Keywords

5.1 Introduction: A Game Designer’s Perspective

From the outset, the reader should bear in mind that this paper is written from the perspective of someone who has spent time in the video game industry as a game designer, among other capacities, dealing with many of these issues on a day-to-day basis. Many of the included examples, anecdotes, and assumptions are based on that experience and are intended to describe one perception of the current state of the industry and how it relates to potential best practices.

5.1.1 A Different Set of Goals

It seems quite common these days to enter into an engaging discussion over the merits of gamification (or more likely, the lack thereof). The participants of these discussions seem to often be professionals who are in a business-oriented role in their respective companies. These individuals typically want to find a way to improve some aspect of their company’s business. Occasionally, educators and academics will be lured into the fray, hoping, in a similar fashion, to either improve various things or to evaluate to what extent such improvement is possible. Customers occasionally stumble into these discussions as well, though they tend to have an understandably negative view of what they perceive as malicious manipulation. These conversations often originate from (or occur in response to) talks or articles about (or by) companies utilizing (and sometimes selling) gamification services. This is understandable enough because who better to talk about gamification than people who deal with it on a daily basis?

The various types of participants all have quite valid and worthwhile viewpoints and can certainly contribute positively to the collective conversation, but strangely, the voices that seem to be missing are the traditional game designers. One might think that game designers would be on the forefront of any discussion concerning bringing game mechanics into real-world situations, but they seem to have a reaction similar to that of the players, often finding the attempts at behavior manipulation extremely distasteful. This distaste sometimes goes as far as considering some gamification techniques to be professionally irresponsible or even unethical.

While the video game industry may be young compared to many others, there are still multiple decades of history and tradition that inevitably shape the design tendencies of today’s game designers. Though the more common trends could surely benefit from an outside influence, there are also a great many lessons and a collective wealth of knowledge that traditional game designers could bring to any attempt at gamification. However, these attempts at gamifying business or education are typically designed and implemented from the perspective of the businessman or the academic. These practitioners come to the world of game mechanics because they see the power that games wield over their respective players, and the gamifiers endeavor to bring some of that power into other aspects of life.

5.1.2 An Entertainment Focus

Seemingly unique among those interested in gamification, a traditional game designer’s primary goal is typically the entertainment of the customer. The designer must take certain business-oriented aspects into consideration as well, such as marketability and accessibility. However, designing a great product is a game designer’s primary method of contributing to a company’s success. The standard (and seemingly logical) rationale is that a better game sells more copies, and even though other factors are also in play (marketing budget, for example), a correlation between game quality and sales does exist (at least as measured by the meta-review site MetaCritic’s scores).

The point of this distinction is to highlight the concept that a game designer is primarily interested in creating a positive (or at least entertaining) experience for the player, while seemingly everyone else who might want to leverage the power of game mechanics has ulterior motives. Therefore, the game designer is best equipped to consider things from the perspective of the player and is best able to to shape the mechanics to both elicit the desired outcome and to ensure that doing so provides the best possible experience for the user.

5.2 Gamification

5.2.1 The Gamification of Games

5.2.1.1 Aren’t They Games Already?

While much has been written recently about the advent of gamification, it is interesting that this advent was largely brought on by the wave of success experienced by social game companies, such as Zynga. Previous successes in the video game industry didn’t bring this sort of interest to game mechanics, so why now? The gamification of games was really the lightning rod for all of the attention, but how can you gamify a game? Wasn’t it already a game to begin with?

There were certainly gamification attempts long before the recent wave of social games (frequent flyer miles, for example), but this new breed of games brought about a change in paradigm that radically altered the way both games and gamification are designed and implemented.

5.2.1.1.1 Formalization

First was the notion that there really can be a formula for success. Traditionally, when one would talk of design rules or best practices, it always seemed implicit that such things were relatively general guidelines that one should thoughtfully consider based on the context at hand. These things seemed more like objectives to pursue (or avoid) than a literal blueprint for success. However, with the advent of social games and the meteoric rise in popularity of titles such as Farmville, the idea that there could be a dominant strategy seemed to take hold in the game industry. Scores of imitators began development on their own versions of the most popular social games, attempting to leverage what were perceived as the secrets to success.

Of course, very few games can actually rise to the top of the charts, practically by definition, but the idea that there are certain mechanics which can repeatedly elicit specific (and sometimes unintuitive or even irrational) behavior is likely closer to the truth than most players would like to think. The industry powerhouses have begun leveraging the expertise of different professions (utilizing mathematicians, statisticians, behavioral economists and psychologists, etc.) in an attempt to better define what does and doesn’t work. By highlighting principles that have been shown to work in other contexts, game developers are able to create and rigorously test in-game implementations to find functional correlates to their real-world counterparts.

The (publicly) unstated goal of this is a formalization of certain designs or mechanics that can be used repeatedly, across multiple implementations and contexts, while retaining its effectiveness. This is incredibly desirable for a large game publisher, as it would effectively be a trade secret that would allow for more effective game development with considerably less risk. However, this is obviously complicated by the fact that these designs and mechanics are public (at least to some degree) because players interact with the products. This opens any game studio up to having their designs quickly and ruthlessly copied. Any formalization that hopes to avoid this is likely to be a meta-formalization, guiding the design at a high level (defining the general principles to be used) instead of directing specific implementations (specifying how to use those principles). While formalization at the tactical level seems to result from the various cloning attempts common to the game industry, formalization at the strategic level is likely the true benefit gained from gamification.

5.2.1.1.2 Optimization

Of course, even the best strategic vision can be undone by faulty implementation. One reason that cloning the details within other games isn’t likely to be a highly successful long-term strategy is that the gestalt of a successful game is quite fragile, and even minor changes can have a dramatic effect on the experience. While a certain mechanic might be wonderfully effective in one game, reproducing the mechanic in another game might give significantly different results (even if the cloning attempt encompasses the entire game).

