Here's a brief explanation of why I'm looking forward to particular games.
Halo: Reach
This video shows why. The most common failure of competitive multiplayer modes is a failure to separate people of different skill levels and, just as importantly, different goals. Some people play for the end results... stats, rankings, challenge completion, etc. Others, like me, play more for the experience itself... the grenade toss into a sniper perch, the point-blank shotgun blast that throws the enemy back, getting killed by your own sticky grenade as the enemy runs toward you. Players like me are less concerned with optimal loadouts and strategies because we prioritize fresh and compelling experiences over winning.
Halo: Reach seems to address this through a refined matchmaking system that goes beyond soft separation and actually divides people into tiers. The fanatics and achievement junkies can have their tier and the "casual" players can have theirs. And, again -- just as importantly, the matchmaking system also asks preferences on Chattiness, Motivation, Teamwork and Tone.
Customization is the other reason I'm interested.
Fable 3
I've recently been surprised at how replayable Fable 2 is. The humor, the joyful art style, the skill options and the moral options are all fun many months after my first (and second) playthrough.
Fable 3 seems to have made co-op actually enjoyable by enabling the use of player-developed characters and making the two players' cameras independent. But it's the single-player mode I'm mostly interested in. The weapons that morph depending on how you use them could be great. And I'm anxious to dive into the latter half of the game, which focuses on making decisions as ruler of Albion. Much will depend on whether the moral options are not so limited as to force us into decisions we don't agree with.
Call of Duty: Black Ops
Black Ops is offering more customization, which is good. For the most part, I'm interested in this Call of Duty for the same reasons I liked the last ones, despite the mediocre matchmaking and inexcusable frequency of connection troubles. I love earning new guns and upgrades, and doing so keeps gameplay fresh by encouraging me to change my loadout often.
I was surprised to learn there will be a bot mode. I haven't heard anything about being able to invite another player to join the fight against bots, but it's a start. I'm still dumbfounded by the industry's failure to repeat Perfect Dark 64's brilliant bot mode. In the interview I saw, Treyarch's spokeman explained it as practice for people before they fight other players (it has a separate ranking system). I wonder sometimes if shooter developers think all shooter fans are achievement junkies like themselves. A bot mode doesn't have to be practice. Sometimes, even the most skilled players want a more relaxing game. And some players prefer more casual play all the time.
Whatever the reasons, was Perfect Dark 64's bot mode not as popular as I think it was? Is popularity not why it was freshened up for re-release on Xbox Live? How many people downloaded Monday Night Combat recently? There are obviously a lot of gamers who like playing against bots, so why not meet the demand and make a profit?
On my sidebar, you'll see a list of other games I'm hopeful for. Most won't be out until next year. I recommend taking a close look at Brink.
Showing posts with label AI. Show all posts
Showing posts with label AI. Show all posts
Tuesday, September 07, 2010
Wednesday, November 18, 2009
moving with NPCs
"Follow me!" the NPC says. So I do.
Well, I try to.
When I walk, the NPC is moving faster than me. I fall behind. When I run, the NPC is moving slower, and annoyingly alternates between walking and running because he or she is incapable of matching my pace precisely.
Why is this a problem? Why does it show up in every single game that has me follow an NPC or has an NPC follow me?
Well, I try to.
When I walk, the NPC is moving faster than me. I fall behind. When I run, the NPC is moving slower, and annoyingly alternates between walking and running because he or she is incapable of matching my pace precisely.
Why is this a problem? Why does it show up in every single game that has me follow an NPC or has an NPC follow me?
Thursday, November 05, 2009
gradual indicators (navigation)
Maps are the most common way to guide players around games. Other methods include beacons, landmarks, a compass, trails, and many other things.
Another option I rarely see is what I'll call gradual indicators. By that I mean, a visual feature becomes stronger/weaker or more/less common as the player nears certain places or objects of importance. These gradual indicators can lessen or even negate the need for UI methods which distract from the gameworld.
For example, elevation can be used in this way. In Rise of the Argonauts, the player knows whether he is running towards the palace or the docks of Mycenae by whether the path is sloping upward or downward. The same method can be used in any number of settings, including dungeons and roads. The degree of a slope can be telling, too. If two paths both lead up a mountain, one can be recognizable from the other if it is significantly steeper.
Vegetation can be a gradual indicator. Grass and shrubs might grow tall in one area and shrink or die out toward another. Oaks might dominate one area while mesquite trees dominate another. Both trees might habit an area, but the ratio between them could indicate to the player in which direction he is headed.
Architecture and objects can be used in the same way. The facades of buildings and their relationships to each other (such as how far apart they are) can indicate if the area the player is proceeding toward is wealthy or poor, old or new, damaged or undamaged (by war, weather, or graffiti), and so on. Statues and decorations can change in frequency or in subject through an environment. The type of shops or their names can change in style and tone.
Even audio can lead players. The roar of a waterfall or pounding of waves can gradually rise in volume as the player nears them. Likewise, the sounds of animals, the firing of weapons, the chatter and laughter or shouts of NPCs, or a thousand other things.
Animations and even environmental A.I. can guide players. As a player nears a battlefield, he might increasing notice NPCs crouched over and weeping. Or NPCs might be increasingly erratic and skittish... increasingly distressed as the player nears the source of that distress. The NPCs might pace, look around fearfully, or chatter to themselves. Think of the A.I. in Batman: Arkham Asylum and how the Joker's men act differently as they become more aware of Batman's presence and are afraid.
There are countless forms gradual indicators can take. The goal is to help players navigate environments without having to take their eyes off the gameworld or see anything that takes their attention off the story they're experiencing.
Another option I rarely see is what I'll call gradual indicators. By that I mean, a visual feature becomes stronger/weaker or more/less common as the player nears certain places or objects of importance. These gradual indicators can lessen or even negate the need for UI methods which distract from the gameworld.
For example, elevation can be used in this way. In Rise of the Argonauts, the player knows whether he is running towards the palace or the docks of Mycenae by whether the path is sloping upward or downward. The same method can be used in any number of settings, including dungeons and roads. The degree of a slope can be telling, too. If two paths both lead up a mountain, one can be recognizable from the other if it is significantly steeper.
Vegetation can be a gradual indicator. Grass and shrubs might grow tall in one area and shrink or die out toward another. Oaks might dominate one area while mesquite trees dominate another. Both trees might habit an area, but the ratio between them could indicate to the player in which direction he is headed.
Architecture and objects can be used in the same way. The facades of buildings and their relationships to each other (such as how far apart they are) can indicate if the area the player is proceeding toward is wealthy or poor, old or new, damaged or undamaged (by war, weather, or graffiti), and so on. Statues and decorations can change in frequency or in subject through an environment. The type of shops or their names can change in style and tone.
Even audio can lead players. The roar of a waterfall or pounding of waves can gradually rise in volume as the player nears them. Likewise, the sounds of animals, the firing of weapons, the chatter and laughter or shouts of NPCs, or a thousand other things.
Animations and even environmental A.I. can guide players. As a player nears a battlefield, he might increasing notice NPCs crouched over and weeping. Or NPCs might be increasingly erratic and skittish... increasingly distressed as the player nears the source of that distress. The NPCs might pace, look around fearfully, or chatter to themselves. Think of the A.I. in Batman: Arkham Asylum and how the Joker's men act differently as they become more aware of Batman's presence and are afraid.
There are countless forms gradual indicators can take. The goal is to help players navigate environments without having to take their eyes off the gameworld or see anything that takes their attention off the story they're experiencing.
Thursday, October 15, 2009
enemies with personality
One of the reasons I expect Borderlands to be a lot of fun is the dynamic comments that give individual enemies personalities:
The amusing outbursts of enemies makes them more than just faceless meatbags. The Halo series has proven how effective this can be.
Giving enemies dynamic outbursts does two things. First, it adds a personal dimension to combat, making it feel like your actually fighting individuals and not the Borg. Second, it adds variation and replay value. It can make one enemy feel slightly different from another despite both using the same model and the same AI.
This isn't only possible with human enemies. Individual dogs of the same breed can have different barks. Individual skaags or other beasts could have different growls, screeches, etc. Voice variation with animals and monsters has an effect similar to that with humans.
Nor does personalizing individual enemies have to be done only through modeling, AI, and voicing. There's also animations. Slightly different hand motions, ways of walking, relaxed and defensive stances, and so on can make enemies which are otherwise the same feel like true individuals.
Every aspect of NPCs changes the experience.
The amusing outbursts of enemies makes them more than just faceless meatbags. The Halo series has proven how effective this can be.
