[In the final installment of his series on Gamification Dynamics, Badgeville's Tony Ventrice puts the concept of flow under the microscope, sharing new research that provides a new window into the popular concept -- as well as examining what aspects of art appreciation translate well to games. The full series includes the original framing article as well as three prior examinations of dynamics: part 1, part 2, and part 3.]
When I was working on iPhone games, I spent a good deal of my time reading user reviews of virtually every successful game to grace the App Store. One thing stood out time and time again, and that was the word "addictive." It seemed to be the highest compliment imaginable. Granted, the audience wasn't the most erudite, but why this word? Isn't addiction a bad thing?
Once you played these games, the language started to make more sense. These "addictive" games had a way of completely absorbing your attention, they were both challenging and simple and had a way of preventing you from setting them down until your battery died.
What the customers were calling addiction is a common phenomenon in successful games and tends to go by the name of "flow" in the industry.
The term "flow" can be attributed to Mihaly Csikszentmihalyi, a researcher who built a career on the topic before video games had even been invented. In an interview, Csikszentmihalyi described flow as:
Being completely involved in an activity for its own sake. The ego falls away. Time flies. Every action, movement, and thought follows inevitably from the previous one, like playing jazz. Your whole being is involved, and you're using your skills to the utmost.
In his writings Csikszentmihalyi typically mentions the following prerequisites to flow:
- Intrinsically rewarding
- Clear unobstructed goals
- Immediate feedback
- Balance between ability level and challenge
To a game designer, these should not be surprising. In fact, these are pretty much textbook instructions to building any type of engagement, from a video game to teaching tricks to a dog.
They're so obvious, in fact, that most discussions of flow tend to focus on only the fourth point: Balance between ability level and challenge.
If you've ever seen flow diagramed, it probably looked like this:
A student named Jenova Chen -- who later went on to found Journey developer Thatgamecompany -- attracted a good deal of attention with his master's thesis, Flow in games, and an accompanying game, appropriately named Flow. Chen's focus was on dynamic difficulty adjustment, exploring the concept of player-controlled adjustment of game difficulty.
The two problems I have with defining flow as something that happens when difficulty is balanced is that it ignores flow that occurs in situations that have no difficulty curve, and it draws the conversation into the realm of general engagement, away from the hyper-engagement that the term flow was meant to discuss.
How does one explain the flow-inducing success of activities such as The Sims or Vegas slot machines -- activities that have no difficulty at all? The realization should be that "difficulty" is only one way to approach a more fundamental factor -- an important detail that may years of assumption may be causing us to overlook.
What if Csikszentmihalyi wasn't completely correct when he described flow as being a test of skill? His early subjects were artists performing tasks that would certainly seem difficult to an observer with no advanced talent in the field of painting.
Daniel Levitin, an accomplished music industry veteran, psychologist and neuroscientist, proposes in his book, This is Your Brain on Music, that it takes 10,000 hours of practice to master anything, be it chess, basketball, painting, writing or playing the guitar. This is a commonly-cited fact popularized by Malcom Gladwell in his book Outliers.
Is it unreasonable to suppose that the experienced artists Csikszentmihalyi was observing had already been through countless hours of practice and that they did not consider their craft to be difficult at all? After enough practice, mastered actions are so automatic and effortless as to be nearly infallible. How likely is an experienced pianist to hit the wrong key, a professional basketball player to mis-dribble, or a professional painter to fail to anticipate how two colors will blend?
I would propose that, for masters (and even experienced aspirants on their way). the little details are not challenging, they're automatic; the brain has moved beyond technical challenge and onto a meta-problem. In the case of artists, the meta-problem is creating self-expression; in the case of athletes, the meta-problem is outplaying the competition. The brain is simultaneously engaged in two levels of constant activity.
- One part of the brain is managing a stream of learned automatic choices.
- Another part of the brain is focused on a stream of subjective, creative choices.
In any example of flow from games to art, to athletics, to writing, both aspects appear to be essential: the creative choices keep the experience pleasurable and fresh, while the undercurrent of automatic choices compels absolute focus and prevents the brain from wandering.
Corroboration for this model comes in the form of research into "choking" under pressure. Researchers have proven that the spectacular failure that occurs when an expert does fail can be traced to a change in focus from meta-objective to the automatic, implying that not only do these two levels of thought exist but that successful experts expend all of their conscious effort focused on the former.
Within the flow-research community there also appears to be some doubt with Csikszentmihalyi's balanced challenge-skill requirement. Researchers recognize that there might be attentional ambiguity when addressing the concepts of challenge and skill.
Implementing Automatic Learned Choices
There are two ways a game can expect to achieve a foundation of automatic learned choices:
- Build on the backs of previous games
- Build on innate human skills
Previous games. In this case, the game implicitly requires a sophisticated set of skills learned from past games, such as a new FPS that assumes hundreds of hours spent playing previous shooters.
All you need to do to demonstrate the expected skillset of a modern FPS is hand Modern Warfare to someone who's never played a game in the 3D era. The significant amount of investment should become clear as the new player struggles to even move in straight line and navigate obstacles.
Innate human skills. In this case, the game relies on skills most humans have acquired unknowingly through thousands of hours of simply being alive. Anyone who's played Where's Waldo was probably, even at a young age, immediately engrossed in spotting that little striped hat and jacket like a pro.
Anyone who's played The Sims probably already had drag and drop mastered, and intuitively knew what to do with a sleepy or hungry Sim. The same lack of training applies with activities like the face-comparing website Hot or Not or answering questions on Yahoo! Answers.
Implementing Subjective Creative Choices
To discuss creative choices, I'd like to start by dissecting the evolution of a flow-inducing classic.
I can remember my introduction to Bejeweled on my roommate's computer in college. My first thought was: This is stupid. My second thought was: What just happened to the last two hours? My third thought was: This is stupid.
That third thought broke me out of my trance and forced me to face what had just happened.
One thing that seemed obvious was Bejeweled's primary objective is to never let you have a thought that isn't dedicated to the game. The game employs a constant, endlessly repeatable decision loop of simple choices.
The other thing was Bejeweled is terribly redundant. At first, there is a creative challenge to optimize your score without letting the blocks reach a failure state, but eventually you realize how to avoid failure entirely and the game literally does become endless.
This is not to say that the original Bejeweled was flawed -- even if there were no sequels and clones (and yes, Bejeweled is itself a clone), Bejeweled would still rank as one of the most successful casual games of all time. But Bejeweled did evolve, and that evolution is telling of both the nature of creative choices and a potential additional requirement of flow.
Opportunities for Creativity
If you compare that early form of Bejeweled I played back in 2001 with its imitators and progeny (Jewel Quest, Bejeweled Blitz, etc), you'll find a number of new additions. You'll see things like bombs, star gems, hypercubes, quest modes, irregular boards, gradually increasing numbers of gem colors, etc. The games typically start out just as simple as the original, but the new mechanics increase the meta-objectives until the game becomes interesting again.
I suspect that what happened is, over time, the basic play of Bejeweled became automatic to the game's core audience. What was once creative became learned and a new layer of creativity was needed -- a new layer of maximizing scores and responding to hazards. Effectively, what was needed was room for growth.
If you'll remember, there are at least four ways people can measure personal growth (learning, order, challenges overcome, and connections). In the case of Bejeweled, the higher growth objective is challenges overcome. The reason for the assumption that flow = balanced difficulty in games comes from the fact that, in most games, flow has traditionally been obtained with challenges overcome as the primary form of growth.
But flow doesn't have to rely on increasing challenges. Facebook is an activity conducive to flow, but Facebook supports no sense of difficulty. Instead, Facebook contains growth in the form of connections and learning -- reinforcing bonds between people and learning from the shared discoveries of others.
Angry Birds takes a different approach to growth in flow. While Bejeweled addresses the need for novelty by increasing challenge through a system of growing complexity, Angry Birds addresses the issue with new content.
Periodically, new levels with new enemies and new birds are released. New content utilizes three growth forms: learning (how simple new mechanics function within the known system), overcoming challenges (solving spatial puzzles), and order (all those new levels need to be completed).
Unlike Bejeweled, which "feels" like the same puzzle over and over again, Angry Birds comes in discreet levels that need to be completed and checked off from the levels list. Even as one level becomes increasingly similar to the last, there is still a compulsive desire to finish the list.
FarmVille demonstrated games can cause flow in a wider demographic of people than was previously believed possible. The difference between FarmVille and many other games is, unsurprisingly, the source of its flow.
FarmVille's flow relies almost entirely on the sense of growth as order -- the order of cleaning, collecting, completing sets, and earning new decorations.
Slot Machines are a common example of flow and present perhaps the toughest test for any definition of flow.
Sure, slot machines can keep the user occupied with an inundation of buttons, spinning pictures, flashing lights and loud noises, but what about the requirements of subjective decisions and growth opportunity?
To address these I'm going to propose that, if we were to step into the mind of a habitual slot machine player, we would discover there is a perception that decisions are being made. If you're willing to follow me for a moment, I think it's necessary to quickly address what a decision is.
At its simplest level, a decision is:
A problem statement, followed by a prediction, and, finally, a determination.
Let's take a few examples:
Decision 1: What would you like for lunch?
To answer this, the first thing your brain does is create a list of possible options: the Mexican place down the street, the sandwich place next door, the Singaporean place around the corner. Then your brain predicts your expected pleasure at each establishment. Perhaps it also predicts the impact on your wallet, time spent, ambiance and any other possible concerns. Finally, your brain arrives at a choice -- a determination of predicted optimal experience.
Decision 2: Put a quarter in this slot machine, or walk away?
Your brain evaluates the history of previous pulls, predicts the likelihood of success on the next pull. Also considered are your current funds, how bad you "need" to win, how "close" you got last time, and how "deserving" you are. All considerations taken, the brain makes a determination of the likelihood of winning and bets accordingly (bet 1, bet 5, 12-ways to play, etc.).
