Grip: The Biology of Compulsion
Tuesday, 03 March 2009
What makes you come back to the game for “one more try” or “just a little longer”? Once again, it can be tied back to the pleasure centre (nucleus accumbens), as we saw with the enjoyment of all games.
You may have noticed Raph Koster and others linking to a Cambridge University study of the neurobiology of gambling showing that the part of the brain involved in reward – the pleasure centre – lights up when we nearly win, as well as when we win. Interestingly, the researchers report that subjects report this experience negatively, even though the pleasure centre is being stimulated. But of course, even though this may be a negative experience subjectively, most subjects who experienced a “near miss” continued to play on. The researchers note that this behaviour happens in both games of skill and games of chance.
I call this phenomena of compulsion in play Grip, and consider it to be a complimentary behaviour to Csikszentmihalyi's Flow, which I deconstructed in neurobiological terms the other week. If Flow is the constant and steady supply of the “reward protein” dopamine from the pleasure centre associated with a period of intense focus, then Grip occurs as a team-effort between the pleasure centre and the decision centre (orbit-frontal cortex), two parts of the brain that are very closely linked. The decision centre generates rewards (dopamine from the pleasure centre) when we make good decisions, and thus encourages us to learn good strategies and behaviours.
A team of Canadian researchers in 2003 reported that the reward system was activated in anticipation of uncertain rewards, which they suggest promotes the learning of better predictors of reward, and the new study is effectively a special case of this observation, showing that the degree of reward (the amount of dopamine released) peaks in the case of a “near miss”. This for me is an expected result, but it is still a highly significant paper not only for confirming this element of the research, but also for confirming that the response is the same in both challenges of skill and in games of chance (Caillois' agon and alea, respectively).This also tallies nicely with Thomas Malaby's conception of play as a disposition toward the indeterminate.
Almost all gamers understand the experience of Grip, although they do not use that term. It's what makes you give the game “one more try”, and why you find yourself still playing that cRPG at three in the morning, despite your intent to stop hours ago. In the recent straw poll on failure in videogames, a common response was that what motivated players to try again was their assessment of being close to success. Grip, which is the activation of the pleasure centre in anticipation of indeterminate reward, can happen in at least four distinct ways.
In games of what Nicole Lazzaro calls “Hard Fun”, Grip occurs when you can almost overcome the challenge. The game frustrates you, slaps you down in failure, but because your decision centre assesses that you could have beaten it you keep playing. So the fight of the fight-or-flight mechanism riles you up and makes you angry (norepinephrine, and epinephrine), while the triggering of the pleasure centre in response to what you perceive as a “near miss” (whether or not you were close to winning or not, as players can often be deceived as to how close they are to victory) provides the Grip to compel you to continue fighting. As I have mentioned before, despite the fact that many games industry professionals consider this to be precisely what games are about, this is a minority approach to play – 20% of players recognise that they enjoy this “angry fun”, while over 40% strongly dislike it.
Closely related to angry Grip is what happens in a game of chance, which connects with the Cambridge paper: when the player comes close to winning, it encourages them to try again. A sense of agency (the paper talks of decisions, but agency need not be based on decision, as in the case of dice) seems to be required for this form of Grip to connect.
This also applies to reward schedules based upon chance, such as the random treasure tables in a cRPG like Diablo (which was discussed briefly last week in the piece on the top cRPG franchises). The mechanism is exactly the same, but the level of frustration does not rise to combative anger as in the previous case. Rather, while the individual becomes a little angry (a small release of norepinephrine), the experience may register more as mild sadness than as anger (likely a change in the neurotransmitter serotonin). Because we recognise that we don't have the direct capacity to influence chance, this kind of Grip isn't combative, it merely compels us to try again in the face of a “near miss”.
Distinct from the anger-motivated forms of Grip, we can also experience this push towards continuing to play when we are in the grip of a reward schedule of the kind that constitutes the central progression structures of cRPGs and MMORPGs (again, discussed previously in the context of the top cRPG franchises). The circumstances here are slightly different as what is involved here is not the “near miss” effect we saw previously, but rather the promise of assured future reward. The decision centre seems to be the primary driver here – it knows you will get to the reward, and thus helps push players on towards that next reward.
The classic example is levelling up – you level up your character, gain new abilities, and perhaps see the promise of what you will get at the next level, and this motivates you to push on to achieve it. If every level your character gains gives you this anticipation of an even greater reward at a future level, the power of habitual Grip is maximal, and players can find it especially difficult to stop. (This seems to be especially true in the case of Goal-oriented players, those who strongly express the Judging axis in Myers-Briggs, or the Guardian temperament).
There is little or no uncertainty involved here – you know you will level up if you keep grinding, the only indeterminacy is how long it will take to get there. For some reason, this is enough to trigger Grip, the brain's response in the face of indeterminate reward. It can be even more compelling among players with highly developed decision centres (which I recently suggested correlates with the Rational temperament) who seem to get sucked into the question of determining the optimal manner of levelling up. Since good decisions stimulate the pleasure centre, the process of determining an optimal course of action helps players maintain interest in grinding – which helps explain why some (but by no means all) cRPG players lose interest when the optimal solution to levelling remains the same, thus denying the pleasure of devising a new optimal strategy.
However, it would be incorrect to assume that this strategic angle is a requirement. In this regard, an interesting comparison can be made between levelling up in a cRPG and playing old fashioned slot machines, penny falls or pachinko (a Japanese gambling game, pictured above - essentially a cross between Bagatelle and a slot machine). The feeling of continuing to play because "a win is coming" can be found here, and with greater uncertainty since despite the player's conviction it is still possible they may not get a payout before they run out of coins or balls. Penny falls machines (common in British seaside arcades) are even closer to the cRPG reward schedules, as prizes such as small stuffed toys ride along with the slowly-advancing tide of coins – players know they will get the prize eventually, but generally misjudge how many coins it will take to get there.
A fourth and final way that we may experience Grip is in the desire to extend an experience of Flow – of which the most commonly manifested form is “one more game” of Tetris, or some other experience that can produce a Flow state (an experience of intense focus within a goal framework, but with the focus upon the experience and not the achievement of the goal). The indeterminacy here seems to relate to a perceived degree of performance – so the player begins a new game in the belief they can do slightly (or greatly) better in the next game.
Since the experience of Flow is intrinsically pleasurable – indeed, Csikszentmihalyi calls it “optimal experience” it is natural that we want to experience more of it. This compelling Grip happens almost entirely absent of anger, and as such appeals to a very broad spectrum of players. The strength of the compulsion is significantly less than in angry or frustrated Grip, and considerably weaker than habitual Grip too, but it is enough to encourage us to play again in the future. You could argue that this was the healthiest form of Grip, but that doesn't mean that players under its influence don't end up playing far longer than they intend to some of the time.
Grip is the term I am proposing for the various mechanisms involving the pleasure centre (nucleus accumbens) and the decision centre (orbito-frontal cortex) that compel players to continue playing a game in the pursuit of indeterminate reward. The immediate experience of Grip need not be pleasant – in angry and frustrated Grip it is perceived as negative – but when it happens it produces compulsive behaviour in which it is difficult for players to let go of the pursuit of the next reward, whether victory in the face of a difficult challenge, the big score in gambling, the next level in a cRPG, or simply the continued pursuit of the intrinsic pleasure of the state of Flow.
I welcome your perspective on Grip in the comments.
It's been a while since I read it, but this formulation really reminds me of this article:
Now, its not surprising that a formulation of grip should sound similar to behavioural psych reward schedules, but have you considered a correlation? I suppose without empirical data you can't strictly correlate, but only compare...maybe someone academic has done it?
Posted by: zenBen | Tuesday, 03 March 2009 at 13:39
Thanks for this article. As an angry/compelling "gripper", I have little understanding of the motivation of MMORPG players who are constantly grinding.
"which helps explain why some (but by no means all) cRPG players lose interest when the optimal solution to levelling remains the same, thus denying the pleasure of devising a new optimal strategy."
This explaination helps give me some grasp of their motivation. :)
Posted by: organic io | Tuesday, 03 March 2009 at 22:32
Reward schedules, indeed. Also look to Skinner for the concept of 'free-feeding weight' and controls ensuring indeterminacy and near wins.
Concepts such as grip may be new to gamers and coining them, a source of egoboo and rep rewards. On the other hand, any good fiction writer, TV writer or marketing designer already knows how the games we play work motivationally. As said elsewhere, if you need to study the affects of traffic and spatial distribution on game design, read a lot of theory or simply pick up a book on professional party planning. The latter will get you there faster.
Posted by: len | Tuesday, 03 March 2009 at 22:42
zenBen: you may recall I posted a summary of John Hopson's GDC talk on reward schedules to Only a Game years ago, and the new book (Beyond Game Design) cites this talk as a reference - so yes, this work in adapting Skinner's reward schedules to the videogames design problems was foundational to what I'm doing here.
The distinction is that here I am talking about the neurobiology behind the decision to continue to play - not simply presenting the schedules that reinforce behaviours. Clearly the two are closely related, but I think this is a significantly different perspective on the issue, since the work on indeterminacy is relatively new to the scene. No-one previously has been able to place indeterminacy as a central factor in the addictive quality of games.
Thomas Malaby's paper is also interesting in this regard - he reaches the same conclusion from an anthropological angle.
organic i/o: there are a lot of misconceptions that emerge from these sort of confusions. When you watch the players, you'll find all sorts of situations that intuitively you feel should not be - and yet they are!
Grinding is one of the more interesting ones, since one can be in the grip of a progress structure and bear inconceivably repetitive behaviour with genuine enjoyment.
I'm doing it just this week myself, "horse farming" in Dynasty Warriors 6. It would be mind numbingly dull to many people, repeating the same two levels over and over again (because these are the optimal choices for the exercise). But I don't mind, because riding a horse in a videogame is an inherently pleasing aesthetic experience for me, and having an excuse to do it over and over again is fun for me, in a way that it wouldn't be for everyone. But then, I express Achiever quite strongly, which tilts me towards the grind. :)
len: thanks for commenting - but I think you are wrong to suggest that the concept being presented here isn't new in some respects (while it does of course draw upon some much older studies).
The research on indeterminacy and the brain only begins in earnest in the 2000's, so you must be conflating it with older work, such as Skinner's reward schedules. This is foundational to what I'm presenting here, but there's also a new perspective in this piece.
I also don't believe that "any good fiction writer, TV writer or marketing designer already knows how the games we play work motivationally". Writers of fiction, myself included, learn to write compelling narratives, but they do not actually learn the neurobiology behind this - much of what goes on in this regard is about learning how to structure a compelling narrative, and how to connect emotionally with a reader. It's coarse-grained stuff - it's very different from what I'm presenting here in my opinion.
I appreciate the goal is the same in many ways, but the mechanics are very different. What's interesting to me in this regard is to look at the mechanics of narrative, then to look at the mechanics of game design, and to find ways to combine these techniques effectively. It's what we at ihobo have dubbed "design-integrated narrative", and it's one of our specialities.
As for marketing, well, good marketing is storytelling, and bad marketing is using sex to sell. :p
In the case of the neurobiology of marketing, this is largely about wholly different brain regions - chiefly the association areas (hippocampus) not the pleasure centre. I'll write about this at some point, as it's interesting in its own right.
"if you need to study the affects of traffic and spatial distribution on game design, read a lot of theory or simply pick up a book on professional party planning. The latter will get you there faster."
Ooh, tip! Do you have a specific title to suggest? I'm very interested in this, and also theme park layout design, but there are no books on the latter that I can find. (Desperate for a lead in this regard!). Any specific tip you might have for a book on the subject of professional party planning would be extremely welcome!
Posted by: Chris | Wednesday, 04 March 2009 at 07:39