This chapter will showcase a step-by-step example of how to analyze an existing economic system with Machinations (Dormans, n.d.) and utilize it to showcase how a given economic system could be changed to test how it would behave under new conditions. After that this chapter will explain how to evaluate these tests and how one could use the evaluations of the tests to build upon them to get a certain result.
6.1. Creating Machination (Dormans, n.d.) prototypes from existing games
To showcase how one can break down the economic system of a game and transform it into a Machinations (Dormans, n.d.) prototype three different games have been chosen to indicate that the approach of doing so can be translated to almost every economic system found in games. The created prototypes will utilize fake numbers and will not show a precise example of how economic interactions in these games are concluded, but rather work as an indicator of how these systems work on a basic level. That being said it should be possible to add correct numbers to systems created by this approach to create systems that allow to be utilized as balancing tools for almost every economic system created in a game.
The first game that will be utilized as an example of how to break down a game to its economic structure will be the Borderlands (Gearbox Software) series. To do this, it is important to consider how in-game economic systems work as explained in the former chapter ‘The financial economy in games’. Based on this it can be said that games should be evaluated from the micro-economic perspective of the players since all forms of macro-economic behaviour can be changed by the designer without limitations (unless a game intent to create a whole macro-economic system that affects all other game systems like reward-systems or ‘NPC’-merchants). Taking this into account, it becomes apparent that the micro-economic system of almost every game is concerned with how a player accesses currency and how the player spends currency. So to analyze how the economic system if a game works one starts by listing all the ways a player earns and loses currency or objects that can be turned into currency.
In the case of the Borderlands (Gearbox Software) series this would be the constant stream of currency gain from looting dead enemies and chests, the income generated from quest rewards and the currency generated by selling unneeded objects. The amounts of value lost by buying new equipment and the ammunition/health needed to fulfil certain tasks and the penalty the player has to pay in case of death. This makes a prototype showing a simplified version of the economic system of the Borderlands (Gearbox Software) series look like Figure 1, which showcases the constant flow of income, items and currency, the player has while playing the game. This prototype visualizes ways a player can lose and earn currency and how selling and buying items affects the players’ current amount for currency and items. It also features a randomly generated skill level that effects how often or how much currency the player loses in certain situations and increases the income and costs depending on the player’s level. This system could already be utilized to balance how much income the player should generate and how much the player should have to spend on each level.
Figure 1
Figure 2
The economic system of Divinity: Original Sin 2 (Larina Studios, 2017) has similar ways for the player to obtain currency and goods. There is a constant stream of income based on quest rewards and loot. The main way of forcing the player to spend valuables is by creating a situation in which the player is forced to utilize consumable items. The other main approach this game uses to remove currency from the player is by selling desirable items that need to be replaced after a certain amount of progress. Other than this the game also features an increased income and cost curve based on game progression and allows the player to reduce the number of consumables needed based on the player’s skill level. The big difference between Divinity: Original Sin 2 (Larina Studios, 2017) and the Borderlands (Gearbox Software) series is that enemies won’t respawn. This creates a situation in which the player is unable to generate an infinite amount of currency and is therefore forced to decide between a maximum of buyable items that can be obtained by the total amount of currency the player is able to obtain. The NPC merchants themselves have regenerating amounts of currency and items, which allows the player to always sell unneeded goods to them. Another difference is that Divinity: Original Sin 2 (Larina Studios, 2017) has the option to invest skill points into a bartering ability that reduces all prices for items NPC merchants offer and increases the amount of currency these merchants give the player for goods. It is also possible to steal goods and currency from each NPC once. Handing the player access to an additional, but also not infinite, option for generating income. All of this can be seen in Figure 3.
At this point, an important disclaimer must be made: Divinity: Original Sin 2 (Larina Studios, 2017) is a game with a very definitive ending. If it were not for that this type of economic system is very likely to run out of resources it can hand to the player. This game avoids doing so by leading players towards different areas and offers new locations constantly since unless new developments are introduced to an area or the player has new options old areas are no longer usable for the player since all tasks and other ways of generating game progress are already exhausted.
Figure 3
Figure 4
In The Elder Scrolls V: Skyrim (Bethesda Game Studios, The Elder Scrolls V: Skyrim, 2011) the economic system also offers a similar approach to the other two games. The player has a constant stream of income that mainly consists of quest rewards and the loot found in the game. The income and costs increase with game progression. The player has also access to a bartering skill similar to Divinity: Original Sin 2 (Larina Studios, 2017) allows the player to reduce the cost of items bought from NPC merchants and increase the income generated from selling goods. NPC-merchants also have regenerating amounts of currency and items, in a way that is comparable to Divinity: Original Sin 2 (Larina Studios, 2017). But unlike Divinity: Original Sin 2 (Larina Studios, 2017) The Elder Scrolls V: Skyrim (Bethesda Game Studios, The Elder Scrolls V: Skyrim, 2011) features respawning enemies in a similar way to the Borderlands (Gearbox Software) series. This allows for an infinite amount of currency generation for the player. The main approaches to remove currency from the player, besides the trade disadvantage from buying and selling goods to an NPC merchant, are the player consuming items that had to be found or bought before and the possibility that the player gets robbed by a thief or bandit. Both of these events can be avoided or at least reduced if the player is skilled enough. A visualization of this economic system can be found in Figure 5.
Figure 5
Figure 6
6.2. Comparing economic structures of games with each other
If the three prototypes of the economic systems described the ‘Creating Machination47 (Dormans, n.d.) prototypes from existing games’ section of this work are compared to each other, by tracing the values of ‘Player Level’, ‘PlayerMoney’ and ‘PlayerItems’ for multiple, which can be seen in Figure 7, Figure 8 and Figure 9 common behaviours become apparent. ‘Player Level’ increases in all three examples for each iteration in the same way, which is the logical behaviour, since ‘Player Level’ is utilized to measure game progression, which has a fixed development in the created prototypes. ‘PlayerMoney’ builds itself up and then decreases in chunks throughout the testing period, but the average amount increases over time. This happens because the simulated players spend chunks of their earned currency to purchase items or lose chunks from specific events like death. The overall increase in currency relates to the scaling amount of income being out of sync with the scaling amount of required spending. ‘PlayerItems’ behave very similar to ‘PlayerMoney’ but the overall build-up ratio and strength of the item decreasing chunks is more stable. None of the fewer features ‘PlayerItems’ have a tendency of increasing over time. This is likely because of the value difference between one average item and one instance of the chosen currency.
If this behaviour is compared to how a player interacts with the economic system of a game, it becomes possible to connect in-game activities like looting or experience gain with the option of doing tasks that reward the player. Doing so allows the player to grow in power resulting in accomplishing the same tasks faster or accessing new areas with higher amounts of loot. The effect that this has on the economic system of a game is represented in the prototypes shown in Figure 1, Figure 3 and Figure 5. This allows a Game Designer to theoretically balance a game economy in a way that leads to players earning and spending an equal amount of currency and therefore a stable economic system. The problem with that is, that it is almost impossible to predict a player’s spending behaviour over a long period of time unless it is in some way forced by the game. This explains why balancing a game’s economic system is a good way of tightening the control of the overall income rate of a player. But it is not enough to control the amount of currency a player has access to at any given moment while playing a game. If this form of control is desirable for a game it becomes apparent, that the economic system as a whole needs to be adjusted in a way that allows the system to either force or incentivize options of gaining or spending currency. These options don’t necessarily need to speed up or slow down player progress but they should be something that is desirable to a player and improve the overall play experience compared to a system without these elements.
More precise information about the results of the test with the simplified economic system of the games The Elder Scrolls V: Skyrim (Bethesda Game Studios, The Elder Scrolls V: Skyrim, 2011), the Borderlands (Gearbox Software) series and Divinity: Original Sin 2 (Larina Studios, 2017) can be found at https://philippstenger.com/wp-content/uploads/2019/02/BaseSystems_ExcelDocumentation.7z
Figure 7
Figure 8
Figure 9
6.3. Prototype description
The following sections of this work will concern themself with the testing of different approaches with the goal of adjusting the ratio in which the total amount of player possessions increases over time and how a desired/fixed amount of player possessions could be achieved for an extended period of time. These approaches are not straightforward solutions, but rather a showcase of how someone can start to utilize the tool Machinations (Dormans, n.d.) to direct an economic system towards a specific economic goal and how to test the system for achieving desired effects and what unexpected problems it may create.
