Tuesday, April 29, 2008

Design Lesson 101 - Grand Theft Auto III

- This week, with the release of Rockstar's Grand Theft Auto IV imminent, I look back at its predecessor, Grand Theft Auto III.

Design Lesson: The ability for landmarks to sufficiently guide the player around the game world dramatically decreases as the size of the world increases.

Grand Theft Auto III is considered by most to be a landmark in video gaming history. It was one of the first, and certainly the most successful, 3D open-world action game at the time of its release in 2001. By moving away from the 2D world of the previous games and moving to full 3D, Grand Theft Auto III ushered in sets of new interactions that could not have occurred in a 2D setting. It also ushered in a complexity in navigation, which would not exist in a 2D setting.

The game world of Liberty City spans three large islands, with each island having multiple districts, such as Chinatown and the Red Light District. Each district has a number of distinct visual landmarks to help navigate you. Seeing a familiar, distinct landmark in-game (such as a casino or airport) is a fantastic way to orient the player to their surroundings and help navigate them, all the while sensually immersing them in the game. In fact, landmarks are a large part of how we navigate in real life (at least for those of us without fancy GPS navigation systems in our cars).

When you consider that the three main islands of Liberty City connect to each other at one point each, it becomes even more important to know one's location when trying to get to a new area in the game. Getting lost can be frustrating and take up valuable time. So the landmarks should fix the problem, right? They should help orient the player and get their bearings straight when they are lost, no?

Not quite, unfortunately. The landmarks help to an extent, especially the ones near high traffic areas such as your safehouse and main mission givers, but the world is just far too large to navigate just from landmarks. Think about driving around a new city. It's usually not hard to figure out the major highways and how to get to and from your house. Everything else, however, takes a while to learn. In games, we don't have time for the player to take a while to learn. If we frustrate players early, they may never come back and play the game.

There are many reasons for it being easy to get lost, but the one as it applies to Grand Theft Auto III is information overload. The game slowly introduces you to all of its areas, which is an ideal way to teach the player of new areas. However, it's far too much information to process at once. It's easy to forget about landmarks or confuse exactly where a landmark means you are spatially.

I would forget exactly what street the gun store was on repeatedly, thereby ensuring I would spend five to ten minutes trying to locate the store that was a block away from my starting point. I would remind myself to remember where the gun store was for next time, take a mental note of where I am, and then the next time I was to return I would remember. Except, I usually ended up forgetting. Damn my stupid brain!

The game offers solutions. One is a mini-map, which tells you where your current objective is. This helps for navigating general direction and help pointing you to new areas of the game. The problem is, if the objective is far away, it only appears at the edge of the mini-map. Also, the objective is only for the current mission. If you are looking for a certain area in game that isn't specifically your goal for the current mission, too bad.

Another problem with the mini-map is that it doesn't tell you where things are relative to one another. I know that the baseball stadium is by the shoreline, but where is the gun store in relation to the baseball stadium? That is far more difficult to grasp spatially, and by the time you figure it out, you're probably close to being done playing the game.

Enter the second solution: the names of districts appears on the screen as you pass into them. This, in theory, should teach the player what districts connect to each other. While it succeeds at that, to an extent, it's still not enough. While I may know that Chinatown connects to Red Light District right away, it took far longer to understand that it connected by traveling north.

So, up comes a third solution: a map of the entire city, top-down, with all of the key locations identified. If you can at least look at an entire map at any time, then it should be a lot faster to learn the layout of the city. This should solve all the problems.

Except, the full map comes as an insert in the game manual and on the back of a poster inside the case of the game; it does not exist in the game. Relying on out-of-game materials to play a game effectively is a poor choice. You can't be sure the player knows about them or hasn't lost them. By using an out-of-game map, I cannot know exactly where I am in the game world at any time. Instead I must look at the mini-map and extrapolate enough data to pin-point my position.

There is no reason that the game could not have at least had a full map of Liberty City in one of its menus. This would allow the player to see their current position, figure out where they are trying to go, and make a mental map. It would relate things spatially that are distant, making macro navigation far easier. The best part is, it's simple to implement.

Liberty City is full of identifiable landmarks that help guide the player, but the city is just too large to navigate without some support. Had the game been the size of one island, this wouldn't be an issue. Within five hours of gameplay, you would know where everything was and not have any trouble navigating at all.

However, the game itself would suffer from a smaller city, so the decision was made to have a large, expansive city. The game needs to support that expansive city with the tools to the player to be able to explore that city without major problems. It's okay to include mini-maps and full maps in your game if it will empower the player to be more successful at the game and alleviate some of the frustrations. As game worlds grow, this navigation complexity will grow and become more and more important to solve.

Bonus Lesson: Just because it's the last mission in the game, doesn't mean you have to make it ten times harder than any other mission in the game.

Seriously, Rockstar, you are killing me. I spend all weekend working up to the finish of the main story missions and then you drop that final level on me. I really wish games wouldn't do this.

The end of a game doesn't have to be a push-over, but levels that take hours and hours to complete are just annoying. The end of the game should appropriately scale with the rest of the game. The second hardest level in the game should be close to the hardest (last) level in the game. It didn't feel like this was the case in Grand Theft Auto III. The final mission felt substantially more difficult than any other mission.

3 comments:

  1. It`s funny how some of these things do stick. After all this time I can still remember the route from the first hideout to the first gun shop. The game has taught me and now I will never forget ;D

    And about difficult ending levels: anyone remember Driver?

    congrtz on the Gamasutra gig, keep up the interesting writing..

    ReplyDelete
  2. I thought GTA III's last level was pretty well balanced -- hard, but you could get through it after a few shots.

    GTA San Andreas final level sent me absolutely nuts. I uninstalled the game in disgust and never played it again... spending 5 minutes trailing the fire truck only to be slightly out when the guy fell off the back and having to do it all over again is not fun at all.

    ReplyDelete
  3. That Bundesliga promises a whole lot of action at the moment, with 5 teams ideal for competing in the prize. Germany's tough showing around the 2006 Universe Cup helped reduce league with the eyes of a lot of spectators, and it has become alot more competitive thus. www.scorespro.com

    ReplyDelete