worlds typically contain systems of resource allocation, production, and consumption, which are often called virtual economies. The operator of a virtual world clearly has an incentive to monitor the virtual economy, and users and outside observers would benefit from e.g. temporal or cross-economy comparisons. Standard methodology of computing macroeconomic aggregates for virtual economies would allow this kind of analysis, but such method is currently unavailable. This study fills this gap by employing the concepts of national accounting and unique log data from a virtual world. In particular, the focus is on virtual economies where the production of new virtual goods takes place as the users expend inputs to produce predetermined outputs along predetermined production paths. The major MMOGs fall into this category. Previous attempts on measuring the aggregate production of a virtual economy have been based on non-standard method and externally collected data. In virtual economies the operator can collect extensive data automatically, a characteristic feature that should be reflected in any standard accounting scheme. Macroeconomic aggregates for a national economy are computed using the UN System of National Accounts (SNA). It is a standard accounting system, and its probably most quoted outcome is the gross domestic product. The most relevant borderline in SNA lies between the national economy and the rest of the world: domestic production is included, whereas foreign production is not. SNA cannot be directly used in a virtual economy, as the concepts of “domestic” and “foreign” are not applicable. In this study, the concepts and methods of SNA are transferred to a virtual economy context. The relevant distinction in a virtual economy is made between production by the users and the creation of goods by the virtual world code. Application of the concepts of SNA and flow chart analysis result in an aggregate measure called the Gross User Product (GUP), which measures the value of the aggregate output of production activities – of both goods and services – by the users of a virtual economy. In the empirical part of this study, the potential of GUP is demonstrated by measuring it for the virtual economy of EVE Online based on extensive log data collected by the operator. Temporal comparisons are performed after purging the GUP values from the effect of significant deflation using a chained Fischer index. A 50% real growth per user is observed for the first half of the year 2007. The composition of GUP has remained rather stable during this period of significant growth. The concept of GUP is general, and it can be used for quantifying virtual economies other than that in EVE Online on the macro level.

My comments:

This paper is one of fruits of HIIT's research collaboration with CCP Games. In my opinion, Castronova et al. would do well to read this papers' criticism of the notion of calculating GDPs for virtual economies.

Machine Ethics for Gambling in the Metaverse: an "EthiCasino"

Anna Vartapetiance Salmasi, Lee Gillam

Abstract— Online gambling produces a substantial turnover. Unfortunately for potential virtual world gamblers and gambling organizations alike, US law had forced the closure of gambling in the Second Life virtual world. However, an Open Grid Protocol could lead to the provision of off-shore gambling in this virtual world. Aside from legal issues, online gambling generally gives rise to ethical issues relating to prevention of harm. We considered the combined legal and ethical issues, and have proposed and begun to construct and evaluate a system with computational oversight: an ethical advisor. The system is grounded in recent research into Machine Ethics, which may offer insights into other legal and ethical matters, and provides a framework for responsible gambling in our EthiCasino (ethical virtual casino) in Second Life.

Blatant misrepresentation of Entropia Universe and the report.


The comment, "Since Entropia's virtual currency is redeemable for real money, it is essentially gambling. Blackjack seems to be the the closest comparison that comes to my mind.", seems particularly inane and warrantless, and is also a blatant misrepresentation of the conclusions of the report and how the Entropia Universe economy actually works.

I personally wonder whether you have a particular axe to grind or are just completely unfamiliar with Entropia Universe.

You appear to have missed the point entirely about the zero-sum nature of the total number of resources within the game (i.e. balancing considerations), the non-randomness of loot generation, the relative differences in individual avatar in-game skills and equipment, and the fact that resource value within the game is largely driven and determined by player demand and hence its "Markup" price.

Any comparison to the game of Blackjack is highly inane, naive and silly.

Dear anonymous commentator: I

Dear anonymous commentator: I am sure that the experience of playing Entropia Universe is in many ways much richer and more complex than that of playing Blackjack, so I can understand why you (as a player?) might feel that this comparison is unfair. But as a way of summarising the core economic aspect of the game -- investing money against an uncertain but proportionate return that can be improved with skill yet remains below the investment on average -- I do think Blackjack is not far off the mark.

