It almost doesn't matter how interesting the level you build is, if there's no goal in it other than the finish line, it will get boring pretty fast. In Spelunky, there are three different goals to aim for, asides from finishing the game, and each one ties to a different randomly generated item in the levels.
The first, and most obvious goal is to collect treasure. Once the terrain has been generated, treasure items of varying kinds are sprinkled across it like delicious candy, and it's Spelunkyman's job to go and pick it up. Each different peice is worth a different amount of money/points, which can be spent as you progress through the levels at shops or to create checkpoints, or stached up to create a high score, which in turn unlocks it's own rewards, which I'll get to later.
But even with collecting treasure, it's not quite as simple as that. The loot is spread out pretty evenly throughout the floor, but it's not always just laying out in the open. There are pots, which usually contain treasure but occasionally also a snake or spider (so you have to be careful when you look inside), there are chests which may require keys found elsewhere in the level, and there are shiny golden idols, which are important into and of themselves. Not only is each idol worth a staggering 15,000 points, almost five times as much as any other treasure item in the game, they present two distinct challenges. The first challenge is to remove it without dying. Like in Raiders of the lost arc, when Indiana Jones picks the idol off the pedestal and has to outrun a giant boulder, different areas have a number of different traps guarding their treasures. From giant rolling boulders to floors collapsing into piranah pools, and even the dreaded falling ceiling, each one must be escaped simply to *get* the idol in the first place. Then, as a secondary challenge, unlike the rest of the treasure, you have to navigate the rest of the level and physically bring it to the exit door before you get your reward. This means you can't attack (short of throwing the idol, which works, but may cause you to lose it), you can't drop bombs or use ropes, and you have to be very careful dropping off of high ledges because you can't grab onto most ledges without dropping the idol. The reward is great, though, because often an idol will be enough to more than double the score you've collected elsewhere in the level. And since you're shooting for about a hundred and fifty thousand points in one run for the top treasure challenge, it's a hearty bite.
The second challenge is also as you might expect, it's to kill monsters on your way through. Each monster has a distinct way of moving and attacking (typically just bumping into you), and you have to be very careful to rack up a hundred and fifty kills throughout the levels. You can't always just throw them into a trap, you have to brutally murder them yourself by landing on their heads or smashing them with your whip. Usually.
Each terrain type has a handful of monsters in it that will appear as the level is generated. The basic caves will have snakes, spiders, bats, and cavemen running around in it, all of which are pretty simple to predict and to smash, although the cavemen take a handful of hits to do away with permanantly. The jungle has frogs, piranahs, monkeys and yes, man eating plants. The ice caves have yetis and aliens, and if you break open a block of ice, you can uncover frozen cavement too. Finally the ancient temple has cultists, and a mix of monsters from all the other levels, just to keep you on your toes. They're placed randomly too, and it's an occasional inconvienence to find and reach, or epecially kill them, but it's always tempting to wreak havok as you pass through.
But there's one more twist on monsters, and that's special levels. In special levels, the normal rules for generating monsters are put on hold for special rules. There are bosses for almost each type of monsters, giant spiders, fish, aliens, yeti or the feirceome giant mummy, and a few special levels for the others - snake pits, undead monsters, and dark rooms (which are like boss levels for traps). Bosses are still only worth one monster despite their eighteen hit points (to a normal monster's one, or occasionally three) but always drop two things: treasure and items. Treasure is great for obvious reasons, but items make it worth to kill the bosses, because for the rare time you generate an item shop, do you really want to spend two levels of treasure just to grab the high-jump-shoes? Bosses all have special areas and attacks all to themselves, and bypassing them is almost always easier than fighting them, but the occasional fight against something so giant is fun, challenging, and rewarding - making this more than a little side-quest.
Finally we come to the last type challenge in the game, rescuing trapped damsels. Damsels will appear on about every second level you play, and can appear absolutely anywhere, giving them equal chances of appearing on flat ground or on a hard-to-reach ledge. Once you rescue them, you have to carry them all the way to the exit (often harder than not), and you can't even throw them like you can the idols, as they only have a few hits before you kill them. Damsels are fragile after all.
