#34: The Psychology of Rewards
(or: The Fun Factor, Part Two)by Shannon Appelcline
December 3, 2001
Last week, in The Fun Factor, I started talking about fun in MCRPGs. As I noted in that article, there are lots of things you can do that will inspire enjoyment in your players, and lots of things you can do that won't ... and you can engineer for most of them.
However, I just barely touched upon one element that is very important, because it does a huge amount to increase the enjoyment of your players. Reward Them.
It seems like a pretty simple formula:
reward = fun
But, if you want to do a good job, it's not quite that easy.
What is a Reward?
Before you can even think about using rewards to increase the enjoyment of players, you need to figure what rewards are, exactly. In other words: how can you reward players? There are an infinity of rewards in our real world alone: salaries, bonuses, trophies, back rubs, and dessert if you finish your meal, just to name a few. Most of these distinct types of awards can be grouped into a few categories in MCRPGs.
Objects: These are physical (virtual) items, such as a new suit of chain mail, an extra week's worth of rations, or The Blazing Sword of Glory and Honor. Things that tend to be called portables in MCRPGs. They can also be real items that are slightly harder to carry around, like a patch of land just outside of town, or a bank account with a yearly dividend. In D&D: gold pieces.
Advancement: Improvement to one's character is one of the most common rewards in RPGs of any type. Your character gets better, whether he gains levels, trains his statistics, improves his skills, or learns new spells. This is experience, however it tends to be laid out in your particular game. In D&D: experience points.
Power: Often closely connected to either objects or advancement, power is notable in and of itself because it is an end rather than a means. Players want to be able to do things, to control reality. They want to have a hand in what happens in the world and to choose how the world works. In D&D: high-level spells or back-stabbing ability or attacking multiple times.
Knowledge: Because this reward can be somewhat abstract, it's sometimes forgotten in RPGs. Nonetheless, players love learning the answer to a long-standing mystery in a campaign, or just hearing a bit of gossip that might be useful in some future adventure. Knowledge can also be something really concretely valuable, like a confession penned by an arch-villain. In D&D: treasure maps.
Recognition: In traditional tabletop RPGs, recognition wasn't worth much; after all, all your handful of fellow gamers are your friends, and they know who you are and what's going on with you. Expanding out to the world of MCRPGs, recognition becomes very valuable. Players love to see their names mentioned on web sites, to be part of the news, or to be mentioned in the game manual. In D&D: convention tournament winners.
So, which type of reward should you offer when you're trying to make sure that players are having fun? All of them, with one caveat:
Moderation: In general if you're offering a reward, it has value only because it's rare. Looking at my equation, "reward = fun", it might seem like you could increase the fun by pouring out the rewards from an ever-full cornucopia. Unfortunately, if everyone has the cool objects or the high levels of advancement, then no one gets much enjoyment from them, and they instead start looking for the next thing, and that's what's going to be fun for them.
The Science of Operant Conditioning
I said that this week's column might be a little controversial, and here's why: I plan to apply some of the science of operant conditioning to the theory of game design.
If you're unfamiliar with it, operant conditioning, also known as Skinnerian conditioning, is a method frequently used to train animals to engage in certain behaviors. I most recently read about it in a book by Karen Pryor called The Lads Before the Wind, in which she describes how she spent years training porpoises for Sea Life Park in Hawai'i.
But, operant conditioning is more than that. It describes a whole method of learning and, most importantly for us, it describes methods for effectively applying rewards. What!? you might think. Can't you just put a reward out there, then proceed on with your life? You could, sure, but operant conditioners have clearly showed that isn't the most effective method of applying rewards.
In operant conditioning, you see, you shape behavior by offering rewards whenever a desired outcome is achieved. (Porpoise leaps out of the water? Apply a reward.) Operant conditioners have learned that certain methods of applying rewards are more effective for training than others. And it seems like a really simple jump to say that those same methods are also likely to cause players to have more fun.
Here's a few of the biggest points:
Invariably rewards will be offered in response to certain actions. Make sure the reward follows the action very quickly. If you don't, your players won't know which action led to the reward, and that'll lead to frustration, not fun.
In other words, if you're going to reward a specific action, don't just stop doing so at some time. That singular action should always produce a reward (or, as we'll see in a second, should always produce a chance of a reward). Again, the result of not doing this will be frustration.
Once you've established a reward in relation to a certain action, begin offering it variably. If a player knew that killing monsters resulted in a reward, for example, after offering a reward every time for a while, you could move to offering a reward for every second monster, then for every third monster. Eventually you'd want to randomize it, perhaps giving a reward about one monster in four. (You should never increase the reward increment over the three or four times that I mention here, or, once more, you'll end up with frustrated characters.)
This is effective simply because of the way that our psychologies work. We know there's some possibility of a reward every time, and thus we're always happily looking forward to it, even though it may not occur.
Or, put another way: reward, don't punish. The power of positive reinforcement is tons more effective than negative reinforcement. Take advantage of that!
