Psychology, Rewards And Motivation

I think games are more fun when they provide frequent small rewards rather than infrequent large rewards. Games are most fun though when they offer both. However, the nature of rewards in games is surprisingly complex, largely as a result of how alien human motivation actually is compared with how simple it is assumed to be by most people. If you reward people, you positively reinforce behaviour. If you punish them, then you negatively reinforce behaviour. It seems simple, and on the face of it it seems to work – but that’s only on the face of it. The truth is that human motivation and the effectiveness of rewards is a lot more complex.

One of the ideas I hear very often is the idea of ‘bucketed XP’ – the idea that XP shouldn’t be a general currency of advancement, but instead that it should be useful only for advancing skills in the proportion in which you have used them. If all your XP comes from farming, it makes sense if you can only advance your farming skills – why should killing a troll make you better at basket weaving? Another variation is that you hide it from players – that skill awards (TMs) come after you’ve earned enough ‘hidden’ XP in a skill, at which point you get your advance. That way it gets around the randomness of the TM system, and ensures people get better at the skills they actually use.

There is nothing wrong with these arguments if what you are looking to do is make player advancement feel ‘realistic’. However, neither of these are actually ‘fun’ systems. Both of them turn the unexpected, ‘oh wow, a TM’ chemical reaction in the brain into a predictable, ‘if I do this then I will definitely get this’ reward. Such rewards work well in some limited contexts – they work well when people are doing boring, repetitive tasks (I will give you ONE SHINY PENNY for every one of these forms you fill in) – in more complex tasks, or tasks where intrinsic satisfaction is high, they actually have a negative impact. People are less creative, less engaged, and (paradoxically) less motivated when rewards are a simple ‘if this, then that’ mechanic.

There is a reason why slot machines and games of chance are fun – an unexpected reward tickles the pleasure centres of the brain much more easily than an expected reward. Additionally, they keep their value – an ‘if then’ reward loses value over time – it stops being perceived as a reward and soon becomes seen as the ‘status quo’. If you want to keep motivating people, you need to keep increasing the reward, otherwise they think the reward is an obligation on your part.

The other danger with ‘if then’ rewards is when you remove or lower them is that it is not, as you might expect, perceived to be the removal of an incentive – it becomes ‘the removal of something I once had’. The intrinsic motivation that may have been there when performing a task doesn’t come back, either. There is a tremendously interesting study that was written up about the administration of a day care centre – growing tired of parents arriving late to pick up their children, they instituted a financial penalty, thinking that it would dissuade parents. They put in place a punishment to negatively reinforce undesired behaviour. And what was the result? That’s right – the number of parents who turned up late, and they turned up *later* than they had before. This mystified the administration, but the reason is pretty simple – the only reason parents wouldn’t do it before was because they had an intrinsic reason – they didn’t want to violate social norms, and felt awkward about putting people out. When that intrinsic justification was replaced with a financial justification, it became a trivial exercise of comparing cost versus benefit. People viewed it purely as an economical transaction.

So, realising what had gone wrong, the administration reverted the penalty – no more penalties for turning up late to pick up kids. The result of that? Rates went EVEN HIGHER. The intrinsic justification doesn’t come back, people just thought ‘hey, a penalty that once was in place is no longer in place – bonus’.

The psychology associated with video games can be used to tremendously destructive effect – look at companies like Zynga – they don’t make games, they make psychological addiction devices. Sometimes you have to know about the psychology so you can make a conscious attempt not to fall into psychology traps. Sometimes you have to know so you can maximize the impact of game features. Sometimes you have to know because you want to manipulate certain desirable behaviour.

Zynga is a company full of unethical, dreadful people because they are using psychological shortcuts to separate people from the contents of their wallets. However, many of the techniques they use can be ‘sanitised’ and used in other games to great effect – I guess the key difference is in intention. I have a problem when the psychology is ‘Hey, we will mess with your brain you so you spend money here’, as opposed to ‘We will mess with your brain so you give our great game a try’. Here and there, I borrow concepts from other games in the hope that we can integrate them ‘cleanly’ and ethically into Epitaph. Login rewards are one of these – I want people to log in, and I want people to log in frequently – but it’s not like you’re going to find us rifling through your wallet while you’re here. Whether it will stay, I don’t know – but I’m certainly willing to experiment to see what works for us.

I think a lot about stuff like this when putting together systems in Epitaph – human psychology fascinates me, and the psychology of video games fascinates me especially. It’s tremendously useful to know about things like this because it lets us have a fair idea of how effective various rewards and mechanics will be in producing actual ‘fun’ from the time people put in. Thus, we don’t have ‘command xp’ – or rather, we do – we have rewards that come with using commands provided you are consuming some kind of limited resource in the process. That way, while the reward is expected (and small enough to be handed out fairly liberally), there is still the ‘lucky find’ that comes with actually encountering a consumable – safes, locks, resource nodes, NPCs – all of these are spread around the game, but their locations cannot be assumed. In addition, even guaranteed rewards (such as the new login reward system) have randomness built in to ensure that there is still the chemical release of a

lucky jackpot.

In the end, it’s important to keep in mind that games are about fun. Game balance, appropriate difficulty curves, engagement, immersion – all of these are ways in which to *define* fun and to maximise it over the long term. But you can forget that all of these things are only measures of fun, and they are not ends in themselves – if in balancing a game you make it a chore to play, then you have failed in your job as a game designer. Keep the fun in mind, always, and the rest will eventually follow.

Leave a Reply