But damsels offer you one of the best rewards in the game, even outside of attempting to rescue nine of them in a fifteen level game; each time you successfully rescue them, you gain an extra hit point. You start with four, and this is the only way to empower youself with more, making the damsels expressly important to surviving to the end of the game. Then again, if you're really good you might want to sacrifice the damsel at one of the occasional sacrificial altars in the game, to gain favor with the gods. You gain different types of rewards for this, but they can be equally tempting.
If you eventually collect enough treasure, kills, or damsel rescues to get a high score in a particular area, you unlock the ability to play special challenges, or to play *as* different characters throughout the game. Those rewards are fantastic, because the different characters each have their own tricks, like the storekeeper's shotgun, or the shortcut guy's ability to dig through solid rock.
Every reward in this game really feels like you've earned it.
Monday, October 26, 2009
Friday, October 23, 2009
Run, jump, plummet...
The level design in Spelunky - yes, even though a computer is doing it, it's still level design - at first looks totally random. You have long strethes of tunnel, deep craters. criss-crossing passageways and more, but even with a really clever program it would be almost impossible to create an actually playable level if you're using totally random level generation.
Note that all I'm going to talk about here is reverse-engineering the game, and may or may not be the way it's actually done. Some bits are more obvious than others, but I make no assumption that I know what's in the code.
The first thing that goes into the level is the path to the exit. You start with about four floors worth of map, with the entrance in one of the top corners and the exit in one of the bottom corners. Then, paths are cut between the different layers, they only need to be a single space wide, but we'll get to the variations in a minute.
Once a basic path is created, terrain is put in. After playing the game for some time I can see that there are a number of different terrain blocks that get used, most of them one floor high and twenty or thirty tiles wide. There are tiles that contain entrances, tiles that contain exits, tiles that contain paths to the next floor, tiles that contain idols, and other tiles that contain nothing but rocks, so far. (What's an Idol, you ask? I'll cover that next post, but it's a game mechanic). There are some special peices of terrain that cross more than one floor worth of space, and others that cross less than one, but for the most part, it's simple cut and paste from the database.
But that would make for a boring game, wouldn't it? Yes it would, but because Spelunky is great, the terrain doesn't stop it's generation there. First, on each terrain peice, there are a number of blocks that may or may not contain a tile of terrain. Sometimes, this allows you extra tunnels, other times it makes jumps harder, and still other times it gives the monsters distinct advantages over your character, for example in a tight tunnel with nothing but a whip (usually) spelunkyman really has a tough time overthrowing a yeti, who has to be jumped on to disable.
These little variations add an almost infinite amount of varaition to the game, even before you consider the enemies, treasure, and other challenges that are generated after the terrain, and especially before you consider the number of different interlocking art tiles that make even identical hallways look just a touch different. You really will play a different game every time you play, the occasional four-level deep pit or endless floor of spiked death mixing up what might be a standard crawl through.
Note that all I'm going to talk about here is reverse-engineering the game, and may or may not be the way it's actually done. Some bits are more obvious than others, but I make no assumption that I know what's in the code.
The first thing that goes into the level is the path to the exit. You start with about four floors worth of map, with the entrance in one of the top corners and the exit in one of the bottom corners. Then, paths are cut between the different layers, they only need to be a single space wide, but we'll get to the variations in a minute.
Once a basic path is created, terrain is put in. After playing the game for some time I can see that there are a number of different terrain blocks that get used, most of them one floor high and twenty or thirty tiles wide. There are tiles that contain entrances, tiles that contain exits, tiles that contain paths to the next floor, tiles that contain idols, and other tiles that contain nothing but rocks, so far. (What's an Idol, you ask? I'll cover that next post, but it's a game mechanic). There are some special peices of terrain that cross more than one floor worth of space, and others that cross less than one, but for the most part, it's simple cut and paste from the database.
But that would make for a boring game, wouldn't it? Yes it would, but because Spelunky is great, the terrain doesn't stop it's generation there. First, on each terrain peice, there are a number of blocks that may or may not contain a tile of terrain. Sometimes, this allows you extra tunnels, other times it makes jumps harder, and still other times it gives the monsters distinct advantages over your character, for example in a tight tunnel with nothing but a whip (usually) spelunkyman really has a tough time overthrowing a yeti, who has to be jumped on to disable.
