Category: Game Programming

Yu Suzuki

May 11, 2003
Cool PeopleGame DesignGame ProgrammingGames
0 Comment

Watched Gamasutra’s newest classic video yesterday. It’s of Yu Suzuki’s 2000 GDC keynote. The video is of the president of Sega of America “interviewing” Mr. Suzuki, though he really just sets things up so Yu can show off his game. “What’s this I hear about 300 characters? What about the fighting system? What is Magic Weather?” To which Yu consistently replies, “I have a video that explains that…”

Not that it was a bad video – not at all, I’m glad I watched it. I just remember those young, heady times – Shenmue was going to change everything. Only it didn’t. And the reason it didn’t is because the game cost too much and took too long to make. While there were some automated systems in the game (such as automated lipsynching) and a fair amount of motion capture, the actual models, textures, and level geometry were all done by hand. And that took about five years. Suzuki himself mentioned that by the time the game was done, his team had grown to over 300 people, and a third of those were Sega employees who were presumably permanently employed. You can’t make a profitable game by paying a hundred people for five years, and I wouldn’t be surprised if Shenmue had the same debilitating effect on Sega that Wing Commander 4 did on Origin Systems.

This is why Warren Spector recently called for the development of tools that could create randomized but realistic level geometry – a tough problem to solve, given how well the human eye picks out patterns and incorrect details. But if such a problem could be solved, games as rich and deep as Shenmue could be made in a much shorter time and with a much smaller team – meaning they could be profitable. Everyone would win – both developers and players.

The other interesting thing about the video was that when Yu Suzuki got back from his inspiration trip to China, the first thing he did – before any models or even sketches were made, before any script was written, before anything else – was to commission musical pieces for Shenmue, so that all the developers could listen to them for inspiration as they worked.

This really caught my attention because on the Spirited Away DVD (which you should go out and buy right now) the “making of” special mentions that Hayao Miyazaki listened to “Always With Me” over and over again during the making of the movie. This song had been created by a musician and lyricist team that had been inspired by Princess Mononoke, and they sent him the song hoping he could use it for one of his movies. He listened to it constantly for inspiration during the making of “Spirited Away” and it eventually became the theme of the movie.

Mike McShaffry once wrote an article for the IGDA called “There and Back Again” (which has since fallen off the web, or I’d link to it). In this article he compared game development to trying to roll a very large rock down a very steep hill to hit a very small target. So of course you want to aim well and push hard at the beginning, because you won’t have much of a chance to affect the rock’s course once it gets momentum. Perhaps having musical themes prepared beforehand could help keep the “rock” on course while it rolls – there’s no doubt that both the Shenmue themes and “Always With Me” are very specific in the images they conjur up.

Jet Set Willy

January 16, 2003
Game DesignGame Programming
0 Comment

I’ve been playing some Jet Set Willy recently.

I never got into the ZX Spectrum much, mostly because it was rare here in the States (as was the MSX, another good older computer). The plot line was simple – Willy has just bought a huge sixty-room mansion. It’s so big that he hasn’t even visited all the rooms yet, but that didn’t keep him from throwing a wild housewarming party for himself. Now all the guests are gone, the house is a mess and Willy just wants to go to bed, but the housekeeper, Maria, refuses to let him until he’s cleaned up every room in the house.

Despite that rather odd plot, Jet Set Willy was well-loved in the Spectrum community. After playing it a bit, I can see why. It was released in 1984, and had two features that were rare in games of that day – determinance and persistence.
Persistence is fairly obvious. The mansion has sixty rooms, each one consisting of a single screen. Every room has a name, and almost all rooms have multiple exits, which lead to other rooms. Almost every room has at least one object Willy has to collect. Therefore, the mansion feels like a mansion – rooms lead to other rooms, and you can go back and forth at will. Which means that it’s possible to map the mansion. And if you pick up an object in a room, that object doesn’t reappear when you revisit the room, because it’s (presumably) in your pocket. Jet Set Willy took persistence one step further – in the space base part of the mansion, there’s a room whose configuration changes based on the rooms the player has visited. Therefore, as you get closer to completing the game, this room fills up with blocks, and the blocks are colored differently depending on whether or not you’ve found every object in each room.

If this doesn’t sound like much, note that Nintendo’s 1984 magnum opus Super Mario Brothers wasn’t persistent at all! Persistence was rare in games in this time period, since it required the machine to keep track of the state of rooms the player wasn’t actually in, and most games didn’t bother. Games that did, like Jet Set Willy, felt more like a coherent experience than games that did not.

Determinance is the opposite of randomness. Whenever the player enters a new room in Jet Set Willy, he is (for the moment) safe. Enemies have predictable patterns, and the player is allowed to stop and watch the enemies before he makes a move. The player won’t ever be killed by some random thing flying in from offscreen that he can’t react to – every part of the puzzle is laid out in front of him and his own skill decides whether or not he can solve it. This doesn’t mean that the game is easy – quite the opposite. But the game is beatable, all it takes is knowledge and skill on the part of the player.

This is another feature that wasn’t common in games of the time. Most computer games were based on arcade games (in spirit if not in actual content). But arcade games are designed to make money, and they do so by making the player lose so he’ll put in another quarter to try again. Therefore it was very common in arcade games to have random things come in and kill the player.

But computer games aren’t arcade games, and Jet Set Willy’s designer, Matthew Smith, realized this. So he threw off that constraint and created a game that allowed the player to take his time and figure the game out.

Both of these advances are important because they allow the player to choose how he wants to play the game. This is vital in my opinion, and I’ll be writing more about it later.