Engineering for Success, Part Two
With these two pieces in hand--a listing of reward you can offer in a game, and an understanding of how rewards can be applied to maximize impact--you can do a lot to begin engineering systems that will definitively increase The Fun Factor of your game. You just need to think, mix, and match. Following are a few examples of game systems, using these precepts, that should be fun for players. I've maxed them out a little bit, to each include a few different types of rewards.
A Combat System: Combat systems are integral parts of just about any MCRPG. But, how can you implement a system that's going to be a lot of fun for your players?
First, start off by considering rewards. It's pretty traditional to offer experience points of some sort, ala Rolemaster or Dungeons & Dragons, but that actually violates one of the central rules I suggest, because the reward comes so long after the actual event (ie, when enough XP have been gained), that players can often end up bored.
Games like RuneQuest or Call of Cthulhu, which allow you to gain increases in skills in very small increments are a little better, but the rewards still tend to be removed from the rewarding effect, because a player usually don't get the skill gain until the end of the current session of gaming.
So instead, why not offer a fairly typical advancement reward, but do it on a much more rapid schedule. Let players gain advancement in skills or characteristics that were used in the combat as soon as the combat is successfully completed.
You can make these advancements very (randomly) easy to gain at early levels, and harder (randomly) to gain at higher levels, but put a pretty low ceiling on how hard those higher levels are to achieve. (Some games, including both D&D and RuneQuest, make those high levels really hard to gain, to the point where it's no fun. ) By making it harder to get higher levels of advancement, you've effectively put your rewards on a variable schedule.
Why not offer something more than this, though? Offer a recognition reward to people in combat too. Every day, list the fifty combats which were the most daring, deadly, heroic, or whatever criteria you decide to use. Set it up so that even the lowliest character can get a chance to see his name in lights if he does something really cool.
Now you've got two rewards, and the advancement reward is pretty well tuned by the rules of operant conditioning; there's some good likelihood that this will lead to fun among your players.
A Courtly System: If there's one thing you learn in designing, developing, and engineering MCRPGs, it's that there are so many players out there that there will almost always be a large number of them interesting in engaging in some activity that you would never have considered for a tabletop game: say hanging around your Lord's court, and getting in good with various nobles. You could offer a variety of rewards in conjunction with this game system.
Knowledge. Let your players learn things about your game world. Let them learn facts about the nobles that might aid them in further court interactions.
Recognition. Allow players to become the favored of certain nobles, get invited to their parties, and maybe become members of their secret societies. Put names of these folks up in lights (er ... except for the secret society members, that is).
Objects. Lords can grant fiefs, forgive taxes, and do any number of other things which might bring tangible gains to your players. These would probably be the highest level of courtly rewards.
Power. And beyond all of that, by becoming nobles themselves, players might be able to themselves give some of these rewards to other players.
Like me again state that these will all be the most effective, which is to say leading to the most fun, if you make sure all of these courtly rewards are quick, consistent, and ultimately variable.
An Alignment System: Back to the days of D&D, alignment has been a common system in RPGs. Unfortunately, it's usually been a system based solely on negative reinforcement. Players start off with an alignment that they choose, and then get dinged for acting out of alignment, and if they act really badly they might even be forced to change alignment and have the possibility, say for a paladin in D&D, of being forced out of their class.
If you want to build an alignment system, consider positive reinforcement and rewards instead. It might be a lot nicer to have players start out the game at a neutral spot. Therefore, when they start drifting to an alignment via gameplay, it'll be a positive thing. Put rewards down the path--extra abilities or recognition that come to people who have achieved a certain amount of alignment loyalty.
There's still the possibility for negative reinforcement in this model, as backsliders could start losing their abilities, but at least the core of this system is built around positive reinforcement instead.
The "Darker Side" of Operant Conditioning
There's a lot more to operant conditioning that's ultimately not of much relevance to the "Fun Factor" issue, but is of more general relevance to the MCRPG engineer. You can use it to encourage certain behaviors in your players, to get them to do certain things.
I call this the "Darker Side", but put it in quotes because I find little offense in this idea, but suspect some readers might. In truth, whenever you're designing a game system you're innately going to be encouraging some behaviors and discouraging others. A combat system, for example, encourages combat, while a system that offered experience for sitting around and chatting with other players would encourage the same. Quite bluntly, it's what I talked about in Thinking Virtually #18, The Game Is what The Game Is. So consider this as a possibility before dismissing it out of hand.
The topic of operant conditioning might be worth returning to for a more complete discussion at a future time.
This completes my own thoughts about The Fun Factor. However, over the next weeks I'm going to be presenting some alternative views--thoughts that other game designers have had about how you can introduce fun into your games.
The articles currently planned are:
They're a fun set of articles, written in response to each other. I hope you'll stay with me as I complete these discussions of "The Fun Factor" throughout the month of December.
See you in 7.