Giving enemies dynamic outbursts does two things. First, it adds a personal dimension to combat, making it feel like your actually fighting individuals and not the Borg. Second, it adds variation and replay value. It can make one enemy feel slightly different from another despite both using the same model and the same AI.
This isn't only possible with human enemies. Individual dogs of the same breed can have different barks. Individual skaags or other beasts could have different growls, screeches, etc. Voice variation with animals and monsters has an effect similar to that with humans.
Nor does personalizing individual enemies have to be done only through modeling, AI, and voicing. There's also animations. Slightly different hand motions, ways of walking, relaxed and defensive stances, and so on can make enemies which are otherwise the same feel like true individuals.
Every aspect of NPCs changes the experience.
Wednesday, September 23, 2009
creature depth
In S.T.A.L.K.E.R., dogs act differently depending on how many fellow dogs are around. When alone, they run away. In large packs, they will attack. And when the player kills most of the dogs in a pack, the survivors lose courage and run for safety.
This sort of depth, simple as it is, adds considerably to the richness of a gameworld. It could potentially add strategic depth as well. Notice that it's not just behavior but also goal priority which changes in that dog example.
Connecting AI behavior to stuff like population and environment is something I'd like to see more of.
This sort of depth, simple as it is, adds considerably to the richness of a gameworld. It could potentially add strategic depth as well. Notice that it's not just behavior but also goal priority which changes in that dog example.
Connecting AI behavior to stuff like population and environment is something I'd like to see more of.
Tuesday, May 26, 2009
civilians and facing mistakes
I'm happy to see civilians running around in Modern Warfare 2's latest trailer, as well as fighters in civilian clothes.
I can only guess the scenes with people running away are during battles and not between. If there are civilians present during battles, with actual risk of civilian casualties, I consider that a big leap forward for first-person war games. War is never just between soldiers.
I have played games before in which it was possible to shoot the wrong people and be punished for it. Anyone who has played the Call of Duty games has been forced to restart from a checkpoint after shooting an ally. I believe that's the wrong way to punish such mistakes.
A reset means the offending action never happened. It's like a deleted scene or an expunged record. For mistakes to truly matter, they must remain part of the player's journey, an irrevocable moment which demands reflection and penance. If that moment can be erased, even in a punishment like a reset, then the player is not made to face it and reflect.
Civilians are important to war simulations (as opposed to arcade shooters) because they epitomize the moral complexity and uncertainty of war. Some civilians are hostile but not threatening. Some are dangerous in their foolhardy support. Some who would run away and seem harmless are dangerous, not by will, but due to the information they possess that enemies could gain by interrogation.
Fun over simulation -- I understand. But such complications of real warfare can make settings and stories infinitely more compelling, and potentially add replayability. Civilians are not a vital element for war games, but represent enormous untapped potential.
I can only guess the scenes with people running away are during battles and not between. If there are civilians present during battles, with actual risk of civilian casualties, I consider that a big leap forward for first-person war games. War is never just between soldiers.
I have played games before in which it was possible to shoot the wrong people and be punished for it. Anyone who has played the Call of Duty games has been forced to restart from a checkpoint after shooting an ally. I believe that's the wrong way to punish such mistakes.
A reset means the offending action never happened. It's like a deleted scene or an expunged record. For mistakes to truly matter, they must remain part of the player's journey, an irrevocable moment which demands reflection and penance. If that moment can be erased, even in a punishment like a reset, then the player is not made to face it and reflect.
Civilians are important to war simulations (as opposed to arcade shooters) because they epitomize the moral complexity and uncertainty of war. Some civilians are hostile but not threatening. Some are dangerous in their foolhardy support. Some who would run away and seem harmless are dangerous, not by will, but due to the information they possess that enemies could gain by interrogation.
Fun over simulation -- I understand. But such complications of real warfare can make settings and stories infinitely more compelling, and potentially add replayability. Civilians are not a vital element for war games, but represent enormous untapped potential.
Thursday, May 21, 2009
kinetics
I bought Sacred 2 for my 360 when it was released, and traded it back in only a few days later. I didn't even want to make use of my money before getting rid of it. So I agree with GameTrailers' review, as usual. But I'm surprised kinetics was never mentioned.
I call it kinetics, but there might be a better term. What I'm referring to is the feel of action: the way animations and sound effects combine to bring actions to life.
Timing is key to kinetics. Imagine a baseball game. When a player swings his bat, is the animation slow or fast? Either could be fun, but the speed affects the feel. Now consider the sound. Is the crack of the bat against ball perfectly synched with the animation? Or does the sound come a split-second before or after the animation?