Because the effectiveness of any portion of a game is so specific to the individual implementation, optimization can be incredibly important to a game’s success. Of course, video games, being software products, typically undergo extensive testing, covering a variety of different areas (testing for bugs, fun, usability, etc.). However, the rise of online, web-based games brought to the game industry (popularizing, if not introducing) a philosophy of testing methods from online marketing, most notably characterized by split-testing. In a split-testing scenario, two groups of players would play slightly different versions of the same game while the developers tracked whichever metrics were intended for optimization. A common example (applicable both in games and elsewhere) would be altering the pricing of an item and tracking which price level generates the most revenue.

This concept could take much of the guess-work or intuition out of game design. Instead of arguing over which design decision to make, a team could simply split-test the options and retain the one that produced the best results. With the most popular social games reaching tens of millions of users, the potential (and the potential benefit) for split-testing was immense. A large number of tests could be run simultaneously, evaluating a massive number of options and keeping only the top performers. The overall effect of fine-tuning practically every part of a game could have a dramatic effect on its profitability.

This very data-driven approach almost unquestionably produces results, though there are a number of downsides or counter-arguments. First and foremost is that these tactics need to be guided by an effective strategic vision. Mass split-testing can enable incredibly effective optimization, but knowing what to optimize for isn’t always obvious. For example, optimizing for user acquisition or virality before a game is effective at retaining or monetizing its users could lead to unsustainable operating costs and churning, unprofitably, through the users in the game’s target market.

An additional criticism of optimization-based development is that split-testing only allows the developers to find local maximums. In other words, it can help your team do what it’s doing as effectively as possible, but it won’t tell you if you’re doing the right thing. Split-testing helps a team refine the details, but the high-level decisions remain in the hands of the developers.

5.2.1.2 Games as a Lens

5.2.1.2.1 Transparency

Even if the goal isn’t to create a profitable video game company, examining games can still prove to be a very valuable learning experience. This can be true even if the goals have nothing to do with games at all. Games provide a lens through which one can examine user behavior in a vast array of environments, engaging with countless mechanics, and modified by a multitude of variables. The problem rarely involves a lack of information. Problems more often arise with the interpretation and subsequent action taken (or not taken). These difficulties often stem from the incredible amount of information at hand.

However, it’s one thing to go diving through millions of lines of your own data to reach a conclusion about user behavior, but trying to come to the same conclusion when the project generating the data isn’t your own (and may well be a competitor) is a very different problem. Luckily, when the project in question is a game, we have a great deal of information exposed to us. Anyone who wants to learn how games work has but to play, test, and analyze everything that the game exposes to its players. This may not be complete information, as the reasons and motivations behind certain decisions may not be readily apparent, but it is fairly straightforward to analyze what has been done and (to some degree) the effects of these decisions. We can look at various implementations of a similar mechanic, attempting to decipher the factors and contexts that led to success (or the lack thereof). We can also look at the evolution of mechanics over time, often within the same company’s purview. It is even possible to track how a certain feature or mechanic changes over time within a single game, allowing us some degree of insight as to the effectiveness of various changes and the internal motivations of the development team. Few situations offer the chance to so easily gain this level of insight from other products on the market (especially without the consent, or even the knowledge, of the product owners).

5.2.1.2.2 Observing “Unfiltered” Behavior

One benefit of using games as a method of research is in the prevention of the interference by the experiment itself. Studying human behavior without inadvertently affecting the behavior of those humans being studied can be difficult to avoid at times, but by using a game as an intermediary, this becomes less of an issue. This is especially true if using a commercially available game (ideally in cooperation with its developers). In this way, players who are already playing the game of their own accord could potentially be tracked and studied. In a real-life context, there are innumerable social conventions that govern people’s behavior to a very large extent, and while some of those tendencies likely map to in-game behavior, stepping out of reality removes a layer of potential interference. This can allow us to see behavior without many of the filters that subtly alter our behavior on a near-continuous basis.

An additional benefit is that games aren’t necessarily encumbered by the same expectations (from the player’s perspective) that are present in a reality-based context. In a real-world experiment, participants may not be aware of the specific nature of the study, but they may still be somewhat guarded or inauthentic due to the participation. This behavior may be further exacerbated if they are placed in an unusual situation or required to perform an unfamiliar task. However, within the bounds of a game, unusual situations and unfamiliar tasks are almost expected. Video games are so varied in their contexts and mechanics that players wouldn’t be as likely to unnaturally alter their behavior just because of the uniqueness of the experience. However, it should be noted that caution is required when questioning players directly on how much they enjoyed the experience or their perceived quality of the game. Inquiries of this type can be among the most difficult questions for players to answer objectively, making it far more reliable to establish behavior-based metrics that can be objectively measured as the player interacts with the game (tracking the length of a play session, for example).

5.2.2 Bringing These Lessons into Non-Game Environments

5.2.2.1 The Customer as a Player

It might be difficult at times to equate certain business objectives with game mechanics, and it can even feel uncomfortable or unpleasant to think of your customer as a player, especially when that subject is an internal customer, such as your employees, co-workers, or even your supervisor/manager. Each situation is different and should be considered within its own context, so the designer of a gamification mechanic must decide how transparent the implementation should be. In other words, they must decide how much information about the mechanic should be exposed to the user. For a forum user who earns a higher post-count by participating, it may not matter that the intent behind the post-counting mechanic is to encourage forum users to post more frequently, thereby increasing the amount of content that is generated for free (and if the forum users were informed of this, it would be unlikely to change their behavior substantially). On the other hand, consider the potential fallout if a manager informed the staff that the purpose of their company’s Employee of the Month program was to incite artificial competition thereby causing employees to expend more effort and work longer hours without costing the department more than a wooden plaque once per month.

