A few months ago I talked about the concept of difficulty levels and the right and wrong ways in my opinion of how difficulty levels should be used. For this entry I’m going to talk about the concept of adaptive difficulty and the subtle and not so subtle ways I’ve seen it used in video games.
The simplest explanation is a system that adjusts the game’s challenge level based on player response. The benefits to this is that it smoothes the difficulty curve of the game, the better the player is the more the game will try to match them and the worse the player is the more the game will hold back. This is a more optimal situation compared to games that lock the difficulty level from the start especially when the only thing the level adjusts are the stats of the enemies. Left 4 Dead besides having 4 difficulty levels also has the director which attempts to match the player’s skill level with more or less infected hordes and items. Looking at this type of system I can break it down further and I have some examples of each sub division.
Raising the bar: for this classification, these are titles that use different challenges or game systems to affect the difficulty. Platformer titles like Super Mario Galaxy, Banjo Kazooie,etc are great examples of this. Basically to beat the game you have to do X tasks or gather X collectibles to reach the final stage and beat the game. However there are more things to do then require to beat the game and the player can effectively choose what challenges to attempt and come back to harder ones later. In the latest Banjo Kazooie title every challenge in the game is separated into three tiers of difficulty: bronze, silver and gold. The higher the tier the better the player has to do in the challenge. To complete a challenge on gold requires greater skill then on bronze but it is not required to get gold on every challenge to win.
In Super Mario Galaxy the majority of the challenges were on the easy to moderate side, but later on in the game you unlock harder challenges like the purple coin quests. These challenges are not required but are there to test players later on. Another element of this is when the difficulty of the game affects the game play.
Going way back in time, the original Theme Park made by Bullfrog allowed the player to choose different difficulty levels. Each level the player goes up starting at easy unlocks a new system for the player to understand. For example on easy the player just has to worry about building rides and organizing the park, on normal the research screen is unlocked requiring the player to put money into research to unlock better equipment and rides. Once again the same theme is there, the player doesn’t have to play the game on hard to win, but the options are there for the player if they want.
Next example comes from Perfect Dark although there are more examples of this. In Perfect Dark there are three difficulty levels that you can play on which are classified as easy, medium, and hard. The difficulty level affects two things, first being how much damage enemies do and second it affects the objectives required. Each mission in Perfect Dark has a main objective and several objectives that have to be done on the way. The harder you play the game the more of these objectives are required. I think this is a great way of challenging the player without just out right stat tweaking the enemies into a cheap experience.
My last example comes fromHinterland for the PC. In Hinterland your mission is to clear a randomized map of monster lairs to settle the land. Before the map is created the player can choose different variables which in turn affect the difficulty. Such as how big the map is ,chance of finding resources and etc. With each choice the bonus given to the player’s score at the end is increased or decreased and this allows the player to tailor their game to how hard or easy they want it.
Tweaking along the way: This grouping is for games that allow the player to alter the difficulty of the game in mid play without having to restart the entire game. In some titles like First Person Shooters altering the difficulty will tone down the enemy damage and accuracy while in other games it can affect the game play itself.
In God Hand for the PS2 the difficulty of the game is measured in terms of levels : 1,2,3 and Die. As you play the game if you are constantly avoiding damage and hurting the enemies a bar will slowly rise and once it fills up you’ll go up one level. Consequently getting pummeled or dying will lower your level, the player can also raise or lower the level by taunting or begging respectively. At higher levels enemies will respond quicker, attack in groups ,do more damage and even spawn more enemies in the level.
The reason why you want to make the game harder is that at the end of a chapter you are rewarded money based on how many enemies you fight and their difficulty settings. Money is used to purchase new moves and essential upgrades which mean that in the long run the game can become easier by raising the difficulty level now.
Before I move on I want to stop for a second and discuss another action game and why it is not an example of adaptive difficulty. Ninja Gaiden Black which is one of my favorites set up each difficulty level to rewrite the game. Different enemies and items will appear as you raise the level up. However while the game changes with each increase or decrease it is not a form of adaptive difficulty like Perfect Dark or God Hand due to one fact in my opinion.
You cannot control the difficulty level after starting the game, in my other examples the player in some way can alter the difficulty of the game in mid play. However in Ninja Gaiden Black once the player has picked a level they are locked into that (with the one exception of playing on very easy mode). I can’t played 10 hours on normal and then decide I want to move it up to hard. My last example comes from one of my favorite titles and I can’t believe I didn’t think of it when I first started typing this up.
The World Ends With You for the Nintendo DS was a Square-Enix anomaly and I could write up an article solely on how unique it was but for now I want to talk about its difficulty settings. In TWEWY the player can alter the difficulty of the game in several ways whenever the player is not in a fight.
1. The difficulty level of the game can be raised or lower at anytime. This determines the stats of the enemies and how aggressive they are. A boss on easy won’t be that tough but a regular enemy on ultimate (very hard) is another story. This also affects what items an enemy can drop, every enemy in the game can have different items they will drop based on what difficulty level you are fighting them on. Of course the best stuff comes at the higher levels.
2. Leveling up in TWEWY mainly determines the max health you have going into a fight. The player can drop what level they will fight at from the menu which will of course leave them with less health. The reason you want to do this is that for each level below your max level you lower it to, the greater the chance enemies will drop items will increase. A great touch is that you can see this per monster from the beastery while affecting this. Bosses have a very low chance of dropping anything and unless you’re very lucky your only way of obtaining the best items is to weaken yourself.
3. Lastly when you are running around in the game you can attempt to control how many enemy battles in a row you will have to fight. Fighting more enemies in a row can be tough if you are low on health but the more you fight at once, the more money and experience you will get at the end.
All this talk about adaptive difficulties looks great so far, but there are several issues that need to mention.
First is that it is hard to make a good adaptive system (duh). The designer needs to be on the ball at making it meaningful to play the game at a higher setting and not just for the masochistic players. There is also the fine tuning of making challenges that can be rated “easy, medium or hard”. For games that have the difficulty affect the game systems, time and resources will need to be spent altering the game at each setting. You’re not just making one level, but instead several variations of the same level. Another major problem is the big opponent to adaptive difficulty systems.
Linearity, while a great way of building a narrative is also a major obstacle in creating adaptive challenges. In games like Left 4 Dead and others that have the game constantly trying to adapt to the player’s style, the problem is that when the AI is dealing with a set # of possible item and enemy placements it means that the challenge is not truly adaptive. The AI can only put so many enemies or have so many spots to leave items that eventually the player will be able to exploit it.
When I play Left 4 Dead with my friends who a lot of us are expert players we know where the item spawns are and where enemies can be. At our level of play we know the absolute extent that the AI can make the game difficult and it is not a challenge for us now.
Another argument is how much should the difficulty be swayed by the player’s actions.
There is a fine line between giving the player a challenge and flat out punishing the player for being good at the game. This was one of the complaints people had with God Hand. It can be very hard to learn a game when as the player is getting good that the game decides to crush them with difficulty. Also if things aren’t affected too much then the system is wasted as well. For example if the only thing the AI will control are item drops then it is a moot point for good players as they wouldn’t need those items in the first place.
When adaptive difficulty is done right it can elevate a great game to an amazing title and allows players of different skill levels to enjoy the same game in different ways. Games that find ways of adapting the difficulty also have the advantage of improving replay ability across the board. It also reduces the need to have to use difficulty levels like “medium easy” or “very very hard” when attempting to balance your game out.
Josh