Now imagine there is no sound when bat meets ball. That would suck, wouldn't it? No game would forget to include that sound. But what about surrounding sounds? Does the crowd cheer when the batter gets a hit? Does the crowd's noise change at all depending on whether it's a home run, a grounder, or a bunt? What about the player's footsteps as he's running to 1st base? Does they get any sound?
Does every player throw or drop his bat the same way? Does every player swing the same way? Do they all run with the same gait?
The point is that even simple actions can have surprising potential for accompanying sounds and animations. And, generally, the more visual and aural details included, the more compelling an action feels.
So back to Sacred 2. Play that game, Play Diablo 2, and tell me which has better kinetics. Compared to combat in Diablo 2, combat in Sacred 2 feels like flag football. There are more sounds, including reactive lines of dialogue (not commentary... screams, yelps, and such), and animations are quick enough to keep combat tense and frantic. Titan Quest also fails by slowing animations and creating a slight disconnect between player attacks and enemy deaths and defense. That one's a good game, but the timing is slightly off and dying enemies often float in the air a moment.
All games are limited by time and expense. But pinpoint the actions your players will employ most often and devote as many sound effects and animations to those actions as possible. Other parts of your games can be sketched simulations, but core actions should be represented as fully as possible.
I call it kinetics, but there might be a better term. What I'm referring to is the feel of action: the way animations and sound effects combine to bring actions to life.
Timing is key to kinetics. Imagine a baseball game. When a player swings his bat, is the animation slow or fast? Either could be fun, but the speed affects the feel. Now consider the sound. Is the crack of the bat against ball perfectly synched with the animation? Or does the sound come a split-second before or after the animation?
Now imagine there is no sound when bat meets ball. That would suck, wouldn't it? No game would forget to include that sound. But what about surrounding sounds? Does the crowd cheer when the batter gets a hit? Does the crowd's noise change at all depending on whether it's a home run, a grounder, or a bunt? What about the player's footsteps as he's running to 1st base? Does they get any sound?
Does every player throw or drop his bat the same way? Does every player swing the same way? Do they all run with the same gait?
The point is that even simple actions can have surprising potential for accompanying sounds and animations. And, generally, the more visual and aural details included, the more compelling an action feels.
So back to Sacred 2. Play that game, Play Diablo 2, and tell me which has better kinetics. Compared to combat in Diablo 2, combat in Sacred 2 feels like flag football. There are more sounds, including reactive lines of dialogue (not commentary... screams, yelps, and such), and animations are quick enough to keep combat tense and frantic. Titan Quest also fails by slowing animations and creating a slight disconnect between player attacks and enemy deaths and defense. That one's a good game, but the timing is slightly off and dying enemies often float in the air a moment.
All games are limited by time and expense. But pinpoint the actions your players will employ most often and devote as many sound effects and animations to those actions as possible. Other parts of your games can be sketched simulations, but core actions should be represented as fully as possible.
Monday, March 30, 2009
AI companions suck
While I love AI as competitors, I don't like AI as allies. Pathfinding, accuracy and such are common issues... but that's not why I hate AI companions.
I generally prefer not to have an AI companion, regardless of implementation, because the AI acts independently of my playstyle and impulsive decisions... thereby interfering.
For example, I might want to be stealthy, snapping necks or sniping from afar. The oblivious AI draws the enemies attention by not hiding, even going in guns blazing.
To a degree, this is avoidable. But the amount of planning and coding (and trial-and-error) required for an AI companion to regularly recognize and adapt to the many possible player behaviors is daunting. And in most action games, all it takes is a moment's hesitation for a companion to disrupt the player's gameplay.
For example, say the player is moving forward when an enemy suddenly comes from around a structure and into view. The player immediately falls prone to avoid detection, but the AI must wait for the player to act to know if he will go prone or charge and shoot. That moment's hesitation gets them spotted and now the player is forced to engage. I'm sure you can imagine any number of similar scenarios where a second's hesitation is a second too long.
Of course, there is some hope. Neverwinter Nights let players choose basic AI behavior, in the vein of "Stay [x] far from me" or "Use your [ranged/melee] weapon". I'm surprised that sort of feature hasn't been included in many other games... particularly FPS games.
It's common in squad-based games for the player to be able to give direct orders like "Go there" or "Focus on that target". Being able to adapt an NPC companion's general style and/or gear would be nice. Yes, being able to select from different companions allows this to a degree. But that's painting in broad strokes. We can do better.
I generally prefer not to have an AI companion, regardless of implementation, because the AI acts independently of my playstyle and impulsive decisions... thereby interfering.
For example, I might want to be stealthy, snapping necks or sniping from afar. The oblivious AI draws the enemies attention by not hiding, even going in guns blazing.
To a degree, this is avoidable. But the amount of planning and coding (and trial-and-error) required for an AI companion to regularly recognize and adapt to the many possible player behaviors is daunting. And in most action games, all it takes is a moment's hesitation for a companion to disrupt the player's gameplay.
For example, say the player is moving forward when an enemy suddenly comes from around a structure and into view. The player immediately falls prone to avoid detection, but the AI must wait for the player to act to know if he will go prone or charge and shoot. That moment's hesitation gets them spotted and now the player is forced to engage. I'm sure you can imagine any number of similar scenarios where a second's hesitation is a second too long.
Of course, there is some hope. Neverwinter Nights let players choose basic AI behavior, in the vein of "Stay [x] far from me" or "Use your [ranged/melee] weapon". I'm surprised that sort of feature hasn't been included in many other games... particularly FPS games.
It's common in squad-based games for the player to be able to give direct orders like "Go there" or "Focus on that target". Being able to adapt an NPC companion's general style and/or gear would be nice. Yes, being able to select from different companions allows this to a degree. But that's painting in broad strokes. We can do better.
Tuesday, March 24, 2009
AI support strategic gameplay
I've decided to let yesterday's question sit for a day, so there's more time for feedback. In the meantime, I'm going to revisit an old subject: how AI competitors can augment human competition.
AI can be more conducive to player strategy than player opponents. Strategy is only possible to the extent that things can be predicted. If you can't foresee something, then you can't plan for it. Of course, sometimes predictability includes the lack thereof (you expect that something unexpected will occur, though you don't know exactly what).
Playing against a particular "personality" many times allows the player to gradually adapt, to develop a strategy over the course of multiple play-sessions. In multiplayer, human personalities are often unstable... meaning that players have fluid styles (due to mood, learning, or battle scenarios), they make impulsive decisions, etc. There's also frequent personality turnover, meaning that you're constantly playing new competitors and there are long intermissions between battles against any one competitor.
AI in the form of "multiplayer" bots serve two main functions. First, they prepare players for facing particular strategies and scenarios. There's wide discrepancy between how much time different players need for this. Second, they can offer a fun alternative to unpredictable nature of human beings. Even players who enjoy the challenge and dynamics of human competition sometimes want a change of pace, a different experience. AI can act as a nice balance and prevent players from burning out.
There are other potential uses, but I just wanted to emphasize the value of AI for supporting strategic gameplay. Don't think of bots as an alternative to players. Interaction with bots can aid interaction between players.
And don't think that AI in a single-player campaign acts in the same way.
AI can be more conducive to player strategy than player opponents. Strategy is only possible to the extent that things can be predicted. If you can't foresee something, then you can't plan for it. Of course, sometimes predictability includes the lack thereof (you expect that something unexpected will occur, though you don't know exactly what).
Playing against a particular "personality" many times allows the player to gradually adapt, to develop a strategy over the course of multiple play-sessions. In multiplayer, human personalities are often unstable... meaning that players have fluid styles (due to mood, learning, or battle scenarios), they make impulsive decisions, etc. There's also frequent personality turnover, meaning that you're constantly playing new competitors and there are long intermissions between battles against any one competitor.
AI in the form of "multiplayer" bots serve two main functions. First, they prepare players for facing particular strategies and scenarios. There's wide discrepancy between how much time different players need for this. Second, they can offer a fun alternative to unpredictable nature of human beings. Even players who enjoy the challenge and dynamics of human competition sometimes want a change of pace, a different experience. AI can act as a nice balance and prevent players from burning out.
There are other potential uses, but I just wanted to emphasize the value of AI for supporting strategic gameplay. Don't think of bots as an alternative to players. Interaction with bots can aid interaction between players.
And don't think that AI in a single-player campaign acts in the same way.
Friday, January 30, 2009
don't ignore player success
In the Call of Duty: World at War single-player/co-op campaign, enemies spawn in an infinite stream until the player pushes forward to seize the objective. That means that how many you kill is irrelevant. They keep coming until your player stands in the targeted location.
