Using Video Games to model Economics

Economics is social science meaning it’s a study of society and our relationships with each other. This means that economists use how we interact in order to create theories about our behaviour. In addition to this, economists use economic models to make predictions on how changes to the economy will impact us. However, Economics is a social science meaning that much of what we learn cannot be tested rigorously unlike the natural sciences. This is because changes to the economy are directly linked to the living standards of someone. Therefore, economists and other social scientists cannot conduct experiments like many natural scientists can where they isolate variables and repeat processes. This greatly limits the amount of work economists can do because it becomes more difficult to link variables together. Additionally, there are financial costs among other costs such as time that makes gathering data a slow process.

In order to help fix this problem, I considered using video games in order to model economics. My reasoning for this is that many games have in game economies and interact with other players, similar to real life. However, with video games there are fewer limits because it isn’t affecting anyone in real life. This means that a more experimental approach can be given to economics. This also means that tasks such as data collection can be automated and because every single transaction would be recorded, the data is completely accurate. This is in contrast to economics where informal sector work is often unaccounted for when calculating the value of economic indicators such as GDP.

I decided to use video games to model a variety of real world variables: population, inflation, economic growth and oil prices. These were selected as they had close analogues in games. However, it was necessary to look at these indicators in different contexts to see which was most accurately modelled. For the first three indicators on the list I looked at the data for Nigeria, India and the USA. Having three countries at different stages of economic development was important to highlight the possible successes and flaws in the modelling process. As the price of crude oil doesn’t vary country to country I opted to look at how well a game could model crude oil prices at different timeframes.

Most attempts in modelling were generally unsuccessful. This was largely due to the video game model failing to emulate events which fluctuate. The reasoning I had for this was that the game often has long periods of stability which is the antithesis of having something that fluctuates unpredictably. This is why it was generally better at modelling long term behaviour which tends to be more stable than short term behaviour. Furthermore, it was more effective at modelling at developed countries due to the same reason – developed countries are economically more stable because of their strong institutions and having generally less corruption leading to behaviour that is easier to model.

Table 2: Incandescent Energy prices over 1 month in Coins (Blue line) by Runescape Wiki

As shown here, the stable prices of the game is unable to match the fluctuations of oil prices on the one month basis.

Table 3: Incandescent Energy prices over 6 months in Coins (Blue line) by Runescape Wiki

Table 4: Oil prices over 5 years in $ (USD) by Bloomberg

As shown by the two graphs the stability in prices is fairly well matched by the game especially the first part.

One other factor that led to the model failing was that the game could be considered more of a ‘free market’ this is because all prices and all transactions are almost exclusively affected by supply and demand. However, in reality there is government involvement which means that market forces are altered and manipulated to ensure a certain objective is achieved. For example, OPEC has been known to manipulate oil prices as it ensures that we have a stable rate of oil production.

Furthermore, the virtual economy lacks features of the real economy such as a government entity. The games I had used had nothing resembling a government entity, however governments are a key feature to the global economy which have a massive influence.

In conclusion, my findings show promising signs as the game was successful at modelling some elements of the real world economy. However, it was limited and couldn’t completely replicate the full features an economy had. On a more positive note, however, there weren’t many games that published economic data meaning the games I had to choose from was a highly limited selection. Therefore, I believe that significant improvements could be made if data from more games was available.

The author is a student of Math and Economics at the London School of Economics and Political Science and can be reached at [email protected]