In truth, certain implementations of the mechanics discussed here could be seen as manipulative or even coercive, and depending on the nature of the business, ethical concerns could come into play. It is beyond the scope of this paper to discuss the issues of user perception and ethics at a useful depth, but as with any business decision, it is important to consider the potential consequences (both direct and indirect). Consider also that individual perspectives can also paint gamification attempts in either a positive or negative light, regardless of the intent of the designer. In the Employee of the Month example above, a more humane (or tactful) manager could also truthfully say that the award allows for the recognition and reward of excellent employees and that such recognition is likely to lead to other rewards such as increases in pay and promotions.

5.2.2.2 The Risk of the Gamified Workplace

The potential human cost (in morale, productivity, turnover, etc.) can make internal gamification attempts a risky proposition. This is a very different environment from a game with millions of players where the developers often accept the risk of running tests that could potentially decrease the quality of the experience for some percentage of their player-base in order to find the better of two options. Though, luckily, it isn’t necessary to split-test across the entire game population. To go even further, it is often the case that a change which makes the game worse for the vast majority of its players can still be seen as “positive” (from the perspective of the developer) if it increases the game’s profitability. This situation is not at all rare in free-to-play games where majority of a game’s revenue might come from a very small percentage of its player-base.

In addition to the potential human costs, there are also the practicalities of hiring that must be considered. In a game with a realistic target market of tens of millions of players, effectively “burning” (ruining the game for them, causing them to quit) a few thousand users might be an acceptable loss if the result is the further optimization of mechanics that affect the rest of the player-base. Essentially, if the cost of acquiring a similar number of users is less than the expected profit from the optimization, then it is a worthwhile loss. However, when considering the workplace, where the target market is the population of people qualified and willing to work in a position for a salary that the company is willing to pay, re-acquiring lost employees can be a very costly and time consuming process. Additionally, the smaller potential sample sizes and the longer testing cycles that would likely be required in a workplace setting would cause the optimization efforts to take longer and will involve far fewer data-points than would be found in a large-scale social game, making the process much more involved and risky. The key point here is that while the dramatic leaps of performance found in video games via optimization might be possible in the workplace, great care must be taken to mitigate the potential risks to the company.

5.3 Relevant Concepts from Behavioral Economics

Traditional economic theory relies on the concept of rationality, the idea that people are generally able to make decisions that are in their best interests (Ariely, 2008). However, there are some situations where people tend to act in ways that seem to be at odds with traditional economic rationality. To a large extent, the field of behavioral economics attempts to examine and explain various situations that seem to cause people to commonly display irrational behavior. Furthermore, these situations that lead to irrational behavior aren’t random but can be engineered for studies and demonstrably predicted to occur in the wild, leading Ariely to describe people as predictably irrational (Ariely, 2008). This ties in closely with many gamification mechanics because many of the typical gamification techniques leverage (intentionally or not) people’s tendency to act irrationally in response to certain situations or stimuli. By deliberately engineering certain elements, a gamification designer can potentially manipulate the behavior of a user, typically in order to achieve a certain business objective. A frank discussion of behavioral economics in terms of intentional user manipulation can, at times, sound malicious or at least distasteful. However, some of these mechanics are practically ubiquitous in today’s society (for example, the practice of ending prices with the digit 9, as in $9.99 instead of $10.00 (Schindler, 2009)), making it useful to be aware of these concepts, even if one has no intention of using them personally.

5.3.1 Relevance

To select the concepts from behavioral economics that were most relevant to the games of today, lists of both these concepts and game mechanics were compared in order to map the motivations of each, attempting to find pairs with corresponding motivational triggers.

It should be noted that the presence of these concepts within a game should not necessarily imply that it was included to intentionally manipulate the players. Game design has evolved over time, with some conventions falling out of favor and others becoming increasingly entrenched. One could logically assume that the design conventions that achieved the desired results were retained while others were discarded. This evolution likely selected for many of the listed concepts even in cases where the designers in question were entirely unaware of the principles they themselves were making use of.

Listed below are nine concepts, further broken into three categories that mapped closely with frequently used types of game mechanics. Each concept is described briefly and accompanied by one or more general examples of a game mechanic utilizing similar motivations. Additionally, each of these is followed by one or more potential implementations of the mechanic in a workplace or non-game setting. This is in no way intended to be a complete list. The range of contexts and mechanics encompassed within the game industry is so vast that a comprehensive list could potentially include practically any cognitive bias. The selections were made based on the author’s perception of their relevance and frequency of use (and of course, limited to some extent by the scope of this work), but there are certainly excellent arguments to be made for the inclusion of others.

5.3.2 Categorization

The concepts resulting from the selection process mentioned above were broken into five different categories based on the typical purpose or motivations behind their use. Most selections could incorporate multiple motivations and many types of potential implementations, so it would be justifiable to place the concepts in multiple categories. However, for the sake of length and clarity, an attempt was made to group them only in the category that seemed to fit best, based on the in-game uses of the concepts. A listing of the categories and a brief description follow.

Encouraging Engagement: The concepts listed here attempt to keep the player participating and returning to the game through either building obligation or preventing the sense of obligation from falling away. These mechanics can serve to increase a number of retention metrics, some effectively acting in a manner similar to a retargeting marketing campaign.

Guiding Action: The concepts listed in this category attempt to keep the player moving forward in the game, often by making clear what action should be taken, preventing decision fatigue from setting in. Many mechanics based on these concepts are meant to boost retention and session times, drawing players in, and imparting significance to in-game action and elements.

Identity Investment: These concepts help to build a player’s sense of identity via the game. The players invest their thoughts and efforts into creating and achieving in a game, thereby developing some sense of ownership over the elements that they have interacted with. This can lead players to develop bonds with those elements, making them reluctant to abandon them, even if they would prefer not to play the game.

5.3.3 Concepts, Mechanics, Non-Game Uses, and Common Errors

5.3.3.1 Encouraging Engagement

5.3.3.1.1 Loss Aversion

Loss aversion is the tendency of the player to avoid losses or even chances at a loss. This tendency can be shown as irrationally strong when the aversion to loss overpowers a desire to acquire a disproportionate amount of gain (Kahneman & Tversky, 1984). This appears in a great many places within modern games and is often manifested in the form of the potential for lost opportunity.