Medal of Honor: Airborne does it differently. If you stand in one place and keep shooting enemies, they'll continue to spawn for a while, prolonging the experience. But the game eventually rewards you for your kills by ceasing to spawn new enemies for that location.
These are different kinds of games -- one involves pushing forward on a linear path, while the other lets the player parachute onto any location of the battlefield and approach objectives from many angles. But regardless of that difference, I'd say Airborne's spawn method is better. Even if reaching locations or destroying AA guns (or whatever) is the main goal, kills should be rewarded.
Every activity central to gameplay should be rewarded. No accomplishment should be completed overshadowed by another.
Medal of Honor: Airborne does it differently. If you stand in one place and keep shooting enemies, they'll continue to spawn for a while, prolonging the experience. But the game eventually rewards you for your kills by ceasing to spawn new enemies for that location.
These are different kinds of games -- one involves pushing forward on a linear path, while the other lets the player parachute onto any location of the battlefield and approach objectives from many angles. But regardless of that difference, I'd say Airborne's spawn method is better. Even if reaching locations or destroying AA guns (or whatever) is the main goal, kills should be rewarded.
Every activity central to gameplay should be rewarded. No accomplishment should be completed overshadowed by another.
Thursday, November 27, 2008
multiplayer stand-ins
Another aspect of Left 4 Dead which should be considered by all designers is the option for an AI stand-in during multiplayer.
Most gamers acknowledge that gaming takes a backseat to other things, including everyday distractions like phone calls. In most multiplayer games, gameplay cannot be paused; and when gameplay can be paused, at least one person is left sitting idly in boredom. Left 4 Dead allows a player to call an AI substitute, so others can go on playing.
It's not a perfect feature. Any AI is imperfect, and a substitution can mean someone is temporarily playing by himself. But I'd still recommend a similar feature for almost any multiplayer game (excluding MMOs).
I'm enjoying Thanksgiving out of town this week, but hopefully I'll be able to keep posting stuff here 'til I get back. My thanks to God for all our many blessings.
Most gamers acknowledge that gaming takes a backseat to other things, including everyday distractions like phone calls. In most multiplayer games, gameplay cannot be paused; and when gameplay can be paused, at least one person is left sitting idly in boredom. Left 4 Dead allows a player to call an AI substitute, so others can go on playing.
It's not a perfect feature. Any AI is imperfect, and a substitution can mean someone is temporarily playing by himself. But I'd still recommend a similar feature for almost any multiplayer game (excluding MMOs).
I'm enjoying Thanksgiving out of town this week, but hopefully I'll be able to keep posting stuff here 'til I get back. My thanks to God for all our many blessings.
Tuesday, November 18, 2008
AI pushbacks
Something I've wanted to see in games for a long time is dynamic enemy AI with enough sophistication to shape a gameworld without overly frustrating players.
If AI characters each have a variety of goals and a variety of behaviors for achieving those goals, then gameplay can continually surprise and each player can have unique experiences to share. But even when games include such variety, they rarely enable AI to regain what the player has fought for. Player progress is rarely tentative.
In Saints Row and Saints Row 2, players seek to conquer the city neighborhood-by-neighborhood. After taking a neighborhood, there's usually a pushback. Whatever gang used to own that neighborhood tries to take it back. But no dynamics are involved in the event. Enemy AI don't move into the neighborhood while the player is away, but simply appear there instantaneously. Because it's a completely scripted event, there are no surprises in the fight to take back that territory. A fixed number of gang leaders of fixed strength and effect appear in fixed locations, which the player must kill to secure a zone with fixed borders. Judging from experience, each territory can only have one pushback.
LOTR: Battle for Middle Earth 2 involves more genuine pushbacks, in its War of the Ring mode. The setting is war, and wars can be lost. This is unlike a typical adventure game, in which it's assumed the player will win the war and only battles can be lost. In War of the Ring (similar to the boardgame Risk), strategic AI decisions are fairly predictable, but territory geography and limited resources combine to make simultaneous offense and defense difficult. To be strong enough to seize new territories, the player must sometimes leave old territories vulnerable to attack. Also, each territory awards particular benefits for ownership. These dynamics supplement relatively static AI to result in dynamic pushbacks. The player's ability to predict the place, time, and manner of pushbacks is limited. A pushback in a particular territory can be repeated multiple times, and the enemy's invasions can be devastating.