These little variations add an almost infinite amount of varaition to the game, even before you consider the enemies, treasure, and other challenges that are generated after the terrain, and especially before you consider the number of different interlocking art tiles that make even identical hallways look just a touch different. You really will play a different game every time you play, the occasional four-level deep pit or endless floor of spiked death mixing up what might be a standard crawl through.
Thursday, October 22, 2009
Exposition Expose.
From the opening screen of Spelunky, everything is randomly generated. The story too, although it is only three lines of setup.
It goes to show how little story is actually needed to make a good game great. Three lines transforms what might have been a fun little pointless platformer into what feels like an epic story, and the three lines aren't even related to what happens in the game, other than for exposition.
The exposition in Spleunky works through all the old movie cliches. Fate guiding one's steps, father's last words, mysterious voices, native guides, and so on. It fits the game's feel, diving into an underground tunnel to search for buried treasure so well that it makes this game feel like an old action movie - indiana jones in particular.
One procedurally game I particularly like is ADOM, Ancient Domains Of Mystery, and while the game is fantastic in many ways it's exposition is one of the simplest of them all, a fixed bit of information tied to a randomly generated story filled in with bits from your character creation. You were born in X, you grew up Y, your parents were Z. You trained to be a fighter and left for the draklor chain where chaos was threatening to destroy the world. A great story, and simple.
Procedurally generated games without any in-game story, like rogue or nethack always feel lacking. They become a world, and while adventuring through them is a number of rewards in and of itself, it's never quite satisfying. Why did you go into the Dungeon? Who were you before? Are you a champion of justice, a madman, or a greedy delver? Nothing is answered without just a touch of exposition.
It goes to show how little story is actually needed to make a good game great. Three lines transforms what might have been a fun little pointless platformer into what feels like an epic story, and the three lines aren't even related to what happens in the game, other than for exposition.
The exposition in Spleunky works through all the old movie cliches. Fate guiding one's steps, father's last words, mysterious voices, native guides, and so on. It fits the game's feel, diving into an underground tunnel to search for buried treasure so well that it makes this game feel like an old action movie - indiana jones in particular.
One procedurally game I particularly like is ADOM, Ancient Domains Of Mystery, and while the game is fantastic in many ways it's exposition is one of the simplest of them all, a fixed bit of information tied to a randomly generated story filled in with bits from your character creation. You were born in X, you grew up Y, your parents were Z. You trained to be a fighter and left for the draklor chain where chaos was threatening to destroy the world. A great story, and simple.
Procedurally generated games without any in-game story, like rogue or nethack always feel lacking. They become a world, and while adventuring through them is a number of rewards in and of itself, it's never quite satisfying. Why did you go into the Dungeon? Who were you before? Are you a champion of justice, a madman, or a greedy delver? Nothing is answered without just a touch of exposition.
Wednesday, October 21, 2009
Spelunking
So, I've been playing this game lately, called Spelunky. It's a freeware downloadable game by Derek Yu, and it is to nethack what mario is to zelda.
For those of you without extensive knowledge of all the gaming world, I'll clairify that little metaphor. Zelda is an expansive top down two dimensional world where you collect items, fight monsters, and generally be a sword-weilding menace. Every map in the game is lovingly touched by dozens of designers, artists and programmers to make the game perfectly balanced and nice to look at. Nethack is a procedurally generated dungeon crawler where the maps are randomly generated (sphagetti dungeons, unfortunately), and you collect weapons to kill monsters and generally be a sword-weilding menace. See the similarities?
By that extension, Mario is a two dimensional side-scrolling platformer game where you go through several worlds of varying types, jumping on the tops of enemies, hopefully collecting powerups and uncovering shortcuts. Spelunky, now that I've finally gotten to it, is a procedurally generated two dimensional sidescrolling game where you travel through several levels of varying types, jumping on enemies, trying to collect powerups and unlock shortcuts. The similarities are obvious, to both nethack and mario, and I feel this game is a stunning and essential addition to both today's gaming climate and the world of procedurally generated games.
You see, this game is good, and not just pong good, I'm talking will probably make it's way eventually to a console good; and while it's only been out for a month or so, the important telltale sign is that a community has sprung up around the game, spaders, overaccheivers, version testers and more. You don't get that on an average play-once-and-discard game.