5.3.3.1.1.1 In-Game Use

As mentioned, lost opportunity is one of the more common occurrences in games. Examples are often some form of chance at receiving an in-game reward. In context, this could be killing an enemy, completing a challenge, or playing another match versus a competitor. Other examples could include the loss of a reward and the materials invested based on a scheduling mechanic (harvesting a crop that the player has planted), losing the chance to purchase an item (either via a time-limited offer or a random shop mechanic), or even the risk of losing position on a leaderboard.

5.3.3.1.1.2 Non-Game Use

In a service or product, the risk of losing access to certain content or features can be used to promote action. This often occurs with demos and trial periods. Once a user has become accustomed to or reliant on a service or product, they could be much more likely to convert, either buying the product or subscribing to the service. Additionally, once they’re signed up, the potential loss of the service makes cancelling a more difficult decision and increases the switching costs, making it harder for them to replace it with a competitor. Additionally, limited-time offers give the user an opportunity (usually for savings, but occasionally for special access) while simultaneously threatening to remove the opportunity if the user doesn’t take advantage of it quickly.

Another way in which loss aversion appears is via the social ties that are formed in certain situations, such as the friends made while taking part in an organization, the relationships with coworkers made at a job, or even the customers encountered in the process of running a business. The potential loss of community and relationships can be a major factor in a user’s or an employee’s decision to leave a service or a job.

5.3.3.1.1.3 Common Errors

Common implementation errors include a reluctance to give users sufficient access to the service or product for them to value it enough to be concerned about losing it. This is evident in products and services without sufficient demonstrations or trial offerings. This also is apparent in cases where there are special offers, but the offers either don’t seem attractive enough for the removal of the offer to be considered a loss or the users don’t believe that the offer is really limited. An example might be a potential customer responding, “No thanks, I’ll just wait until the next sale next week.” In such cases, the users don’t really have anything to lose. Another potential problem occurs when a business owner is reluctant to penalize users sufficiently enough to create a risk of loss. Many services with free offers likely give too much away in the interest of building a user base.

5.3.3.1.2 Maintaining Intrinsic Focus

This concept contends that the addition of a tangible reward to an activity previously performed for its own enjoyment replaces that enjoyment with a form of payment (Heyman & Ariely, 2004). This essentially turns an intrinsic motivation into an extrinsic one (Pink, 2009). A person might play a game because they find it fun, but if they were paid to play the same game, the original enjoyment would be largely replaced by the prospect of payment. This is very troublesome in video games because game designs frequently call for players to perform similar, repetitive actions in order to extract more playtime out of a given amount of content. It is a formidable design challenge to encourage a player to do this without the repetition beginning to feel like work.

5.3.3.1.2.1 In-Game Use

This concept is often implemented through the player being forced to repeatedly defeat very similar enemies or challenges for the purpose of driving progression in the game (often in order to gain experience points or currency). Games often attempt to disrupt (or at least mask) the repetitive nature by adding contextual variations to the encounters. These variations often take the form of altering the visual depiction of the enemies (while their appearance may be different, the encounters are often essentially identical) or having non-player characters send the player on a mission or quest while providing a snippet of narrative as a justification. However, when effectively done, these variations can help prevent the feeling that the players are “grinding” or “farming” content. Another manifestation of this is allowing players to purchase highly sought after items with in-game currency earned though a regular and predictable method. Instead of playing “for fun,” the players are encouraged to “grind” for their currency if they want the prized item. Instead, providing a method to obtain the item via a form of randomness could help maintain the intrinsic quality of the experience.

One common criticism of microtransactions is the fear of the players’ ability to spend money becoming more important than their skill or time invested into the game, effectively moving the motivator from intrinsic to extrinsic. This can be avoided to some extent by only allowing purchased bonuses to modify what the player has earned in-game, allowing the base of the effectiveness to remain in the realm of the intrinsic.

5.3.3.1.2.2 Non-Game Use

This concept can occur in any situation where a user is being rewarded for performing a desired behavior. A general defense would be only giving non-monetary rewards for the behavior, though even so, this can appear in something as simple as rewarding forum users for the number of posts they make (granting special access based on post-count, for example). Once users become accustomed to receiving a reward for a behavior, they may not continue the behavior once the reward is removed.

Following the in-game example, instead of allowing a user/employee to earn a reward through the repeated performance of an additional task, allow them a chance to win the reward. An example of this might be entering the user into a lottery each time an item is recycled. Even though the final reward might be monetary, being rewarded with a chance to win is psychologically removed from actual payment (even if the economic value of each entry is easily calculated).

5.3.3.1.2.3 Common Errors

In online communities, a commonly quoted ratio, 100-10-1 attempts to illustrate the ratio of users who consume (100 %), interact with (10 %), and create (1 %) content (Wilson, 2011). In an effort to encourage the behavior exhibited by the 1 %, community owners may attempt to incentivize that behavior in some way. However, those users already making up the 1 % are already exhibiting the desired behavior, so the encouragement attempts to draw the other users into participating as well. This is has the potential to drive away the users performing the activity with a high degree of quality and integrity because they felt intrinsically rewarded for doing so, replacing them with users who attempt to maximize the rewards they can extract (not an unreasonable expectation since the reward was what enticed them to participate). In short, it is replacing an altruistic behavior with one only done for payment.

5.3.3.1.3 Pseudocertainty

Pseudocertainty effect is present when someone makes a decision that treats an uncertain outcome as certain (Tversky & Kahneman, 1986). One implication of this concept is that normally risk-averse people can become risk-seeking when given the opportunity to gain a small chance at a large reward. Additionally, the value that they tend to place on the chance is often far greater than its actual economic value (Tversky & Kahneman, 1979).