Now, The Godfather 2 gets its turn: "The Godfather 2 is dynamic and reactive, families react with complex "countermoves" based on the players actions and leveraging very human behaviors like revenge and coercion." Though the game includes a strategic map, The Don's View (shown above), "revenge and coercion" sound like ground-level interactions. I'm anxious to see how this turns out.
If AI characters each have a variety of goals and a variety of behaviors for achieving those goals, then gameplay can continually surprise and each player can have unique experiences to share. But even when games include such variety, they rarely enable AI to regain what the player has fought for. Player progress is rarely tentative.
In Saints Row and Saints Row 2, players seek to conquer the city neighborhood-by-neighborhood. After taking a neighborhood, there's usually a pushback. Whatever gang used to own that neighborhood tries to take it back. But no dynamics are involved in the event. Enemy AI don't move into the neighborhood while the player is away, but simply appear there instantaneously. Because it's a completely scripted event, there are no surprises in the fight to take back that territory. A fixed number of gang leaders of fixed strength and effect appear in fixed locations, which the player must kill to secure a zone with fixed borders. Judging from experience, each territory can only have one pushback.
LOTR: Battle for Middle Earth 2 involves more genuine pushbacks, in its War of the Ring mode. The setting is war, and wars can be lost. This is unlike a typical adventure game, in which it's assumed the player will win the war and only battles can be lost. In War of the Ring (similar to the boardgame Risk), strategic AI decisions are fairly predictable, but territory geography and limited resources combine to make simultaneous offense and defense difficult. To be strong enough to seize new territories, the player must sometimes leave old territories vulnerable to attack. Also, each territory awards particular benefits for ownership. These dynamics supplement relatively static AI to result in dynamic pushbacks. The player's ability to predict the place, time, and manner of pushbacks is limited. A pushback in a particular territory can be repeated multiple times, and the enemy's invasions can be devastating.
Now, The Godfather 2 gets its turn: "The Godfather 2 is dynamic and reactive, families react with complex "countermoves" based on the players actions and leveraging very human behaviors like revenge and coercion." Though the game includes a strategic map, The Don's View (shown above), "revenge and coercion" sound like ground-level interactions. I'm anxious to see how this turns out.
Wednesday, November 12, 2008
military's AI in MMOs
The U.S. military is about to use MMOs as a testing ground for their latest human-simulation AI.
According to the Army's Director of Research and Laboratory Management, these AI characters "think independently, have emotions and speak in slang". In other words, these are far beyond the bots you might be accustomed to seeing in MMOs. These bots don't just grind. They will seem to converse with players.
There's no mention of the extent to which these Army bots will partake in gameplay. The focus is clearly on simulating social intelligence and personality, rather than combat prowess. I'm sure our military has impressive AI for that as well.
Unfortunately, I doubt the Army's test results and strategies will be shared with the game industry, though I could be wrong. If our military perceives such AI has having only training value, then they might share. But if they believe such AI has intelligence and counter-intelligence applications, as I expect, then they will guard their methods against enemy insights.
So don't expect this to lead to gaming progress.
But do feel free to make a mini-game out of identifying impostors! You'll be helping our troops in the process.
According to the Army's Director of Research and Laboratory Management, these AI characters "think independently, have emotions and speak in slang". In other words, these are far beyond the bots you might be accustomed to seeing in MMOs. These bots don't just grind. They will seem to converse with players.
There's no mention of the extent to which these Army bots will partake in gameplay. The focus is clearly on simulating social intelligence and personality, rather than combat prowess. I'm sure our military has impressive AI for that as well.
Unfortunately, I doubt the Army's test results and strategies will be shared with the game industry, though I could be wrong. If our military perceives such AI has having only training value, then they might share. But if they believe such AI has intelligence and counter-intelligence applications, as I expect, then they will guard their methods against enemy insights.
So don't expect this to lead to gaming progress.
But do feel free to make a mini-game out of identifying impostors! You'll be helping our troops in the process.
Friday, August 08, 2008
a lesson from Beijing
After watching the opening ceremony at the Beijing Olympics, I'm reminded of how a different beauty emerges from large masses of people as from groups and individuals. Over two thousand drummers synchronize their actions to create an awesome spectacle that's simply not possible in any other way.
Games are increasingly able to include vast numbers of objects and characters -- coordinately directly or indirectly, or acting separately. Keep in mind that sheer numbers can change aesthetic and narrative possibilities. Through masses, you can create experiences for players that are essentially different from encounters with individuals and groups.
