I still remember saving up for Zelda and finally opening up that gold NES cartridge. IMO the original Zelda and the follow-up SNES version are among the best games ever made.
The opportunity to explore and build your character was amazing for its time. The replay value was so high opposed to other popular NES games that required you to start from the beginning each time you turned it on. Zelda is a great case study for new game designers to learn from.
This type of analysis can be performed for any of the early Nintendo-designed NES games. They knew what they were doing. The first few levels of Super Mario Brothers are ingenious at how they subtly teach the player the context of the world and the physics of moving Mario. It is the artistry of these subtle touches that distinguished these games from ordinary, competent platformers long after they'd been knocked off to oblivion.
"In most modern games, there would be fewer enemy types and the rooms would ramp in intensity by combining monster types together."
This is a very interesting observation. It seems that modern open-world RPGs (Skyrim, for instance) have opted for larger and more elaborate worlds at the expense of enemy variety. This is probably the result of the "territory race" that has been taking place in the genre for well over a decade now: each new RPG is expected to have a world twice as big as the last. The player's sense of novelty, and of world scope, is more easily cued by landscape ("room") variety than by enemy variety -- and so, focusing resources on landscape variety usually yields better payoffs.
Even still, I can't help feeling that something has been lost along the way. While I am extremely pleased with the immensity and variety of modern-day world maps, I do notice the emptiness of them, or the repetition of NPC and enemy types. (While enemy level-scaling is a nice feature that keeps enemy encounters relatively fresh, it can also be a crutch that keeps developers from thinking about variety of encounter design).
"It is possible to achieve the feel of non-linear level design by taking a linear path and adding short offshoots."
also called a "maze".
"As with the other levels, there is a minimum amount of re-traversal required to get through this level, as the critical path is extremely linear."
"The flow is generally linear and ramps well, as with Level 1, but the designers stop the player in Room 6 and won't let him continue until he gets the ladder in Room 8."
"One surprise was that the silver arrow is not technically on the critical path even though you can't beat the end boss (Ganon) without it."
So we see that these mazes are linear, except in parts where they have branches, and some branches are required, and there are a few loops.
So, what exactly is the point of this analysis? That this great game has structure, and isn't a completely open world? OK.
Why is minimal dead space (i.e. space not on the critical path) not a design criteria? In level 9 (in the appendix[1]), for example, there seems to be a large amount of space that isn't on the critical path at all. I understand that this increases the non-linearity of the level, but it seems to me that it would also significantly increase the amount of re-traversal required to get through the level.
Level 9 is a war of attrition between the player and the dungeon. The dungeon is intentionally designed with a lot of non-critical areas in order to get the player lost, so that he'll run across more sets of enemies, and eventually be ground down to a point of insufficient resources (health, potions, ammo, etc.) to complete the level. In this sense, beating Level 9 is really an exercise in efficiency, rather than the free-roaming exercise of exploration that carried the player through previous dungeons.
There's also a very subtle psychological dimension at play in Level 9. By the time the player gets to Level 9, he is so used to random exploration, bombing of random walls, discovery of secret areas, and so forth, that his natural instincts are to bomb every wall and explore every room. It's precisely these inclinations that will land him in trouble in Level 9. Almost every wall is bombable; almost every room seems interesting; almost every path is interconnected. Accordingly, in order to beat Level 9, the player must "unlearn" a lot of the behaviors that made him successful, throughout the game, up to that point.
The amount of entertainment I got out of these classic games is in many ways the same, if not more than I get out of modern RPG and FPS games. The graphics have gotten way, way better but the fundamentals that make it fun are the same. In fact some modern games, even though they look great, they just aren't as much fun to play.
Those old games seem a little slow now because the bar has been raised as far as speed and eye candy. But the amount of thought and craft that went into them made them so much fun to play.
The comment on the article about all the maps fitting together reminds me of html/css sprites. clever use of 1 file for multiple uses. great article. never knew this stuff...
Not exactly. A rectangular world that wraps around on both axes is indeed a torus, and appears in many video games, but not Zelda. Hyrule in each of the first four Zelda games in 2D is a flat rectangular world with impassable edges. (Perhaps a game cheat device could let you walk through a wall and wrap around the edges; it's arguable whether that would mean the world is toroidal.)
The opportunity to explore and build your character was amazing for its time. The replay value was so high opposed to other popular NES games that required you to start from the beginning each time you turned it on. Zelda is a great case study for new game designers to learn from.