5.3.3.1.3.1 In-Game Use

This is often used in games by giving players near constant chances at a reward. Random rewards are common in games, where you may defeat hundreds of enemies during a play session with each of them having a chance to reward the player with items or currency. Typically, there is a table of potential rewards, each with a certain chance to drop from a given encounter. Of course, the quality of these rewards is inversely proportional to the chance of the item appearing. Players are often given practically worthless items, while others are both incredibly rare and valuable. This can turn each encounter into a lottery or a pull of the slot machine handle, essentially setting up a potentially addictive intermittent reward schedule (Lee, Sturmey, & Fields, 2007).

This is also a very common mechanic in analog games such as collectable card games, where it typically takes the form of small packs of cards that must be purchased to play. Each pack has a certain number of cards, but the buyer has no way of knowing what cards are inside. Most are very common and not at all desirable, at least after forming a basic, usable collection. However, some are quite rare, making each pack a small gamble.

5.3.3.1.3.2 Non-Game Use

As in the recycling lottery example listed above, this type of mechanic can be used to incentivize any trackable behavior. For each performance of a desired behavior, enter the user into a periodic lottery. The prize doesn’t need to be incredibly valuable to trigger the effect; it just needs to be large enough to be meaningful to the winning individual. The goal is to essentially enable a certain degree of uncertainty so that the users/employees will work to increase their chances. The fact that both the chances and the increases may be quite small is at the root of this cognitive bias. People will work to enhance their chances to an extent that is far out of proportion to the economically expected gain.

5.3.3.1.3.3 Common Errors

As with the intrinsic/extrinsic errors, enabling people to directly earn their rewards can cause problems. When “earning” a reward, people are much more likely to be cognizant of exactly how much gain they are getting from each unit of work provided. It is the introduction of uncertainty that short-circuits the logical analysis. Additionally, a reward system based on chance could still potentially run afoul of logical analysis if its rule set is too transparent or explained in a way that makes the motivational intent obvious.

5.3.3.2 Guiding Action

5.3.3.2.1 The Paradox of Choice

This is the concept that people almost universally see an increase in the number of available choices as a positive change even though making a choice becomes much more difficult as the number of choices increases. People tend to be fairly comfortable making a choice when presented with a relatively small number of options (between three different items, for example), but once that number increases beyond a certain limit, it is much easier to simply choose not to choose (Iyengar & Lepper, 2000).

5.3.3.2.1.1 In-Game Use

Even if unintentional, we can see increasingly limited choice in some games as they have progressed over time. Modern games often give the player a highly pre-defined, scripted experience, seemingly in imitation of a cinematic/theatrical experience. One may quibble with the design sensibilities of such games, but one upside to this is that players are actively pulled though the progression of the game. They have few meaningful choices to make, so they encounter little decision fatigue. The path forward is always clear and unambiguous, so less content is wasted (either by players quitting before finishing the game or by bypassing it, as could happen in a non-linear game).

Tutorials that guide the player though the beginning of a game are incredibly common, to the point of being expected. In open-world games, conspicuously marked non-player characters send players on simple quests, helping to ensure that there is never a question about what one should do next. Even achievements are structured to guide the player on to the next challenge. The less time a player spends deciding what to do, the less likely it is that the decision to stop playing comes up.

5.3.3.2.1.2 Non-Game Use

Similarly, in today’s popular web services and social networks, it is common to have a very simple functional tutorial that walks a new user through setting up and using the service. The goal would be to help the player progress though the service’s social loop as far as possible before having to make any difficult decisions. Similarly, some services use checklists and scores to guide new users along the preferred paths. Even when not rigidly enforced by a tutorial process, if a checklist is readily available, following it can often serve as the path of least resistance that a user can default to if lost or unsure of what to do next.

Sometimes a service or product may be quite complex, making this simplifying process seem untenable. However, the rigid tutorial path is really a mechanic that exposes complexity gradually. If complexity is actually required, it may be beneficial to only allow a user to access the various features one at a time until some level of knowledge or proficiency is reached. Additionally, presenting the user with only a small selection of possibilities is applicable almost any time a product or service is offered for sale. (In-game item shops could often benefit from learning this lesson!)

5.3.3.2.1.3 Common Errors

The most common errors here are simply dropping an inexperienced user into a product or shop with a dizzying array of options. In a shop (physical or online), it is advisable to offer as few items of a given type as possible and to ensure that the differences between the items are clear and unambiguous. Similarly, complex tools may need simplified views or robust tutorials to prevent novice users from being overwhelmed.

5.3.3.2.2 Scarcity/Urgency

When an item is seen as having limited availability or when the time to act is similarly limited, people irrationally value the object in question (Cialdini, 2006).

5.3.3.2.2.1 In-Game Use

Scarcity often appears in games via rarity of items. The items may be theoretically infinite in supply, but if there is only an infinitesimal chance of the item appearing at a given chance, it can effectively be quite rare and valuable. Additionally, layers of scarcity can rest atop one another as a means of increasing tension. An example could be a particular enemy who is very rare who has a similarly low chance of rewarding the player with a special item when defeated.

Scarcity of opportunity should also be considered. A player who is only allowed to undertake a certain number of missions per day is likely to undertake more missions on average than if there were no limit at all (assuming that the arbitrary limit is appropriate). Similarly, if a player is planting crops but only has a limited number of slots in which crops can be placed, the player is very likely to plant as many plants as possible. Urgency will also drive a player to attempt to capitalize on as many opportunities as possible in a given time period.

Multiplayer games are of special note here. When competing with other players, it is possible to make practically all resources scarce and their accumulation urgent. When players perceive that competition exists (even when the only competition is indirect), they often are driven to accumulate resources far beyond what they would otherwise deem necessary or desirable in order to prevent the competition from benefiting.

5.3.3.2.2.2 Non-Game Use

It is possible to introduce scarcity or urgency through the use of limited editions or with a limited-time offer. As an additional bonus, limited editions often have the ability to sell at a premium price. As with games, scarcity of opportunity can be introduced by only allowing users to take a certain number of actions per day. If a social network wanted to encourage messaging, it might limit the number of posts one could make per day, thus ensuring that users would want to make the most of their scarce posting slots. A new service might launch as invite-only to create a scarcity of availability. Then it could offer a limited number of invites to its user-base, who would be encouraged by their scarcity to invite others.

Introducing competition among customers or employees may seem counterintuitive, and it should likely be done with a certain amount of caution. However, a common way of doing so is via contests where people compete with each other, likely performing a valuable activity, with the most effective participant winning a meaningful award. The effect can be amplified with tight time limits, ensuring that the participants are acting under the effects of urgency.

5.3.3.2.2.3 Common Errors

Introducing scarcity or urgency in a way that appears excessively contrived or outright deceptive seems fairly common in these attempts, and this can backfire. On the other hand, the more common error may be not imposing any scarcity or urgency at all. Prospective buyers often want to be left to “just think about it for a little while.” However, if a user feels no impetus to act upon first seeing an offer, the chances are that additional time will move the subject farther and farther from the transaction.

5.3.3.2.3 Variable Reinforcement Schedules

A variable reinforcement schedule rewards a user for a certain action or behavior, but does so in an irregular pattern. This causes the user to know that the action occasionally leads to a reward, but it prevents the user from knowing what, if any, pattern is behind the decision. This makes the reward mechanism largely a black box to the user and introduces a situation where the only variable that the user can control is the frequency of the action (naturally leading to a dramatic increase in the desire to perform the action) (Lee et al., 2007).

5.3.3.2.3.1 In-Game Use

This mechanic is incredibly common in the game industry, existing in some form in almost every game that allows players to receive rewards. One of the most direct implementations is the typical reward received from defeating an enemy. These normally have some random element associated with the reward they offer, making each kill a mini-lottery. There are a myriad of other implementations using this mechanic, being present in practically any feature that relies on randomly generated results. Even an individual attack within a combat can display these properties, as the success of an attack is often determined by chance (sometimes with a small chance to do extra damage, which is often accompanied by reward audio/visual feedback).

It should be noted that the “randomness” mentioned is typically more of a weighted distribution of outcomes and can, at times, be entirely predetermined. However, the key issue isn’t the randomness. The important issue is that the player remains unaware of the exact mechanism being used (especially if the mechanic is not randomness).

5.3.3.2.3.2 Non-Game Use

In a workplace setting, this could manifest itself in the form of bonuses or positive feedback distributed on some schedule that isn’t obvious to the employees. This type of reinforcement also presents itself naturally in some settings, such as user comments or feedback. Additionally, a product or service could create reminders or notifications that are sent to users at certain intervals, either with encouraging feedback or rewards.

5.3.3.2.3.3 Common Errors

A common error here would be a lack of planning for the introduction of a variable reinforcement schedule to any aspect of a business. Stability and consistency are seen as generally valuable traits in an environment, so workplaces seem eager to eliminate anything seen as unpredictable. However valuable they may be, consistency and predictability don’t seem like tremendous motivators.

5.3.3.3 Identity Investment

5.3.3.3.1 Commitment

Commitment is the concept that people tend to want to fulfill agreements that they have made in order to avoid the cognitive dissonance that occurs when breaking these commitments (Cialdini, 2006). The utility of this can be illustrated by asking users for a commitment to perform an action at a later time (which is easier to agree to than asking for immediate compliance, likely due to hyperbolic discounting) and then comparing compliance rates when the agreed upon time comes with the immediate compliance rates of a control group. This tendency to avoidance cognitive dissonance is further manifested when asking users to comply with a further request, increasingly the likelihood of compliance with each successive request.

5.3.3.3.1.1 In-Game Use

One of the most direct implementations of commitment is via the use of appointment mechanics recently popularized with social games. In these mechanics, a player might plant a crop or begin construction on a building, with the understanding that the growth or construction will complete at a certain time in the future. These times typically start quite low, at only a few minutes, encouraging the player to return to the game for frequent, short sessions. This can be very habit-forming due to the repetitive nature of the activities and because each time a player returns, the fulfillment of another commitment is registered, reinforcing the player’s identity as a player of the game in question. Eventually, the timers extend, keeping the repetitiveness from becoming too extreme while at the same time, offering the player a feeling of progression, as now the full benefit of the mechanic can be achieved in only one session per day. However, the optimum session frequency is often kept at one day in order to maintain the habit by allowing it to conveniently fit within one’s daily routine.

Additionally, multiplayer games make use of this though periodic events that required the simultaneous cooperation of multiple players to complete. This forces the players to self-organize, essentially making social commitments to each other. The commitment elicited increases as the required player skill and progression for the successful completion of the event increases due to the increased difficulty of finding replacement participants.

It’s also worth mentioning that any tiered progression with rewards for advancement (even if the rewards are only feedback-based) can act to support commitment by helping to establish that the player is a person who is successful at progression in the game. Though it sounds a bit circular, with each successful step in a game’s progression, the player is tacitly committing to continue playing (as with the increased compliance with multiple requests mentioned above).

5.3.3.3.1.2 Non-Game Use

This can be leveraged in non-game setting in a similar fashion by allowing people to commit themselves to a desired action at a future date, incentivizing them to do so with discounts or initial periods of reduced cost. Additionally, allowing the user to set certain appointments within the scope of the product can assist in retention. For example, a user might be able to choose which day a monthly charge should occur on, causing the arrival of the charge to be an action the user specifically scheduled. A user may also choose how or when notifications and reminders are sent, making their interruption more acceptable.

In terms of commitment increases, a user could be introduced to a product or service in a manner that is very low risk and potentially highly beneficial. For example, the user could be prompted to sign up to a newsletter containing valuable information. This is both quick and free for the user, making it an easy first commitment to make. Additional requests for compliance could gradually increase the cost of the commitment.

Enabling commitments between users or between employees of the company and the users can add the element of social commitment to a product or service as well. A company could offer free training webinars, instructing users on how to get the most out of their product. These webinars, while free, might only occur periodically and could require prior registration, creating a scheduling commitment. To further add social commitment, these events could be invitation only, with each existing customer receiving a limited number of invitations for their friends. This would allow them to give something valuable to their friends, who would then have a reciprocal commitment to attend.

5.3.3.3.1.3 Common Errors

The most common errors that businesses make in this area consist of only making one request from potential users and making that request at the very beginning of the relationship. At this point, it is incredibly easy for a user to end the relationship before it even begins. As a first step in avoiding this problem, businesses often entice users with trial periods, which is certainly a step in the right direction, but that carries a different psychological effect. Labeling an offer as a “trial” implies that the user will decide if they want the product at a later time, giving them an option to opt out without the corresponding cognitive dissonance. However, enabling the users to commit now and pay later can trigger the commitment effect even if they aren’t contractually obligated to continue the service after the introductory period.

5.3.3.3.2 IKEA Effect

The IKEA effect is Dan Ariely’s term for the concept that one’s valuation of an item is disproportionately increased by the personal labor one has invested into its creation (Ariely, 2010). By investing one’s time and effort creating something (and likely expending a considerable amount of cognitive energy making decisions during this process), people tend to develop an emotional attachment to the creation, causing them to value it much more than would be justifiable by the creation’s actual economic value.

5.3.3.3.2.1 In-Game Use

This is highly relevant in many types of games, but none more so than persistent, online games. The persistence of online games along with the ability to share the experience with other people adds a sense of permanence to a player’s creative efforts that typically exceeds the importance of creations found in other types of games.

It is common in many modern games to be able to create an avatar through which the game world is experienced. Avatar creation and customization can, at times, be incredibly complex and in-depth. While the process of doing so may be a slight barrier to entry, the resulting creation is one that the player can feel a great deal of ownership over. Additionally, many games are based around the advancement and progression of this avatar. Players may literally invest thousands (or even tens of thousands of hours in extreme cases) of hours of their time into the progression of their in-game avatars. This can obviously become a significant commitment on the part of the player, holding a great deal of personal value, even if the economic value is negligible.

Customizations and progress apart from an avatar are also relevant here. The ability of players to customize their environment is present in many games, giving the player some control over the world that their avatar inhabits. This can take the form of player housing or even simply the ability to place decorative flowers in a field. Some players are willing to spend vast amounts of time and effort customizing their environment to their preferences.

In-game pets are also worth mentioning here. Non-player characters within a game that take the form of a player’s pet (often in the form traditional pet, such as a cat or a dog, but sometimes being very non-traditional and context based) can elicit strong bonds from a player. This is naturally amplified as the opportunities to customize and advance the pet increase.

5.3.3.3.2.2 Non-Game Use

In non-game settings, the obvious example is to follow the IKEA model of requiring some amount of choice in the selection of items and some further amount of assembly in order to garner an investment of time and effort from the user. However, non-game settings can also use the concept of online persistence and avatar customization. Almost any online product or service could potentially benefit from allowing users and their creations to be visible to one another in some way. Once visibility is possible, users have an extra incentive to spend time crafting their online presence and acquiring or developing any items that other users may encounter. Additionally, methods to advance or progress this online presence can be incorporated (unlocking additional features or access, for example).

The value of customizing an environment can easily carry over to a physical setting as well. If employees are given latitude or even encouraged to heavily customize their workspace, it could seem considerably more valuable to them than a very generic space. Additionally, if customization options exist in a product or service and are utilized by the users, this could raise significant switching costs, deterring users from swapping to a competitor.

Incorporating virtual pets into non-game settings may not be commonplace, but humans’ tendency to become quickly attached to pets could make such attempts worthwhile. It’s also worth noting that some offices have introduced pro-pet policies to great effect.

5.3.3.3.2.3 Common Errors

The most common error here is simply not providing sufficient opportunities for users to invest time and effort into customization or progression. However, when offering options of this nature, it is useful to consider whether or not the options are actually meaningful. A myriad of superficial decisions that don’t make a substantial impact on a user’s experience could easily drive people away.

Mass customization isn’t for everyone, and some users won’t want to invest the time and effort to do so until they are more familiar with a product or service. It can be beneficial to provide a solid set of default options and a few sets of pre-defined options to select from. Also, emphasizing that the customizations can be easily altered at a later point can remove some of the pressure from new users.

5.3.3.3.3 Sunk Costs

The concept of sunk costs implies that even though money or effort invested in the past has no bearing on future decisions, these past investments still carry significant weight when considering those decisions (Arkes & Blumer, 1985). This may overlap with the IKEA effect to some degree but differs in the motivations behind the concept, with sunk costs creating more of a feeling of obligation than of value.

5.3.3.3.3.1 In-Game Use

Many elements within games can trigger the sunk cost effect. As with the IKEA effect, these elements are most pronounced within persistent, online games for similar reasons. Players often spend great amounts of time and effort on in-game progression and achievements, creating a sense of obligation to continue playing the game even once the enjoyment has vanished. Players can feel like all of their previous time and energy has been wasted if they abandon their current game for a new one. Additionally, these games often encourage, or even require, the creation of social bonds with other players in order to progress past a certain point. Players can find it incredibly difficult to leave their friends and communities behind, often remaining an active user of a game simply to maintain the social ties. These aspects of persistent, online games lead to the potential for them to have very long lives and low churn-rates.

An additional manifestation of sunk costs within games is the development of mastery. In addition to progression through content, players may also have a significant feeling of progression their personal skill level. Some games can be quite complex, requiring a great amount of time and effort to simply learn to play the game at a high level. Once a player is accustomed to possessing a certain level of mastery in a particular game, the idea of beginning a new game, having to start from scratch and go through the long learning process again, can be incredibly daunting.

5.3.3.3.3.2 Non-Game Use

One way to enable the sunk cost effect to users is to again introduce some aspect of advancement or progression. However, unlike the IKEA effect, the point isn’t to enable the feeling of ownership, it is to increase the potential switching costs involved in changing to a competitor. The more that the user feels that time and effort would be wasted upon discontinuing the service the less likely it will be. Building the opportunity for deep levels of progression can aid in this, as can gradually increasing the complexity. By acclimating to the increasing complexity over time, the learning curve is kept low enough that it doesn’t push members away, but once it reaches a certain point, the idea of learning to replicate all of the desired functionality on an entirely new service can be overwhelming. In some cases, it can even be difficult to find out if all of the desired results are possible with another service. Additionally, enabling a community within the product or service can aid in this much in the same way as it does in-game.

An equivalent to in-game progression could be the storing or creation of content within a service or product, for example, uploading years of photos or contact information to a particular service. This service could dramatically increase switching costs by making this content difficult to move to a new service.

5.3.3.3.3.3 Common Errors

Common error here would be not providing the opportunity for progression or community ties of sufficient depth to trigger the feelings of obligation and wasted effort when considering leaving. It seems common for companies to add various progression or community features at a very superficial level, possibly in order to simply expand the feature-set on offer, without enough depth for players to find them meaningful.

5.4 Common Pitfalls

5.4.1 Implementation Errors

Errors of implementation are typically less serious than their counterparts listed below. They are often noticed with the gamification attempt in question simply isn’t working. There could be countless potential errors here depending on the specific type of implementation being attempted. However, some common problems could include:

  • Rewards that aren’t meaningful to the users

  • Obviously false or manipulative scarcity/urgency

  • Excessive or irrelevant choices

  • Allowing users to earn rewards (turning intrinsic behavior extrinsic)

  • Aggressive or unwanted use of social sharing features

Regardless of the error, implementation problems can typically be identified and improved via split-testing and optimization. Occasionally, an error here can significantly damage the performance or reputation of company, but that’s more likely an error stemming from insufficient testing than any specific gamification issue gone awry.

5.4.2 Design Errors

Design errors, on the other hand, can be particularly insidious, often stemming from a high-level problem with the company’s strategy. As with many problems in business, a common source is the prioritization of short-term, financial results over the long-term sustainability of the business or project.

These errors often take the form of incentivizing the wrong behavior. Consider the simple (and not uncommon) example of a website wanting to increase the number of page-views in order to increase ad revenue. They might make an attempt to introduce a gamification mechanic that rewards the writers on the number of page views garnered. However, this plan may hinge on the leadership assuming that higher quality articles get more page-views, which could be an entirely faulty assumption. Writers, eagerly responding to the incentives, will likely attempt a number of strategies to increase their performance, obviously keeping the successful ones. Some successful ways of increasing views might include publishing a large volume of short, low-quality articles, using multiple pages or slideshows when unnecessary, or using excessively inflammatory or controversial titles in order to draw clicks from users who ultimately leave disappointed. These tactics may drive page views, but they may not take the web site in the direction that its management had envisioned.

The danger of starting with an erroneous gamification design is that your plan just might succeed. With the potential for optimization and split-testing to be so incredibly effective, if you want to increase a certain metric, you must be very sure that it is in your best interest to do so. Otherwise, your optimization efforts could lead to a destructive cycle where you could literally optimize your company out of business!

5.4.3 Assumption of Universal Applicability

A final problem that is somewhat different in character than the others is the assumption of universal applicability. Simply stated, this is the failure to realize that people can be very different and can react to a given situation in very different ways. It is very likely that some (if not all) of the previously discussed concepts and implementations elicit very different results from different types of people. However, it seems quite common (especially in the video game industry) for features (and even entire games, on occasion) to be copied wholesale. Just because one company is presumably achieving positive results with one feature in one game with one set of players does not mean that this success will automatically readily transfer to a different game by a different company with a different set of players. Though it may seem obvious, it bears mentioning based on the feature sets of many games on the market today.

5.5 Conclusion

5.5.1 Implications

Discussed above are a number of potentially powerful concepts, and typical businesses and game studios are likely using some of them on a day-to-day basis in both their products and their management processes without even realizing it. They turn out to be, as often as not, less a new tool in the management toolbox and more of an explanation about how your existing tools actually work. Even though gamification has gotten a great deal of attention recently and always seems to be a target for controversy, our lives are all gamified to some degree. Gamification in some form is likely to be an increasingly important factor in the direction and the formation of the world around us, so it behooves us to try and understand what’s really happening inside the black box.

From a very general perspective, it would be useful for practically everyone to become familiar with the basic techniques of gamification, if for no other reason than for the awareness of the many different ways that outside parties might affect one’s behavior. From the position of someone hoping to make use of gamification, whether in business or academia, developing an understanding of how these techniques work is key to making the best use of their effects, regardless of what the intended result may be.

Taking this a bit further, anyone in a position that reasonably calls for them to affect change in another person’s behavior or knowledge stands to benefit from the well-considered use of some of the techniques mentioned in this paper. However, when trying to initiate a behavior modification, caution should always be exercised, as unintended (and often unfortunate) consequences are a frequent occurrence. The key to the effective design and implementation of these mechanics lies in understanding what exactly is being incentivized and what the long-term results will be.

5.5.2 Future Work

An area that currently seems lacking in the knowledge base of both gamification and behavioral economics is a thorough analysis of the affects of each on various segments of the population. With the advent of modern data-mining tools, segmentation itself is an increasingly useful concept, and combining this with a large-scale database of gamification and behavioral economics data could provide incredibly valuable insights. As a simple example, we might find the optimal number of choices that could be given to each segment of the population in a particular situation, which could potentially lead to improvements in practically all online shops. In addition to the typical demographic and economic population segments, it could also be useful to study these effects on segments based on personality type, market preferences, or even segments selected via data-mining for concentration of effect.

Additionally, as discussed earlier in this paper, when compared to the possibilities in consumer gamification, there are considerable difficulties in developing testing techniques that can be as effective in many enterprise applications, which are often geared towards some type of internal customer. It could be very beneficial to these enterprise attempts to explore how these difficulties could be overcome, most notably those of scale and test cycle times.