6.3.1. The goals of the prototypes
As already stated the objective for the prototypes will be the creation of approaches that are able to stabilize an economic system in a way that the average of valuables in possession of each player becomes more consistent and more predictable. By evaluating the data found in https://philippstenger.com/wp-content/uploads/2019/02/BaseSystems_ExcelDocumentation.7z two values and one behaviour can be viewed as the defining point of a stable economic system.
The first of which is the mean of currency and items a player has during its playthrough. The prototypes created in this work will be limited to the mean of currency and goods a player has within a limited amount of game progression. If the mean is similar between multiple players the expected result in the game is that players are likely to obtain similar amounts of currency and items and spend similar amounts of currency and items in similar periods of time. The strength with which the amount of currency and items differs from the mean at each step of progression can be calculated by utilizing the sample variance of all steps of progression. It is possible, that different spending behaviours like saving up the currency and then purchasing multiple items at once can heavily influence this way of evaluating the stability of the economic system. This means that the mean value of currency and items shows the average amount of currency and items a player obtains and spends in certain game progression intervals and with that demonstrates the average spending behaviour of players. This makes this value an effective method to determine if players use their earned currency and items effectively and if the game offers enough possibilities and necessities to spend the earned currency or to utilize collected items. This can be determined based on having similar means across multiple playthroughs, while also considering the number of extreme cases. An economic system with similar means and a low amount of extreme cases can be considered stable and an economic system with fluctuating means and a large number of extreme cases can be considered unstable.
The second important value to consider is the maximum of currency and goods a player is able to obtain during one playthrough. The prototypes created in this work will be limited to the maximum amount of currency and goods a player is able to obtain within a limited amount of game progression. If the maximum amount of currency and goods in a game is high it likely translates to the fact that the game offers too much currency and items to the player, has not enough or too cheap ways to spend the earned currency or only offers uninteresting or unnecessary ways and situations spend or use currency and items. Another possibility is that the economic system of a game features increasing amounts of income depending on the players’ progression in which case the maximum amount of currency and items at a certain level of progression must be measured. By measuring the maximum amount of currency and items players have the same problem that was mentioned for the mean also exist. Because of that, the spending behaviours of players will also heavily affect the results of this test, since players that save up more currency and spend it in larger chunks are not necessarily able to obtain more during the same level of progression. As a consequence of that comparing the mean of currency and items to the maximum value shows the maximum amount of currency and items a player is able to obtain during a certain game progression interval and with that demonstrates the maximum income potential of a player. That being said the maximum amount of currency is a good value to measure how often and in which ratio related to the players’ income the need or wish to spend obtained currency or items arises and if the game has enough ways to fulfil that need or wish. This can be determined based on having similar maximums across multiple playthroughs, while also showcasing the number of extreme cases. An economic system with similar maximums and a low amount of extreme cases can then be considered stable and an economic system with fluctuating maximums and a large number of extreme cases can then be considered unstable and therefore unpredictable.
The behaviour that is important for the definition of a stable economic system is the overall trend of mean and maximum currency and items on specific levels of progression. This showcases how much currency and items a player has at each point of progression during a playthrough compared to all the other points of progression during a playthrough. For the prototypes created in this work, only a limited amount of points of progression during a playthrough are listed. Evaluating the behaviour of the in this way collected data allows a translation to in-game expectations. An example of this is the higher or lower income and spending between different progression levels, which are confirmed if the total amount of currency and items increases or falls but the percentage of how much of the gained currency and items are spent by the player stays the same. Another example is that the balance between income and spending may vary between different levels of progression. This is usually readable if a player obtains currency and items faster or slower on different progression levels compared to others, even so, the spending habits are the same, or vice versa. By measuring these behaviours spending habits of players must also be taken into account evaluating the maximum and mean of the currency gives no direct feedback about the possibility that a player collects currency and items for multiple progression levels and spends it all at once. In this case, the trend for all affected levels of progression is measured in comparison to each other, but the overall behaviour can still be compared across multiple playthroughs. This behaviour can be utilized to determine the overall economic growth the player entity receives between different levels of progression if economic growth exists within one level of progression or how the income and spending work on different levels of progression compared to each other. This can be explained based on the increase or fall of the overall income and spendings a player has on different levels of progression, the increase or fall of income and spending within one level of progression or a percentage-based change between income and spending. An overall increase in income and spending on different levels of progression showcases the economic growth of the player entity, while an overall fall shows the opposite. As long as the percentage between income and spending stays the same the economic system can still be considered stable. Having the overall income and spendings rise or fall between progression levels usually translates to desired or undesired imbalances within the game, an example of this would be enemies that offer more loot than other enemies on the same progression level. Another way this could happen is if there are ways for a player to improve something that is not tracked by the progression level. An example of this would be a progression level that tracks a player’s in-game level, but it is also possible to obtain equipment that allows completing the tasks in one progression level faster. If the percentage between income and spending changes slightly within one progression level it is usually an indicator that the game guides the player by separating different objectives to not overwhelm the player with too many new things at once, but still allowing the players to choose what they what to obtain first. Having missions that are focused on obtaining a special skill for a character compared to missions that are focused on obtaining currency or items are a good example of this. Large differences between income and spending for multiple levels of progression are most likely an indicator of an unstable economic system.
