Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
October 23, 2014
arrowPress Releases
October 23, 2014
PR Newswire
View All
View All     Submit Event





If you enjoy reading this site, you might also want to check out these UBM Tech sites:


 
Predicting Behavior: Dissecting rewards
by Nils Pihl on 11/06/13 05:21:00 am   Expert Blogs   Featured Blogs

The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.

 

Those who criticize gamification because extrinsic rewards can be demotivating are missing the point. There's a problem with how we talk about rewards that we should address, ourselves, as well.

In the episodes "What is Play" and "What are rewards?" of Decoding the Game I argued that it was important to understand rewards and motivation in terms of the intrinsic and the instrumental, which prompted an important question:


How is this any different from intrinsic/extrinsic distinctions?

First of all, let me make it clear that I am not unaware of this distinction being the dominant way of distinguishing and categorizing rewards, and that it is passionately promoted by people with a lot more authority than I have. That being said, I wish to borrow your collective ears for a moment to argue the merits of intrinsic/instrumental distinctions.

When examining a reward using the intrinsic/extrinsic model (IEM), we are describing a different set of reward attributes than the ones I think are the most predictive of behavior. IEM describes the mechanics of reward delivery. In this popular model, an extrinsic reward is something somewhat tangible given to the player, whereas an intrinsic reward is something merely perceived. Where an intrinsic reward (in this terminology) might be "feeling satisfied", an extrinsic reward could be a point, a badge or some in-game currency.

In an almost libelously simplified way, IEM deals with who gives the player the reward, where the reward is coming from, and if the reward is tangible/visible to others. Intrinsic rewards are those rewards that the player affords herself ("I will take pride in this") and extrinsic rewards are those that the designer/game gives to the player ("Here's something for your trouble").


So what is the distinction you're making?

In the intrinsic/instrumental model (IIM), the attributes being described are less mechanical, and have more to do with why the reward is rewarding. IIM deals with the player subjective value of a reward, rather than its origin. The distinction is borrowed from value theory (philosophy), and is meant to describe why and how we perceive things as valuable. More importantly, IIM resonates rather well with a lot of data from behavioral economics and behavioral psychology, even though the distinction is rarely made by the people conducting the research. I argue that there is a compelling case to be made that, even at the level of the brain, we respond differently to intrinsic and instrumental rewards.

In IIM, the word intrinsic has a completely different meaning than it does in IEM, the IIM definition being "something that has value in and of itself". An intrinsically valuable thing is a thing that we value because of how we experience it, not because of some use it has. The taste of chocolate is intrinsically valuable, whereas a voucher to buy said chocolate is instrumentally valuable.

This means that something that is extrinsically rewarding in IEM can be intrinsically rewarding in IIM, for example: Being praised by a loved one. IIM considers praise intrinsically rewarding (it just feels good), where IEM considers praise to be extrinsically reward (it required another party to give it to you).

However, interestingly, everything that it is intrinsic in IEM is also intrinsic in IIM, but not vice versa (as demonstrated above).

1) All IEM intrinsic things are IIM intrinsic. 

2) Not all IIM intrinsic things are IEM intrinsic.

Neither of these two claims are axiomatic definitions, but are observations we've made. 


And you're claiming this helps with predictability of behavior?

Let me first change IEM to internal/external, so we can tell the two models apart. That gives us:

1) All internal rewards are intrinsically valuable. 

2) Not all intrinsically valuable things are internal rewards.

Moving on. There's a substantial body of research that our industry likes to cite (Ariely [1], Glucksberg [2]) that shows that external rewards can be demotivating and hurt performance. What I find interesting is that there is a very high correlation between demotivation and instrumental rewards.

Praise (external) is rarely (ever?) found to be demotivating, but money can be. In fact, money can even be demotivating by proxy. At this year's GDC Shanghai Dave Mark of Intrinsic Algorithm gave a great talk on the psychology of numbers[3], where he shared a couple of interesting anecdotes:

A Snickers bar was less motivating when the subject knew its price.
(Here's some enjoyable chocolate vs. the work you are doing is worth 1 dollar)

A bottle of wine was less appreciated when the subject knew its price.
(Here's a token of my appreciation vs. this dinner was worth 20 dollars to me)

Lawyers prefer free work to discounted work.
(I'm doing something because it is the right thing to do vs. I am not being paid what I am worth)

I'll freely admit that I don't recall which studies Dave was refering to (if you know, post it in the comments). I submit to you that external rewards are likely to be motivating when they are intrinsic, and risk being demotivating when they are instrumental.


What does any of this have to do with gamification?

We all like pointing out that the mechanics of a game is not what makes the game, but I submit to you that the mechanics of a reward is not what makes the reward, either. This is not a problem unique to gamification, but something that we often fall short on ourselves, in our own discourse. When we reduce our reward systems into internal/external rewards, we are failing to take the player's subjective experience into account, and miss an opportunity to make valuable predictions.

IIM is a more precise scalpel than IEM, but the two are not mutually exclusive. In fact, once IEM terminology is changed to internal/external we find that there is no conflict between the two models. I hope you'll join me in advocating for this use of the terms, so that we can have clear and productive conversations about design and engagement, and be more credible than gamification proponents.

[1] Ariely, et al. (2005). "Large Stakes and Big Mistakes". Federal Reserve Bank of Boston
[2] Glucksberg, S. (1962). "The influence of strength of drive on functional fixedness and perceptual recognition". Journal of Experimental Psychology
[3] Dave Mark. "Psychology vs. Structure". GDC Shanghai 2013


Related Jobs

Nexon America, Inc.
Nexon America, Inc. — El Segundo, California, United States
[10.22.14]

Localization Coordinator
Petroglyph Games
Petroglyph Games — Las Vegas, Nevada, United States
[10.22.14]

Producer
DeNA Studios Canada
DeNA Studios Canada — Vancouver, British Columbia, Canada
[10.22.14]

Analytical Game Designer
Xsolla
Xsolla — Sherman Oaks, California, United States
[10.22.14]

Senior Business Development Manager






Comments


Ryan Braley
profile image
I thought I'd share one of the interesting questions we threw at Nils as he was writing this post:

If you're duping items with the Horadric Cube (or just printing money), what kind of reward would it be, according to IEM?

Would it be extrinsic, because it produces a tangible good, or would it be intrinsic, because the reward comes from the subject?

For that matter, is a badge still an extrinsic reward if you made it yourself?

:)

Tasley Porter
profile image
Clearly Nils is the expert in this thread, but based on the information he's shared I'd like to make a guess.

The item created from the Horadric Cube isn't a reward at all. It's merely the product of your labor, which is itself an intrinsic reward. Now if we ask whether I use the Cube to make an item and then sell it I think the question becomes: will I feel more rewarded by giving it to someone or selling it. Tricksy ....

The badge, as the Cube, shouldn't be a reward at all.

How'd I do? lol :)

Nils Pihl
profile image
I'd say that IEM gives us the following (since it is not particularly concerned with the subjective experience of the player):

The OBJECTS created by the cube are extrinsic rewards (in that they are tangible and separable outcomes), regardless of the nature of the particular item.

The CREATION of the objects is arguably an IEM intrinsic reward, in that it gives the player an outlet for creativity, completionism, autonomy/agency/impact.

The SALE of a duped item is extrinsically rewarding.

Dave Mark
profile image
Thanks for the mention, Nils. Many of the examples in my lecture can be found in the book, Predictably Irrational by Dan Ariely.

http://www.amazon.com/gp/product/0061353248/ref=as_li_ss_tl?ie=UT
F8&camp=1789&creative=390957&creativeASIN=0061353248&linkCode=as2
&tag=davemark

Also, “Thinking Fast & Slow” by Daniel Kahneman.

http://www.amazon.com/gp/product/0374533555/ref=as_li_ss_tl?ie=UT
F8&camp=1789&creative=390957&creativeASIN=0374533555&linkCode=as2
&tag=davemark

Nils Pihl
profile image
Thanks for checking in, Dave, and thanks again for the great talk.

Ian Sturrock
profile image
I think you're missing some of the nuances of the IEM, here.

First up, a quick note on terminology. Self-Determination Theory is where the vast majority of the data on IEM comes from. They have probably 40-odd years' worth of empirical evidence, that has driven the whole concept of the IEM -- which makes me think that 'IEM' is the wrong term, since a 'model' is typically used when no experimental data is available. So, I'd prefer to refer to it as SDT, rather than IEM.

Secondly -- praise can be either extrinsically motivating, or intrinsically motivating, depending on context, relationship between praiser and praisee, etc. Praise is a form of feedback. Feedback that is perceived as *controlling*, is an extrinsic motivator. E.g. "Good girl". Praise that is perceived as *informational*, is an intrinsic motivator. My suspicion (not yet backed up by research) is that some personality types are more likely to perceive any given piece of praise as extrinsic or intrinsic.

I would suggest reading some of the SDT literature that deals with the research that has been done. _The Handbook of Self-Determination Research_ is a good starting point for general use. For game-specific studies, _Glued to Games_ by Ryan and Rigby is the main text. Alternatively, watch Deterding's Google Tech Talk on "Getting Gamification Right" --

http://www.youtube.com/watch?v=7ZGCPap7GkY

He manages to get across the practical, take-home points from SDT for game designers and gamification designers, very accessibly.

Nils Pihl
profile image
Thank you for your replay, Ian.

IEM (by which I only mean the classification of rewards as intrinsic/extrinsic) exists outside of SDT, and has been around for longer than SDT has. I also happen to think that SDT is an incomplete model, overrated and making poor predictions...

Perhaps I haven't been reading carefully enough, but I can't recall SDT making a clear definition of when praise is intrinsic rather than extrinsic. Part of the problem, I feel, is that SDT doesn't make sufficient distinction between rewards and motivators.

Behavioral psychology is still a field in its infancy, but we still throw around things Maslow's Hierarchy of Needs and SDT like they had way more evidence than they actually do.

SDT has a rather tough time predicting, or even explaining why, things like the "paradox of choice" happen. That increasing someone's choices (and autonomy) can be drastically demotivating, and it's not entirely clear (to me) how SDT proposes to incorporate that empirical finding into their model.

Ian Sturrock
profile image
IEM led directly to SDT, starting with White's concept of competence as a motivator, followed up by Deci's early work.

You are welcome to think what you like of SDT... but until you can point to studies that disprove what they've spent the last four decades proving, you'll have to forgive me if I give more credence to science than to thoughts. :)

The concept of praise potentially being either informational feedback (intrinically motivating), or controlling feedback (extrinsically motivating) is a fundamental one to SDT -- you'll certainly find info on that in most or all of the sources I mentioned above.

I don't throw around Maslow, personally, because I think his work is quite limited and dated. But SDT is entirely evidence-based -- all the mini-theories that make it up were formulated after studies were done, and evidence gathered. Again, loads more info in the Handbook (or just on the SDT website).

The paradox of choice is an interesting one, with some strong parallels to game design, but there isn't as much empircal evidence for it as there is for SDT. Game Theory could perhaps be a better route for understanding the paradox of choice. Imagine a game in which there are hundreds of options any time you want to make a decision -- you wouldn't play it, because the time required to research each decision to the point where you had enough information to make an informed choice, would be too much of an investment. As Sirlin's shown, having too many choices also increases the risk of a dominant strategy emerging. Again, thinking back to game design -- we want *meaningful* choice, not just choice. Three or four options, with "predictable consequences" (Doug Church), are probably more autonomy-supportive than a hundred choices, at least for most gamers in most games.

Nils Pihl
profile image
What is it you feel that they have proven that is so compelling, and how many competing theories can easily include the same data? How many competing theories have they managed to falsify?

SDT is promising, but young, is still makes poor predictions. If you need game theory to explain the paradox of choice, then game theory needs to be included into our models, and that's exactly what we're promoting with behavioral engineering (that also incorporates memetics).

Ian Sturrock
profile image
Read the Handbook. You don't appear to be aware of even the basics of around 40 years' worth of empirical research, which is what SDT is, and which you think is... young and promising?

There are no competing theories that include the same data, because SDT was derived directly from the data. Seriously -- if you've not even read any of their studies, which is what it sounds like, how can you possibly be qualified to dismiss the entire field?

The games industry is full of people who've read an article or two on psychology and assume they can now pontificate about why people game. The gamification industry is full of snake oil salesmen. Please prove you're not one of them. You've got some promising ideas in the OP, but you're kinda young. And you've certainly not presented any actual data at all. (Unlike SDT.)

Nils Pihl
profile image
Ok let's get a couple of things straight here, Ian:

Yes, I stand by my opinion that SDT, and the entire field of behavioral psychology, is young and promising. Pretending like I haven't read the studies because I disagree with you is not a terribly strong argument. You are free to show me that I'm wrong in the specific criticisms I have leveled against SDT in this thread. You're defending your favorite book, and that's commendable, but you're not actually addressing my claims.

Yes, there are competing theories that CAN incorporate the same data. Note that I said "CAN". A lot of the findings of SDT, in my opinion, does not lend special or exclusive credibility to SDT in favor of other research endeavors. Behavioral economists, for one, often look at the same studies you refer to to support and augment their positions. SDT and behavioral economics is not the same. SDT is only ONE theory even in behavioral psychology, not THE only one. In fact, SDT itself doesn't claim to be the panacea or behavioral-theory-of-everything that you seem to think it is.

Third, I am not dismissing the entire field, nor have I claimed to be qualified to do so. I have always included, recommended and commended SDT in my professional career - but I think that it has come to a point where people outside of academia give it more credence than it deserves. Our industry is full of people that have read a few studies, or the Handbook, and now think they can pontificate on the state of behavioral psychology. I am not dismissing SDT, I am pointing out that there are areas where it makes weak predictions. Strawman arguments don't impress me, but I'm happy to indulge you while you dismiss a fabricated argument instead of the one you seem to have misunderstood.

"The games industry is full of people who've read an article or two on psychology and assume they can now pontificate about why people game."

Surely, they must read the entire Handbook before they get to pontificate.

"The gamification industry is full of snake oil salesmen. Please prove you're not one of them."

I haven't tried selling you or anyone else here anything - I have invited you to examine terminology that I feel adds clarity to our discourse, and have strongly criticized gamification. My last three entries here have been almost purely philosophical, and I've only made very few predictive claims, none of which you or anyone else has seen fit to challenge. It seems like you're arguing just to look clever, because you haven't actually challenged any of my claims, you've just made the claim that the distinction Intrinsic/Extrinsic can be attributed to SDT. I think Huizinga, Caillois etc would disagree. SDT was born out of IEM, not the other way around.

Thank you for thinking I have promising ideas, and for showing that you think my age is relevant to the merits of my argument. I think that says more about you than it does about me.

Ian Sturrock
profile image
Heh. I don't actually know your age. I was just pointing out that "young and promising" sounds pretty patronising, if you hadn't spotted it.

You don't appear to know the basics of the 40-odd years worth of empirical research that has gone into what you call the IEM. All I've been doing is trying to show you where you can find the basics. Because if you do have the potential to teach us something -- which would be great (MUCH more research is certainly needed in this field), it'd need to be done the way that knowledge is always advanced -- incrementally, either building on or disproving the research that has gone before. If you aren't aware of what you're dismissing... well, it's a bit like what would happen if a game designer said, "Hey! I have this great idea for a new sort of game. How about if we take World of Warcraft, but instead of playing it with graphics and sound and stuff on a computer network, we play it non-digitally, round a table, using dice to resolve actions, and bits of paper to record our characters' statistics."

Ben Lewis-Evans
profile image
Ian is correct that SDT does spend quite a lot of time talking about the difference between controlling praise and non-controlling praise. With, as Ian says, non-controlling praise supposedly being the one that doesn't decrease motivation. SDT also makes a distinction between expected vs unexpected praise (or any ''extrinsic' reward for that matter), in that expected praise is supposed to be more potentially demotivating than unexpected praise.

However, SDT is certainly not without its detractors and doesn't stand unchallenged in psychology. For example the whole field of behaviourism, particularly radical behaviourism, doesn't really agree with SDT and the data it puts forward. Rather a radical behaviourist would point at their own data to first show that a) there is no such thing as an 'intrinsic' reward and b) that reinforcements are what is important, not rewards.

Even putting such extreme behaviourist arguments aside there are other issues with SDT that can be raised. First is that SDT is often presented as a blanket statement that external rewards are bad in all situations. However, in reality SDT has a caveat that it generally only applies to situations/tasks that are already intrsinically motivating/interesting. This is a caveat that Deci admits to, but sometimes doesn't make 100% clear in his writing.

SDT has also been criticised based on the core/classical methodology that it uses to assess the impact it uses the impact of rewards. In that a) if you go back and read the very early Deci papers where the methodology was developed the sample sizes are low with quite a few non-significant results b) that often no pre-reward measures are taken to compare with the post-reward situation c) that rewards in reality are often repeated, whereas the classical methology that SDT is based on tends to use single period reward delivery and that d) rewards in real life are often on-going and not arbitrary removed, whereas the classical methodology for SDT only shows evidence for reduced motivation after the removal of the reward.

The form of the reward might also matter. With some saying that extrinsic rewards that are task relevant aren't as likely to produce negative effects. Also, again there is the important, although circular, distinction between a reward and a reinforcement.

Personally, I also haven't seen a decent attempt by SDT to address potential personal differences in reward type sensitivity. But, perhaps they do in the handbook (sorry, haven't read that - my comments are based on academic journal papers)

If you are interested in some commentary on SDT you can check out these two papers for example (not to say they are he only ones, but just the ones I could grab quickly) - Paper 1 http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2731358/pdf/behavan00
009-0003.pdf and Paper 2 http://www.psychology.uh.edu/faculty/Eisenberger/files/16_Detrime
ntal_Effects_of_Reward_Reality_or_Myth.pdf

As an aside, Nils, I find your use of 'behavioural psychology' a bit confusing. What do you mean here? Behaviourism? Because SDT is more of a cognitive/social psychology theory in my opinion.

Nils Pihl
profile image
Again, I don't call SDT by the name of IEM. I just abbreviated the distinction intrinsic/extrinsic (which is in no way unique to SDT) to IEM. Unless you want to call Gabe Zichermann a proponent and champion of SDT, you'll agree with me that the intrinsic/extrinsic distinction is common outside of SDT.

I see you prefer having the cake and eating it too, criticizing me for calling SDT young but then saying "MUCH more research" is needed. We're apparently in agreement, but you still choose to call me out for glob-knows what reason.

You also choose to insist that I am unaware of SDT. I don't know how you imagine that will help your argument, should you choose to present one.

I really don't know what compels you to say, and then repeat, that I am unaware of research that I myself have referred to earlier in this series, and I fail to see how you have addressed any of the actual predictive claims that I have made.

Again, I am not claiming that IEM is a scientific theory, nor am I presenting IIM as one - I am discussing terminology that I think adds clarity to our discourse.

What are you trying to achieve?

Ian Sturrock
profile image
They do mention differences between people, for sure. And they also demonstrate that a given person might have different motivators in different contexts, too. I think (again) that more research is needed. My suspicion is that some personality types are far more hostile to extrinsic rewards than are others (though, an SDT researcher might argue that this is because those people haven't been quite so conditioned to want extrinsic rewards, as others have).

In particular, there's a need for more research into how what I call 'in-game extrinsic rewards' (gold, XP, etc.) are perceived -- do we think of them as informational feedback (as Ryan and Rigby argue in _Glued to Games_), or as controlling feedback (as we would if they were 'out-of-game extrinsic rewards', given that there is some evidence from behavioural economics, neurology, etc. that we tend to treat virtual rewards very similarly to their 'real' equivalents).

For me, behaviourism is ethically suspect, and practically dubious in its long-term effects. Yes, we know that you can use operant conditioning to modify behaviour -- but we also know that humans are considerably more complex and rebellious than (say) rats in a box, or donkeys rewarded with carrots and punished with sticks. Absent near-brainwashing levels of conditioning, behaviourist attempts at behaviour modification in humans tend to only work in the short term, due to the overjustification effect (among other things). Although variable reward ratios can certainly push that into the medium-term, if we're talking game design.

Nils Pihl
profile image
Thank you for your thoughtful response, Ben. I always enjoy your commentary.

Our use of "behavioral psychology" here is meant to be the widest umbrella term possible. It's an imprecise and sometimes even borderline libelous labeling of disparate fields, but what we're trying to do is find the most predictive parts of several different fields of research, including:

Cognitive neuroscience
Functional neuroscience
Behavioral economics
Neuroendocrinology
and psychology, in its many guises

There are certainly elements of behaviorism that can aid in making certain predictions, but it is by no means our central focus - nor is SDT. We focus on making predictions, not advancing one field over the other, and we take pride in our track record without claiming to dismiss the important research that we borrow from.

We don't engage in holy wars that we have not ourselves started.

For someone with your strong background in academia, I can see how you would take issue with the umbrella term - but please keep in mind what our endeavor is here:

Sharing and exploring terminology, models and concepts that aid in making predictions, particularly pertaining to gaming. These are heuristics that you are welcome to try, and even falsify, but in no way an attempt from us to position ourselves as scientific authorities.

Ian Sturrock
profile image
You suggested there are gaps in the research, in the area of praise. I pointed out that there aren't, and offered suggestions for further reading. Given that the way SDT treats praise is a fairly basic part of SDT, it did seem likely that you'd missed a few other nuances.

What am I trying to achieve? I'm trying to help you to do enough reading and other research that you can formulate your ideas into something useful.

Zichermann misuses and misunderstands the terms intrinsic & extrinsic. See his dialogues with Deterding.

(And, naturally, much more research is *always* needed!)

Nils Pihl
profile image
You say there are no gaps regarding praise? I disagree, and so do you:

"In particular, there's a need for more research into how what I call 'in-game extrinsic rewards' (gold, XP, etc.) are perceived -- do we think of them as informational feedback (as Ryan and Rigby argue in _Glued to Games_), or as controlling feedback (as we would if they were 'out-of-game extrinsic rewards', given that there is some evidence from behavioural economics, neurology, etc. that we tend to treat virtual rewards very similarly to their 'real' equivalents)."

Saying that "controlling" feedback is extrinsic while "informational" feedback is intrinsic is somewhat of a copout if you can't predict how the subject will perceive the feedback, as you yourself pointed out. How do you know if a piece of feedback will be controlling or informational? What is SDT's distinction, in your own words?

Simply asserting that the feedback is perceived as controlling when motivation-levels go down is not very impressive.

I love the Zichermann/Deterding dialogues, but saying that Zichermann is just wrong is a "No True Scotsman" fallacy. IEM is obviously used by people outside of SDT, and is therefore worth addressing.

Here is a prediction that I stand by, and that I feel has ample support from many fields of research: A subject will have measurably different responses to an instrumental and an intrinsic reward. That gives the IIM terminology some predictive strengths that IEM lacks.

Ian Sturrock
profile image
You have a prediction! Excellent. Now -- go do the research -- test it out.

Yes, IEM is misused by Zichermann. I think Deterding proved that beyond doubt, unless one is Zichermann (and hence has a massive interest in being taken seriously as a 'gamification guru'). It is used outside of SDT, for sure. Notably in Flow theory, where researchers found a strong correlation between intrinsic motivation and frequency of experiencing a flow state. But then, the Flow theorists based their use of IEM on the work of SDT researchers... so for me, SDT and your 'IEM' are intrinsically (sorry!) connected.

Saying that "controlling" feedback is extrinsic while "informational" feedback is intrinsic isn't a copout -- it's a finding based on empirical research. Like the rest of SDT. Unlike your claims. So far. That doesn't mean that your claims don't have any merit -- just that they need to be tested before we can find out if you're right.

There are lots of finely nuanced distinctions between controlling and informational feedback. It's not necessarily possible to predict how a given person will perceive a given piece of praise. But there are certainly autonomy-supportive forms of praise (i.e. those that are likely to be perceived as informational rather than controlling). Examples include: praising effort rather than ability/intelligence; numerically useful feedback (e.g. if you're doing an exercise programme and you are informed that you finally achieved what is widely classed as an Intermediate-level deadlift for your weight). I gave you a few examples of the reverse (blatantly, obviously controlling feedback) in my first post.

I'd love to see the whole field shaken up by new findings. Maybe you're the person to do it. But you need to show us the data.

Ian Sturrock
profile image
(For some more insight into praise, by the way, check out the work of Carol Dweck, notable _Self-Theories_ -- it's not part of SDT, but is directly relevant, I think.)

Nils Pihl
profile image
"Saying that "controlling" feedback is extrinsic while "informational" feedback is intrinsic isn't a copout -- it's a finding based on empirical research. Like the rest of SDT. Unlike your claims. So far. That doesn't mean that your claims don't have any merit -- just that they need to be tested before we can find out if you're right."

Seems like a claim that would be particularly difficult to falsify, Ian. If you can't predict when something is controlling, you must be asserting that something is controlling to make sense of your data.

I see this kind of backward reasoning from both behavioral economists and SDT proponents. I also see it by you, in your comments here.

Feel free to share with the readers here what the predictive definition is of controlling motivation. I freely admit to being unaware of a helpful definition, here, and welcome your input. Give the definition, or link to a relevant academic paper that provides a definition. You've added very little to the conversation with your last couple of posts.

You fail to address criticisms leveled by me and by Ben Lewis Evans, instead just saying that I should read a book.

Edit: I'll read Dweck.
Edit2: I seem to have neglected to read the last paragraph of your last comment. I recognize that you have now given us somewhat of a definition through examples.

Also, I am not claiming to be about to shake up the whole field with new findings. Get real, Ian. Stop with the strawman arguments. I am starting to suspect that you're just trolling me.

I have simply said that adding a simple lens can aid with predictions. I am not an academic, but this lens has served me great in my career and I offer it to you for scrutiny and experiments. I am not trying to position myself as an authority above Ryan, Deci, Rigby or anyone else (except Zichermann. HAH!) I repeat over and over that I am not trying to dismiss SDT wholesale, only saying that it is still an incomplete model, to which we seem to agree.

Ian Sturrock
profile image
Again, it's not a 'claim'. That's not how SDT operates. The researchers do the research, then they work out what it shows (even if that's not what they might originally have expected). The controlling/informational feedback distinction has been examined in great detail, over dozens of studies and several decades. Here's one source that shows that distinction, if you like:

http://www.selfdeterminationtheory.org/SDT/documents/1984_Koestne
rRyanBernHolt.pdf

I'm pretty sure 'the readers' can use Google though.

Nils Pihl
profile image
Commenting to notify you that I made edits to my previous post.

Edit: Thanks for the link. I seem to recall having read it once before, but you've convinced me that it's worth reading again. Overall, thanks for the conversation but I think I have to put a cap on it for now - maybe we can pick it up again after the next episode is out?

Ian Sturrock
profile image
We're a bit closer to agreement, then.

Any scientific field of enquiry has gaps. We have no perfect understanding of anything.

I felt that if I offered you useful sources of info on praise, you could update your own understanding of the current state of research into 'IEM'. That way you can offer a newer, more complete version of your model.

At the moment, there's a big disconnect between academic research into games, and game design / gamification design practice. I'm keen to work on bridging the gap. For me, as more of an academic than a game designer, at least these days, that means ensuring that my research is relevant to game design, and can feed directly back into the design of better games. I don't think it's unreasonable to expect something in return, though -- which is, I'd like to see writers on game design and gamification, ensuring they do adequate reading of extant research, before creating new 'models'. I think it's probably at least as much of a feature of the game design mind as it is a bug, incidentally. Game designers are very much used to taking a small amount of info (e.g. the intended setting and background of the game) and turning it, intuitively and then iteratively, into a workable system for interacting with. But that approach is flawed if your goal is to advance human understanding rather than 'just' make a fun game. :)

I'm reminded of Schell: "We don’t understand human pleasure at all; it’s like we’re alchemists trying to understand how to turn lead into gold, but we haven’t even discovered the periodic table yet." Plausible-sounding stuff, but sadly it just demonstrated his ignorance of the entire field of positive psychology!

Edit: for sure -- I'll look out for your next blog posts with interest.

Curtiss Murphy
profile image
What about Control? Say it's 10 PM on a Sunday night and I'm setting my clothes out for Monday morning, when I realize, CRUD! I didn't wash my white work shirt. I run around in a panic.

So, of the actions and rewards that follow, which are extrinsic and which are intrinsic? Late night shirt washing has clear tasks, immediate feedback, and increased difficulty cause the time crunch. It feels a bit like being in Flow. And, when I finish, I'm happy for getting it done so quickly, and the next day, I enjoy feeling of being cleanly dressed. Intrinsic, intrinsic, intrinsic. And yet, aren't clean socks themselves an extrinsic reward? Hmmm...

Now, consider the same situation, except add the fact that every Monday, my pain-in-the-butt boss checks uniforms. If I show up in a dirty shirt Monday morning, he's gonna dock my pay! And all of a sudden, the motivation aspects have become extremely complicated.

Even in this trivial situation, it's hard to decipher whether something is intrinsic or extrinsic. And yet, there's much less confusion when it comes to CONTROL! That's what I focus on when I'm guiding new designers. Cause it's so much more practical! Rather than trying to decipher every nuanced interaction in the filter of extrinsic or intrinsic. Try, instead to pay attention to whether we're trying to control our players! Cause sure as heck, our players will notice.

Know what else gets forgotten? The idea of baseline rewards. These are external rewards that are sort of about control, except that really, they're things players EXPECT to see. Like being paid for work, or getting XP for killing a mob. Even though they're clearly extrinsic, they're probably required.

@Nils - I deeply appreciative this explanation of "Instrumental" and now that I've read more about it, I think the distinction between it and extrinsic might benefit from a few more iterations.

Nils Pihl
profile image
Curtiss, all these problems you mention is why I prefer IIM to IEM.

IIM deals with WHY we consider something to be rewarding. If I iron my shirt BECAUSE my boss is an asshole, it is clearly work, and instrumentally rewarding.

If I iron my shirt BECAUSE I enjoy the feeling of neatness, just for myself, it is intrinsic.


The difference between instrumental and extrinsic, to me, is very clear: That something is instrumental simply means that it has some USE or utility - it is not the ULTIMATE outcome.

Making money is instrumental, because you want to use the money to buy a boat (instrumental), because you want to increase your social status (intrinsic).

Or making money is instrumental because you want to buy a boat because you love sailing, just for you. Intrinsic.

Bart Stewart
profile image
After reading the transcripts of the previous entries in this series, and thinking about the extrinsic/intrumental definitions to try to understand the distinction being advocated between them, I think I'm coming down on the side of "they're functionally similar enough to be interchangeable for good game design."

That's actually because I agree that subjectivity -- what the player cares about -- does matter. More specifically: Game rewards are motivating when the kind of reward obtained matches what a player naturally values, and demotivating otherwise. And one important way of assessing motivation (and thus predicting behavior) is distinguishing between what comes from inside and what comes from outside.

If I'm wired to care more about externals than internals, so that I typically measure my happiness and self-worth by accumulating tokens that other people say have value, and I feel good when I'm effectively following rules defined by others, then the games I'm likely to find most enjoyable are the ones that give me valuable objects that others can see. In a game, that means rewards with extrinsic value defined by other players, such as money or achievements or leaderboard status. For the same reason, I'll be considerably less interested in playing a game whose rewards are more about telling me that I'm a good person or giving me inapplicable knowledge because those are rewards whose value is intrinsic to the player, and thus not relevant to how I'm perceived by other people.

Conversely, if I naturally feel best when I'm effective at making my actions align with my internal beliefs regardless of what other people claim is right or wrong, then the games I'll find most enjoyable are the ones whose rewards have intrinsic value to me. I'm most likely to be motivated to keep playing a game whose rewards let me feel like a good or clever person. Those rewards have value to me even if they have no overt utility in the world of the game. Examples of this could include a rescued character saying "thank you!" and occasionally sending the player notes listing the things they were able to do because the player saved their life (something you rarely if ever see in a game because extrinsic rewards are more easily understood and more valued by a majority of today's gamers), and revealing deep lore about the game world (which has no functional value and is interesting purely on its own merits). The gamers who find delight in these kinds of rewards are exactly the people who are likely to find exterally-valued (i.e., "extrinsic") rewards demotivating.

In both of these cases, the key design point is that you maximize a player's desire to keep playing when the form of the reward matches what the player values in him/herself. That boils down for me to deciding who the target audience is for the kind of game I want to make, and looking for mechanics and outcomes (rewards) that match the interests of those players. If I want to appeal to explorers, then I'm going to focus on creating gameplay that lets players discover interesting things about the world of the game even if that knowledge has no direct application, because that's what's fun about discovery to that kind of gamer. I may also include some features tailored to token-accumulation (because that will attract more players), but not to the point that those rewards outweigh the intrinsic rewards that my targeted explorer players prefer.

That argument also applies to games intended for other kinds of players. In all cases, I believe the goal should be to align reward types to the preferred play styles of the people you want your game to attract and retain. That's why, even if I don't see a significant difference between thinking of external-value rewards as extrinsic or instrumental (though I can see how an instrumental reward can be used to help satisfy an intrinsic desire), I do see value in talking about the nature of reward types as this series has been doing. I'm looking forward to reading the next entries.

Nils Pihl
profile image
Thanks for your thoughtful response, Bart. I enjoyed reading your challenging commentary, and hope that you'll stick around for the coming episodes.

I disagree about the terms being interchangeable, and argue instead for both distinctions describing interesting attributes of a reward.

Both distinctions have their place, and I hope that the game industry will start embracing IIM more.

That being said, I happen to believe that IIM is more predictive than IEM, whereas IEM is quite useful in a design meeting when discussing mechanics.

The behavior of the player and the mechanics of the game are not the same.


none
 
Comment: