I was reading an article on Hyperbole Games earlier, in which Grant discusses fine tuning games. His background in economics gives him an interesting perspective and the rest of this post is on a tangential topic, so you won't learn the bulk of what he read here - go check it out Before I start on my tangent I'd like you to have a think about this image:
Grant's article touches on the nature of feedback mechanisms in game. Some games reward good play with bonuses, granting victory points, resources or some other important thing. Some penalise bad play with penalties, taking away important things or restricting player's actions in future turns. Most games do both to some extent and in a lot of situations it is possible for a designer to implement either. In a two player game granting one player a two point bonus is functionally identical to penalising the other player by the same amount - but they can be received very differently. Humans are naturally risk averse and are reluctant to choose an action that risks penalties over one that risks not getting as great a reward. So this decision can be meaningful.
I've read the argument a few times now that in game design the carrot is superior to the stick. After all, the carrot makes people feel better than the stick and you want people to have fun while they're playing your game. There's definitely truth to this, given two functionally identical options it's better to make your players feel great than feel bad, but I think it's misleading to make these decisions by comparing how players feel when faced with the carrot or the stick. I'd like you to think for a bit about this image:
This image, "lunch atop a skyscraper" is world famous and has been replicated however many bajillion times. There are lots of things about it that sing to its viewers, but I'd like to focus on one thing in particular here. What is the difference between what the builders are doing here and the actions of the children in the first image I asked you to consider?
There is a difference, I'd definitely feel different in those situations, but they're functionally very similar. Both groups are sitting down and having lunch. Both are balancing on a bar of about the same thickness without their feet touching the floor. Mechanically, the actions that they are performing are more or less identical, but I think most people would agree that they're not the same thing.
I think that the important difference is the risk; if one of the builders fell it is very likely that they would die. The contrast between this and their apparently blasé attitude gives the image a lot of its weight. You can observe a similar reaction in yourself watching videos of free running, pay attention to what happens to your heart rate and reactions watching people tricking on the ground to those bouncing around rooftops. If that's not convincing go and take up rock climbing and see if the bottom of the wall feels different to the top. We've whole ranges of activities that basically fall into the category of "Risking huge negative outcomes that then don't happen because of our skill" and it feels awesome.
So I don't think that "carrot vs stick" is a fair comparison. Really a player who's done well will either get "carrot" or "not stick" and a player who's done poorly will get "stick" or "not carrot" so the question of how to make players feel most awesome comes down to the comparisons between these things.
If the "not stick" can be made to feel more awesome than the "carrot" and the "not carrot" feels more miserable than the "stick" then switching to a punitive game system improves the players experience, even though the penalty generates a more negative emotional reaction than the bonus it has replaced. In general this doesn't happen because making avoiding the penalty feel awesome enough to outweigh a bonus often makes the penalty so absolutely diabolical that it can't help but undermine the gain. A lot of games that feature player elimination fall into this trap.
I'd argue that this isn't inevitable though, due to the nature of how players react to rewards and penalties. If a game contains a reward players want it, they strive to achieve it and they feel cheated if the game makes it impossible to achieve. On that note I'd like to have a brief aside...
Prophecy, I hate you for pied piper. I hate you for writing an achievement for slaying two or more rats. There are only two rats in the game. Only two. And they're so weak that they'll die whenever anyone sneezes near to them. The same player never kills both of them. Of course the deck is shuffled and the enemies reappear a couple of times - but it doesn't matter. Both players kept the rat card as proof of their kill. They'll never get the second, never! For both of them keep their half of a bonus, doing nothing but telling them to get the other half. For the whole game they'll be forced to keep a useless card in their play area that's sole function is informing them about the bonus that they might get if the other player storms off in a rage and their cards are added back into the deck. The dead rat just sits there, taunting me with the bonus I'll never get, mocking with its cold dead eyes.
...I needed to get that out of my system. As I was saying, if there's a bonus players want it to happen and get annoyed if it doesn't. The same isn't true of a penalty, players will actively try to avoid a penalty and while they might balk at a penalty so improbable as to be toothless, nobody is upset if it's unlikely or will only emerge as the result of a major tactical blunder. This makes a penalty significantly different to a bonus in that it never has to happen, or at least doesn't have to happen often.
This provides a little bit of wiggle room for the designer. If you can design a penalty such that avoiding it feels awesome, but that is relatively easy to avoid then there's room to make your game better with a stick than with a carrot. Somehow I consistently feel like I've done something clever when I avoid a begging card in Agricola, but it's extremely rare to be forced to take one. I'm not sure if this is my mentality, but the game is highly praised enough that I'm willing to attribute it to some hijinks on the part of the designer.
Penalties are also great because it feels good to subvert them. I've had this experience in Sentinels of the Multiverse where henchmen or villains have launched attacks that are intended to be punishing, but are turned into healing or reflected as damage or somehow launches into a combo that blows up everything.
So carrots are good, but sticks can give us highs that are just as good if used well. Aiming to make players feel great is certainly good advice, but as any good S&M practitioner will tell you, feeling great doesn't just come from nice things happening.