By using this site, you agree to our Privacy Policy and our Terms of Use. Close

Difficulty settings wouldn't be necessary if game AI would work like a good DM in the old table top RPGs. A good DM challenges the party, yet leaves a way out and will adjust difficulty on the fly not to kill off the party and end the game. Games need to figure out how to dynamically scale difficulty to challenge players from start to end without vertical learning curves or obstacles that become frustrating. Unless ofcourse you venture into an area you are not supposed to be in yet.

There are many ways to scale difficulty for example rubber banding has been used in arcade racing games for decades to keep races close. Online lobbies offer boost and increased drafting to keep racers together. The last of us also does something similar by throwing you much needed items when you get low. The AI director in Left 4 dead keeps the players on their toes. Unreal Tournament had a setting where players that run out ahead were handicapped by making them easier to hit, while doing the reverse to players that die a lot. In a game about learning patterns like Dark Souls, enemy speed can be flexible as well as hit windows and damage modifiers. A difficulty setting can determine how much the engine can scale the difficulty or not at all.