Most importantly, Spelunky is a solid peice of proof that dungeon crawlers are not the only type of game that can be successfully procedurally generated. You could, theoretically, procedurally generate an entire game of *any* genre, and while the obvious place the industry started is on the simple end of the scale, I'd wager that eventually you're going to wind up with more complex games for more complex genres.
For those of you without extensive knowledge of all the gaming world, I'll clairify that little metaphor. Zelda is an expansive top down two dimensional world where you collect items, fight monsters, and generally be a sword-weilding menace. Every map in the game is lovingly touched by dozens of designers, artists and programmers to make the game perfectly balanced and nice to look at. Nethack is a procedurally generated dungeon crawler where the maps are randomly generated (sphagetti dungeons, unfortunately), and you collect weapons to kill monsters and generally be a sword-weilding menace. See the similarities?
By that extension, Mario is a two dimensional side-scrolling platformer game where you go through several worlds of varying types, jumping on the tops of enemies, hopefully collecting powerups and uncovering shortcuts. Spelunky, now that I've finally gotten to it, is a procedurally generated two dimensional sidescrolling game where you travel through several levels of varying types, jumping on enemies, trying to collect powerups and unlock shortcuts. The similarities are obvious, to both nethack and mario, and I feel this game is a stunning and essential addition to both today's gaming climate and the world of procedurally generated games.
You see, this game is good, and not just pong good, I'm talking will probably make it's way eventually to a console good; and while it's only been out for a month or so, the important telltale sign is that a community has sprung up around the game, spaders, overaccheivers, version testers and more. You don't get that on an average play-once-and-discard game.
Most importantly, Spelunky is a solid peice of proof that dungeon crawlers are not the only type of game that can be successfully procedurally generated. You could, theoretically, procedurally generate an entire game of *any* genre, and while the obvious place the industry started is on the simple end of the scale, I'd wager that eventually you're going to wind up with more complex games for more complex genres.
Sunday, October 18, 2009
Zzigy the Warrior
One thing that often bothers me is random name generation. Almost every 'random' name generator program anywhere is actually a single random number generator that picks a name out of an existing list. So, everyone who can't think of a name for a familiar in say, kingdom of loathing, winds up with one of four names, not an *actual* random name.
Names are complex things, I can give that up. Michael isn't exactly a word you would nicely find by putting in seven random letters, and words are generally more complex than that in the first place. So what can be done? Well, dwarf fortress has a good start to it (although like everything in the game it's way too complex). Each dwarf you randomly generate has a number of spaces in their names, and each one can be filled with one of a group of name sections like "shu" or "Meh". Unfortunately, almost all of them are from the same list, so it is possible for you to wind up with Shushu Shushu as a name if you're really unlucky. Each of the names has it's own description and if you're looking to build a dwarf name you can read them. But there's hundreds of them, and that's what makes the names seem random. Combined, you wind up with a few tens of thousands of combinations and that's what you're picking out of, but still, only a small fraction of them are both readable or interesting.
I've started fooling around with another program that generates random strings of characters. It's nowhere near random name quality, but it's certainly better than coming up with "Gort" or "Shushu" every thirty seconds. So far it generates two to five sections composed of one vowel and one consonant, in no particular order. There's about a ten percent change of getting a fantasy-name-joiner in there too, like a ' or a - . You get a lot of unreadable gibberish, but nothing unpronouncable save the rare cases where you have x's matching up and that's *still* not that bad. I'll let you know as it continues to shape up.
And for anyone wondering how tinyworld is coming along, I've stripped out all the code I had for making multiple floors, and am attempting a higher-level way to work it out. I'll keep you up to tabs if I ever get that working.
Names are complex things, I can give that up. Michael isn't exactly a word you would nicely find by putting in seven random letters, and words are generally more complex than that in the first place. So what can be done? Well, dwarf fortress has a good start to it (although like everything in the game it's way too complex). Each dwarf you randomly generate has a number of spaces in their names, and each one can be filled with one of a group of name sections like "shu" or "Meh". Unfortunately, almost all of them are from the same list, so it is possible for you to wind up with Shushu Shushu as a name if you're really unlucky. Each of the names has it's own description and if you're looking to build a dwarf name you can read them. But there's hundreds of them, and that's what makes the names seem random. Combined, you wind up with a few tens of thousands of combinations and that's what you're picking out of, but still, only a small fraction of them are both readable or interesting.