You might point out that the slot machine problem statement and predictions are irrational, but I'd suggest that rationality is not a prerequisite to decision-making or flow.
The growth of a slot machine is a little more nebulous to define. It may be that there is no sense of growth -- that the potential for financial gain is enough -- or it may be that slot machines contain at least some perversion of order, the implication that behind all the apparent randomness a system exists that only needs to be observed to be learned.
If you're like me, it might bother you that growth is a topic here, considering it was already covered previously in its own section. The reason for re-addressing growth is this: growth is a requirement for flow, but flow is not a requirement for growth. Without growth, the subjective choices of flow will eventually cease to be interesting. Without flow, a game can still engage a user (although perhaps not as compulsorily).
Three Aspects of Flow
- a stream of learned automatic choices
- a stream of subjective, creative choices
- a source of growth
Looking at the three aspects above, it should be clear that not any experience can be "gamified" to include flow. In fact, the requirements are rather limiting, particularly the two endless streams of interaction. Most activities seeking game mechanics are not as interactive as games. A few websites that seem to have flow include Facebook and YouTube. What makes these sites special?
Facebook and YouTube are two of the biggest content sites on the internet. They achieve flow through binge content consumption. On these sites the stream of automatic choices are clicking and viewing. The stream of subjective choices are deciding whether to like the current content, possibly typing out a knee-jerk reaction, and then deciding what to view next.
With limitless content and endless loops of rapid-succession view/respond transactions, these sites fulfill the basic requirements of flow. A constant novelty of content offers an opportunity for growth in the form of learning and possibly connections. A possible implication to be taken from these sites seems to be that, to maintain a necessary level of content volume, user-generated content is almost essential. Other sites that have flow include Q&A sites like Yahoo! Answers and Quora, and commenting/social bookmarking sites like Digg, Fark, and Reddit.
Another possible source of flow is in the act of creating the content itself. Yet content creation possibly faces the same "expert" limitation as creating art or music -- in most cases not everyone is qualified to create content. If the audience is narrow and qualified, this is not an issue. Wikipedia has managed to maintain flow within a small community (although apparently it runs the risk of being bogged down by process).
For the more general population, Facebook is an example of a site where creating content is almost as easy as consuming it. A review site like Yelp also has little obstruction to obtaining content-creation flow, at least until a user runs out of restaurants she's eaten at.
If anything can explain the "magic" of games, I'd say it's flow. While flow is not unique to games, games bring flow to the masses. Perhaps Guitar Hero illustrates this better than anything. Most people, for better or worse, will never learn to play a musical instrument well enough to reach the meta-level of creative expression.
Guitar Hero took the aspiration of musical flow and cut out the thousands of hours needed to reach it. A player can pick up the plastic guitar and be transported directly into a wonderful world of flow. It's not the same flow, not even close, but to someone who hasn't tasted the real thing, slipping into a trance, doubling down on star power and setting a high score while the music harmoniously blares is pretty sweet.
The Fun of Art
The Question of Art
In the comments of previous entries in this series, the question of art came up. What makes looking at paintings fun? What makes listening to music fun?
More than anything I've discussed so far, I feel this is the most sacrosanct. Art is commonly considered a "personal" experience. Art says something to the soul, speaks to our deepest sense of self and identity and is generally regarded as safe from conclusive analysis.
In explaining how art is fun, I can only go off of my own interpretation of what art is and what it means.
I believe art is a combination of the following aspects of fun:
- Intellectual Stimulation
Sensation describes the way things look, sound, feel, smell and taste. As you remember, sensation was an aspect of fun I discounted in the process of making my list. While it is certainly valid, it is not an aspect I found practical to gamification. You don't need a game designer to make a pleasing sensory experience -- you need an artist (not to say they couldn't be the same person).
Order was an aspect of growth. I believe it relates to art in the sense that art can have style. Not everyone appreciates or even notices it, but for those that do, identifying and analyzing style is an exercise of testing/comparing new instances against an implicit set of rules. A sense of harmony can come from witnessing a consistent demonstration of style.
The recognition of a new style (or the evolution of an existing style) involves building a new order -- a new model of implicit rules.
Intellectual Stimulation describes the act of consciously deciphering meaning from art. While not all art has explicit meaning, some does. The exploration and debate of this meaning is both an exercise in finding order (validating personal opinions reinforced by the artistic expression of others) and something of a challenge (figuring out what the art means through deciphering clues or following a line of reasoning) -- both identified as aspects of growth.
Identity defines art in terms of individual and collective experiences. As an individual, artistic preferences identify who you are; they speak to what sensations and emotions you value, and even your sense of order or intellectualism. As part of a group, a shared artistic experience that "moves" you and others in a similar manner can strengthen social connections and validate your sensational preferences, measures of order and/or intellectualism.
This marks the last installment in my investigation into the aspects of what makes games fun. The objective was not to demand all games contain all possible aspects of fun, only to inform the discussion and make the pursuit of distilling and designing fun more productive. I hope you've found this journey as thought-provoking as I have.
If you would like to catch up with the previous installments of these series, you can find them linked below: