2006-09-26

Difficulty Levels In Computer Games

I bought Dead Rising when it came out in the UK and I've been playing it a while but it has got me thinking about the difficulty curves in games. I messed up a few missions in Dead Rising and that made points later in the game more difficult. In turn that escalating difficulty made leveling up more difficult which made more points of the game difficult. For example I failed to kill a psychotic clown, had I killed him I would of unlocked a shortcut to another area. If I had done that then later, when I was trying to save four survivors then I would have succeeded. This would have given me more experience points making it easier for me to finish later points of the game. This is an inversely proportional difficulty, that is the more skilled the player is the easier the game becomes. I'm going to look at this in more detail later but I'll look at a few of the models I've seen in games.

Flat Model Difficulty

A lot of games fall under this category to one degree or another. A lot of Real Time Strategy games are like this (Dawn of War as one example). At the beginning of each level you are back to square one and given a fixed resource to work with. It doesn't matter if you just scraped through the previous level or managed to succeed with flying colours. Generally these games also follow a game trend where the levels become increasingly difficult; it's a game mechanic rather than following more of a real world example where you'd assume the dwindling resources of your enemy would mean that you were in a stronger position as you progress. Many first person shooters also are like this. In Half Life you start each level on a roughly equal footing to any other player. You might have more ammo or health but generally a good game designer will balance this out by providing a stock up area to put people on equal footing.

Often these games will increase the challenge by allowing the player to pick the difficulty (maybe dynamically, or when they start the game). An experienced player can increase the difficulty to match their skill to insure the game is still a challenge for them.

Proportional Difficulty Model

I can't think of many examples of the model. Sin: Episodes is one example. In that, if you die the game begins to reduce the difficulty of the enemies deployed automatically. If you are doing very well the game will increase the difficulty automatically providing a more difficult challenge to players that can handle it. It's important to provide players with the ability to control this themselves as well so a good player that isn't looking for it to become frustrating can still tailor their game. This model helps insure a challenge for all players without, hopefully, becoming frustrating.

It's a good model to implement as it makes the game fun for everyone while not consciously making them change the difficulty. However a developer may find it more difficult to scale events dynamically like this.

Neverwinter Nights also uses this for random encounters though the fixed encounter fights don't scale like this. So players may find it easy (where it is picking lower level encounters) until they get to a scripted fight where they'll find it impossible (because the boss doesn't scale to challenge them correctly).

Inversely Proportional Difficulty Model

Finally we have the model that made me think of this in the first place. This model is understandable as it rewards skilled players and penalizes poor players. The first thing you need for this model is some level of persistence in the game. Grand Theft Auto isn't an example as though you might fail to achieve missions or progress in the game each mission is an instance which has no knock on effect to the rest of the game. In dead rising achieving certain missions will reward you with experience points (giving you new abilities) and routes which will help you in later missions. Long term strategy games such as Rome: Total War and Civilization also create a situation where a bad start might effect the entire rest of your game. Roleplaying games tend to also create a situation in which a good player can hoard items and gain levels easier than a poor player.

Though rewarding people who are skilled is good (as it makes them feel they've achieved something) this is a bad situation. Good players might well find the game too easy and poor players will find the game impossible. The easiest way to rectify this is to offer difficulty levels so the player can find a difficulty level that they are happy with. Another way is to playtest all rewards to insure they don't give a huge long term advantage.

I like this model, it is just a difficult one to balance properly.

2006-09-02

Insomnia

Maybe insomnia setting in should be reserved for my personal blog but failing to sleep made me decide to talk about a game I was playing last night; Bang!. I found it a really fun game and I was a bit surprised at myself for not playing this before as I can be a bit of a Western fan. Anyway the Wiki entry is pretty good at describing it. I found the dynamic of the game interesting. The fact that some people want to defend the Sheriff even if the Sheriff doesn't know that.

Right maybe I'll try that sleep thing again.