I've started fooling around with another program that generates random strings of characters. It's nowhere near random name quality, but it's certainly better than coming up with "Gort" or "Shushu" every thirty seconds. So far it generates two to five sections composed of one vowel and one consonant, in no particular order. There's about a ten percent change of getting a fantasy-name-joiner in there too, like a ' or a - . You get a lot of unreadable gibberish, but nothing unpronouncable save the rare cases where you have x's matching up and that's *still* not that bad. I'll let you know as it continues to shape up.
And for anyone wondering how tinyworld is coming along, I've stripped out all the code I had for making multiple floors, and am attempting a higher-level way to work it out. I'll keep you up to tabs if I ever get that working.
Tuesday, October 13, 2009
Mini Gaming
Mini games are one of the tender areas inside gaming that nobody is quite sure what to do with. Sometimes the mini games are as simple as pushing one button a hundred times in as short a time as you can. Other times it's more complicated and integrated with the controls of the rest of the game you're playing. One of the most popular methods of minigaming is the quicktime event.
Quicktime events are, to quote a narcissistic australian I'm sure you know, times when you are forced to press X to not die. It's simple, often unexpected, and can bring about long periods of gameplay you are forced to redo, because no matter how skilled you are, or how much time you put into training, if you fail to quickly and properly react to a particular series of all-but-random events, you are brought to the game over screen. So why are they so popular? Because they're easy to do! Programmers have to take like, fifteen seconds to code in 'if this button is not pressed by this time, the player dies', while if they had made it an actual in-game challenge it would have taken them hours, weeks or days to build it.
Allow me to say that I think minigames are absolutely required in today's gaming world. In shorter games they can provide some pacing as well as adding in length for much cheaper than standard levels, and in longer games they can provide a fast-paced moment of pickup much needed in slower sections of the game, especially if they provide in-game rewards for doing them.
Just so long as you don't go gold saucer on us, requiring hours upon hours of play to get any reasonable rewards.
Quicktime events are, to quote a narcissistic australian I'm sure you know, times when you are forced to press X to not die. It's simple, often unexpected, and can bring about long periods of gameplay you are forced to redo, because no matter how skilled you are, or how much time you put into training, if you fail to quickly and properly react to a particular series of all-but-random events, you are brought to the game over screen. So why are they so popular? Because they're easy to do! Programmers have to take like, fifteen seconds to code in 'if this button is not pressed by this time, the player dies', while if they had made it an actual in-game challenge it would have taken them hours, weeks or days to build it.
Allow me to say that I think minigames are absolutely required in today's gaming world. In shorter games they can provide some pacing as well as adding in length for much cheaper than standard levels, and in longer games they can provide a fast-paced moment of pickup much needed in slower sections of the game, especially if they provide in-game rewards for doing them.
Just so long as you don't go gold saucer on us, requiring hours upon hours of play to get any reasonable rewards.
Thursday, October 8, 2009
See Dungeon Runner... Oh, forget it, it's a boring joke anyways.
When randomly generating a dungeon, one thing you have to think about is the challenge of it. Is the average player going to be able to breeze through it? Will only the most experienced and quick-witted be able to pass? Are there going to be some challenges harder than others? Some challenges that allow the player to take a figurative breather? How quickly do they get more challenging, if at all?
Dungeon runner takes a very simplistic look at this, and like the shapes of the rooms that eventually get stitched together, the monster encounters are basically cookie-cutter fights, although even a cookie cutter can surprise you sometimes.
Basically, monsters are divided up into groups. Typically there is exactly one group per room, and they spread themselves out fairly randomly. That only matters on the first shot, though because the AI is extremely basic. If you kill a monster in one shot without it seeing you, you don't draw agro. If you don't kill it in one shot, regardless of whether the rest of the monsters in that group are able to see you, or the proximity you are to other, different groups, you draw aggro from exactly every monster in that group and no others. (in case you don't know, drawing aggro means to make a monster aggrivated, and chase you to attack). That's a fine system, but there are some 'scout' monsters who random-walk way outside of the room they are grouped in, allowing you to accidentally double-up a fight even when you're already losing.
There are several simple 'shapes' I've seen these cookie cutters come in. Medium, Swarm, and Boss.
Medium groups contain a half-dozen average melee units and one or three ranged units. Once in a while they'll contain a ranged unit who can heal your opponents or poison you or the like. Mostly it'll just be one knockdown per melee opponent, and then fight until they're all dead. No real variation, minimal strategy. It's not even much different based on what type of enemy you're fighting. All the melee enemies can knock you over exactly once. I suppose it's a charging bonus or something.
Swarm groups are larger than medium groups and in addition to a small handful of melee guys and a larger handful of ranged guys, they contain about ten or fifteen fodder guys whose only purpose is to absorb your fire and get in your way by getting as close to you as possible and doing the smallest amount of damage possible for you to want to kill them all. There's still not a lot of strategy here, you use an area attack once they get in close, then kill each of the fodder guys in one hit. It doesn't hurt that everyone gets area attacks of every element, so you can do it no matter who you're fighting. I'm sure there's other strategies, but this one works and it's simple.
Now we get to the boss fights, which is where the strategy actually comes in. Boss fights typically contain a good number of melee units, a few ranged units, and a boss enemy selected from among the available enemies that you're fighting. That does include the fodder units, but that just means they do less damage, they still have a huge number of hit points. Now, you actually have to prioritize your targets, because if you just attack the boss, then the other units will kill you. However, ignoring the boss will lead to the opposite scenario, forcing you to respawn and once again meet up with the now lackey-free boss and wipe the floor with him. I've found the best strategy is to take out about half the flunkies, then the boss, and clean up the flunkies last. It's simple, but the fact that the bosses have names, drop more loot than normal, and guard treasure chests puts just enough variation in the kill ten enemies, wait thirty seconds for health to recharge. Kill then enemies...
Of course, the names are 'randomly generated', too. Or, randomly picked from a list for that particular group. I'll never forget the first "Particuarly Rude Broodling" or "Pyrus the Smoldering" I fought. At least, until January when the game clears away.
Dungeon runner takes a very simplistic look at this, and like the shapes of the rooms that eventually get stitched together, the monster encounters are basically cookie-cutter fights, although even a cookie cutter can surprise you sometimes.
Basically, monsters are divided up into groups. Typically there is exactly one group per room, and they spread themselves out fairly randomly. That only matters on the first shot, though because the AI is extremely basic. If you kill a monster in one shot without it seeing you, you don't draw agro. If you don't kill it in one shot, regardless of whether the rest of the monsters in that group are able to see you, or the proximity you are to other, different groups, you draw aggro from exactly every monster in that group and no others. (in case you don't know, drawing aggro means to make a monster aggrivated, and chase you to attack). That's a fine system, but there are some 'scout' monsters who random-walk way outside of the room they are grouped in, allowing you to accidentally double-up a fight even when you're already losing.
There are several simple 'shapes' I've seen these cookie cutters come in. Medium, Swarm, and Boss.
Medium groups contain a half-dozen average melee units and one or three ranged units. Once in a while they'll contain a ranged unit who can heal your opponents or poison you or the like. Mostly it'll just be one knockdown per melee opponent, and then fight until they're all dead. No real variation, minimal strategy. It's not even much different based on what type of enemy you're fighting. All the melee enemies can knock you over exactly once. I suppose it's a charging bonus or something.
Swarm groups are larger than medium groups and in addition to a small handful of melee guys and a larger handful of ranged guys, they contain about ten or fifteen fodder guys whose only purpose is to absorb your fire and get in your way by getting as close to you as possible and doing the smallest amount of damage possible for you to want to kill them all. There's still not a lot of strategy here, you use an area attack once they get in close, then kill each of the fodder guys in one hit. It doesn't hurt that everyone gets area attacks of every element, so you can do it no matter who you're fighting. I'm sure there's other strategies, but this one works and it's simple.
Now we get to the boss fights, which is where the strategy actually comes in. Boss fights typically contain a good number of melee units, a few ranged units, and a boss enemy selected from among the available enemies that you're fighting. That does include the fodder units, but that just means they do less damage, they still have a huge number of hit points. Now, you actually have to prioritize your targets, because if you just attack the boss, then the other units will kill you. However, ignoring the boss will lead to the opposite scenario, forcing you to respawn and once again meet up with the now lackey-free boss and wipe the floor with him. I've found the best strategy is to take out about half the flunkies, then the boss, and clean up the flunkies last. It's simple, but the fact that the bosses have names, drop more loot than normal, and guard treasure chests puts just enough variation in the kill ten enemies, wait thirty seconds for health to recharge. Kill then enemies...
Of course, the names are 'randomly generated', too. Or, randomly picked from a list for that particular group. I'll never forget the first "Particuarly Rude Broodling" or "Pyrus the Smoldering" I fought. At least, until January when the game clears away.
Monday, October 5, 2009
Form or Function
I'm sure you can all think of a point in time where you were playing a game, running along, and you run across a peice of the story that doesn't fit with the game world. A character dies, for instance, and while you have a dozen scrolls of ressurection in your backpack, you can't save their lives.
The delicate balance of gameplay versus story is the most evident when dealing with resurrection. If you can bring a character back from the dead with no effort, any tension the game might be trying to build is lessened, because any threats are suddenly less so. If resurrection becomes too much of an obstacle, the game becomes hard and unforgiving. Worse still, since you should have more powerful abilities and more resources available at the end of the game than you did at the beginning, there have to be new, more challenging ways to keep the tension in the game.
This particular argument was raised by a discussion on Final Fantasy. The series has been handled well over the years, breaking ground in many different fields, having new stories with common elements and so on. In several of the games, however, you have characters die on you as part of the story. PCs, and NPCs alike. Then, you are jolted out of your suspension of disbelief because for ten hours you've been fighting monsters and getting killed regularly, only to cast a little magic spell (typically called 'life', so there's no confusion as to simply recovering consiousness) to pick your mangled body up off the ground, and it doesn't work in the cutscene you're watching. Or afterwards. You can drag 'dead' party members around for days and ressurect them at the drop of a hat, but if someone dies in a cutscene they're gone for good, and forget about raising that dead NPC king who was the only one who knew where the secret treasure room was.
And that brings us back to the challenge of ressurection in games. If you could raise any NPC from the dead who died, towns would no longer be in threat of being destroyed. The baddies would never be any sort of a threat, because all you'd need to do is get strong enough, buy a million pheonix downs, let them wipe all life off the world, and start casting ressurect!
One way to get around this would be to make there be something special about the PCs. They could have special amulets that allow the bodies to be raised if they fall - but most NPCs don't have that amulet (or whatever it's decided to be). a character could be in real danger if that item or trait is taken away, allowing a game to threaten the characters when needed, but make things easy otherwise. But, would it work?
The delicate balance of gameplay versus story is the most evident when dealing with resurrection. If you can bring a character back from the dead with no effort, any tension the game might be trying to build is lessened, because any threats are suddenly less so. If resurrection becomes too much of an obstacle, the game becomes hard and unforgiving. Worse still, since you should have more powerful abilities and more resources available at the end of the game than you did at the beginning, there have to be new, more challenging ways to keep the tension in the game.
This particular argument was raised by a discussion on Final Fantasy. The series has been handled well over the years, breaking ground in many different fields, having new stories with common elements and so on. In several of the games, however, you have characters die on you as part of the story. PCs, and NPCs alike. Then, you are jolted out of your suspension of disbelief because for ten hours you've been fighting monsters and getting killed regularly, only to cast a little magic spell (typically called 'life', so there's no confusion as to simply recovering consiousness) to pick your mangled body up off the ground, and it doesn't work in the cutscene you're watching. Or afterwards. You can drag 'dead' party members around for days and ressurect them at the drop of a hat, but if someone dies in a cutscene they're gone for good, and forget about raising that dead NPC king who was the only one who knew where the secret treasure room was.
And that brings us back to the challenge of ressurection in games. If you could raise any NPC from the dead who died, towns would no longer be in threat of being destroyed. The baddies would never be any sort of a threat, because all you'd need to do is get strong enough, buy a million pheonix downs, let them wipe all life off the world, and start casting ressurect!
One way to get around this would be to make there be something special about the PCs. They could have special amulets that allow the bodies to be raised if they fall - but most NPCs don't have that amulet (or whatever it's decided to be). a character could be in real danger if that item or trait is taken away, allowing a game to threaten the characters when needed, but make things easy otherwise. But, would it work?
Subscribe to:
Posts (Atom)