Still, I have only played the game briefly in 2005, and rely mostly on secondary sources, particularly the paper at hand. I would be thrilled to have a more knowledgeable person argue for a different description -- but only when there is a commitment to mutual respect and sincere pursuit of knowledge and understanding. I will not tolerate attacks ad hominem. If you are able to present your arguments politely and clearly, you will find me a very receptive audience.

Gambling or not?

This might not be the correct place to initiate a discussion as the intention here is to leave comments.

Nevertheless, gambling aspects of Entropia Universe (EU) are very controversially discussed within EU’s community. Although, there are many similarities to gambling, the intention of the platform is clearly different.

Gambling per its definition is a system were the result (gain or loss) is only related to luck. Within EU however, gain or loss is not only related to skills (more sustained costs to achieve the same loot with less skills) but also to the ability of the participant to resell looted items with markup.

My personal impression is, that the gambling like nature of EU’s loot gives thrill to the game and should it make more attractive. The heavily right-tailed nature of the loot distribution leads to some interesting discussions on community forums as mentioned in the paper and there are many opinions about the loot system (how might loot be triggered?, how is it implemented?).

The aim of the publication was to reveal the expected payback of mining as this information was not available officially. Please note that many aspects of EU have been unknown and are still unknown. For instance, decay of an item is not explicitly stated and hence unknown to a participant and several methods have been invented by the community to assess it. Mindark intentionally created a black box and its content has to be discovered.

Although, we were able to describe loot using statistical methods, we still don’t know its implementation (only the provider knows and he won’t tell as correctly stated by Vili). It is however of crucial importance to know sustained costs and payback if partaking in EU or other similar universes. As many things have still to be discovered, we can judge only partially the implications of EU’s implemented loot distribution and related economy as Vili requested in his comment. So there is quite some room for further investigations and publications ;)

Virtual Economy in Sheep's Cothing

It is easy to be seduced by the similarities between Virtual and Real economies. The idea of having perfect manufacturing and consumption data on which all types of economic models can be created is somewhat intoxicating i suppose to a theoretical economist. BUT, virtual economies are NOT "real"! and a guy playing Guiter Hero is not a "real" musician. and what I mean by "Real" is quite simple: I will argue that the first medium-based manifestation of an entity is the "real" version, and that when the entity becomes manifested in a second medium, it can no longer be described as "real" or even "the same as the first" manifestation. by "medium" I am referring to a "generalized platform definition". but this is not just a game of semantics. the importance is that just because 2 things appear quite similiar, it does not mean that we extrapolate on that to rationalize or validate all kinds of other observations. and we know this to be true of just about any entity we can think of..
In Virual Economies, the only real laws that the site is bound by having nothing to do with money(ie. Pornography,etc). Rulers of VE's, can and do:
1. revalue currencies at will
2. discriminate
3. subject
4. create inflation, stagnation, etc
5. create scarcity and excess in milliseconds
6. create and destroy vast "wealth" in milliseconds

I do believe that as time goes by, Virtual Economies will have to trend towards rules, laws, and other conventions that keep "real" economies "trust worthy". but in the meantime, do not give in to a false sense of security that comes from working through "real" models that seem to apply well to virtual does make me think however, if you could limit a model to a single Virtual World/Economy, and define all of events that could occur in VW, that could not occur in Real LIfe, and you were able to assign a probability to each of these events, you could I suppose factor in the sum of these probabilities into your real model. think of it as a Virtualized Wiener Process....

the image that just popped into my head is the TV commercial that shows an office worker strolling the photocopier to get his copy. I can model that task pretty simply(basic modal model stuff). but in teh commercial, out of nowhere, at a fantastic speed, comes a fully padded football player and absolutely levels the officeworker. That is precisely the way we need to think about Virtual Economies. Until someone guarantees that Linebackers will not be loose in the offioe, we cannot trust in real world models to describe or forecast Virtual Economies

Ofcourse EU is gambling.

If you disagree it's because of two possibilities;
A - you have a vested interest in the success of EU due to your own investments (time/money)
B - you have no idea