6.3.2. The approach for the creation of the economy prototypes
All prototypes are created utilizing the tool Machinations (Dormans, n.d.) and represent simplified versions of the economic systems of the games Divinity: Original Sin 2 (Larina Studios, 2017) the Borderlands (Gearbox Software) series and The Elder Scrolls V: Skyrim (Bethesda Game Studios, The Elder Scrolls V: Skyrim, 2011). These prototypes will utilize fake numbers to showcase how the systems work but do not represent the actual values of the chosen economic systems. That being said it is possible for the systems to be translated into the readers’ own games by inserting the actual values which will allow us to predict and balance one’s in-game economy.
After the explanation of how a game is reduced to a simplified version of its economic system, this work will go more in-depth about how an economic system can be adjusted to fulfil a specific goal. This will be done by changing the economic system of The Elder Scrolls V: Skyrim (Bethesda Game Studios, The Elder Scrolls V: Skyrim, 2011) with the goal of creating a higher level of economic stability. To achieve this, multiple approaches directed towards that goal will be created and evaluated. Promising approaches will then be improved over multiple iterations, while unsuccessful approaches will be dismissed. After a number of iterations, the last promising approaches will be compared to each other and evaluated based on their reliability, use cases, and potential problems. To evaluate these prototypes the data values of how much currency and items the player owns over a period of time will be collected in a sample size of 100 for each approach. The data generated this way is then compared between different approaches to determine how the stability of the economic system as whole changes and how extreme cases are affected. Stability is defined as similar values for mean, median and a low variance across multiple playthroughs. Extreme cases are defined by their strength, which refers to the total amount of currency or items at any point, the higher the value the stronger the extreme case. The important factors to evaluate the extreme cases for the approaches are their strength and how often extreme cases occur.
This whole process will be documented and explained in a way that allows utilizing this method for other games.
6.3.3. Prototype limitations
The prototypes created for this work have several shortcomings. The first of which is that the duration a prototype is evaluated is limited to a certain progression level. This allows assuming if a system has the expected results or not but does not confirm if problems are avoided on a later progression level. This can result in problems if a game adds more levels of progress later on.
Another limitation is, that all prototypes are strictly focused on the economic perspective of the system. It is important to understand that good solutions for an economically balanced system can contradict themselves with the desired effects and reasons an economic system is added to the game in the first place. This makes it necessary to understand why an approach can work for some games but not for others.
The last and most problematic limitation is, that these prototypes are focused on creating approaches that adjust the economic system of a game based on probability-based player decisions. But real players can be different. If a player discovers a dominant or at least a good, strategy for achieving economic growth many others will follow. This results in general economic behaviour that is more focused on successful strategies than a focus on all strategies. In addition to that, it is likely that players come up with ways to interact with the game that were not anticipated. These new ways are not documented within the prototypes and therefore neither tested nor evaluated. Based on this it can be said that it is almost impossible to predict the behaviour of a single player within a large number of players and it is possible for that single player to influence the others. This can result in the majority of players behaving in an unpredicted way.
Despite these limitations, these prototypes are a good way to get a first impression of the workings of a game’s economic system and give designers a good option to balance it. But they can’t guarantee that the system works outside of the tested progression area, must be evaluated early to exclude approaches that won’t align with the design goals of the economic system and they don’t replace playtesting but rather give an extra layer to test the feedback received from playtesting in a larger scale, but with less reliable results. If a designer keeps these shortcomings in mind these prototypes can become a great tool to make better games.