Games are increasingly able to include vast numbers of objects and characters -- coordinately directly or indirectly, or acting separately. Keep in mind that sheer numbers can change aesthetic and narrative possibilities. Through masses, you can create experiences for players that are essentially different from encounters with individuals and groups.
Thursday, July 24, 2008
companion milestone
In a hands-on preview of Fable 2 at Kotaku, Michael McWhertor notes that the game's combat is fun "especially when the dog comes to your aid, gnawing on the limb of some recently dispatched foe". That's a sort of enjoyment that is common in games, but the potential is largely unexplored.
I'll venture to say that what Molyneux's team has done with canine companions is a monumental achievement; a stepping stone toward an important trend in future games. He's capitalizing on the joy of love.
As the philosopher Peter Kreeft once said, love is about "with-ness". It's a desire and/or choice to be together with something; in presence, in experience, in ideas, etc. Love is about communion.
When you love any being, you take joy in simple exposure to it. A newborn baby can't even focus its eyes, yet family and friends are captivated by anything and everything he or she does. Likewise with pets and wild animals. There's joy in merely watching and listening to them.
The deeper, more complex NPCs become, the more potential they have to entertain the player through exhibitions of personality, quirks, and dynamic experiences. In McWhertor's Fable 2 experience, the dog in combat is doing nothing very different from companions in Diablo 2, yet the player's interest in the dog's actions is significantly deeper. A well-rendered, clever (AI), dynamic pet evokes a stronger emotional response than a 2D, simple-minded skeleton without a name.
One of Bungie's brilliant moves with Halo was giving NPC companions dynamic comments. The player kills an enemy and a soldier behind him says, "Aww, man. That one was mine!" (or something like that).
NPC mistakes can be made endearing, rather than frustrating. Trivial dialogue can be made more memorable than story arcs. It can be done through deepening characters... adding individuality and dynamicity. Fable 2's dog is an important step that barely hints at things to come.
I'll venture to say that what Molyneux's team has done with canine companions is a monumental achievement; a stepping stone toward an important trend in future games. He's capitalizing on the joy of love.
As the philosopher Peter Kreeft once said, love is about "with-ness". It's a desire and/or choice to be together with something; in presence, in experience, in ideas, etc. Love is about communion.
When you love any being, you take joy in simple exposure to it. A newborn baby can't even focus its eyes, yet family and friends are captivated by anything and everything he or she does. Likewise with pets and wild animals. There's joy in merely watching and listening to them.
The deeper, more complex NPCs become, the more potential they have to entertain the player through exhibitions of personality, quirks, and dynamic experiences. In McWhertor's Fable 2 experience, the dog in combat is doing nothing very different from companions in Diablo 2, yet the player's interest in the dog's actions is significantly deeper. A well-rendered, clever (AI), dynamic pet evokes a stronger emotional response than a 2D, simple-minded skeleton without a name.
One of Bungie's brilliant moves with Halo was giving NPC companions dynamic comments. The player kills an enemy and a soldier behind him says, "Aww, man. That one was mine!" (or something like that).
NPC mistakes can be made endearing, rather than frustrating. Trivial dialogue can be made more memorable than story arcs. It can be done through deepening characters... adding individuality and dynamicity. Fable 2's dog is an important step that barely hints at things to come.
Friday, June 27, 2008
contextual powers
One of the great movements in design right now is contextual powers.
Assassin's Creed is a great example. A single player-command during combat (block + timed prevention strike = counterattack) can result in one of at least half-a-dozen moves. Which counterattack is seen depends on the relative position of defender-to-attacker and perhaps timing as well. Climbing is also contextual in Assassin's Creed. Simply direct Altair's movement and he'll use whatever actions are necessary to climb or leap that way. It's simple, fun, and occasionally surprising.
Fable 2 apparently makes spell-casting and ranged combat contextual as well. I'm anxious to see how that will work, but it also has me looking ahead to games of next year and beyond. How else might contextual powers be applied?
STORY
Game events are often activated by invisible triggers placed in the environment. When the player crosses a particular threshold or interacts with a particular object, a scripted event is set in motion. Triggers could also be used to activate variations in AI. When an AI character witnesses a particular event, crosses a particular threshold or such, then a change in personality occurs.
In Neverwinter Nights, child NPCs are invincible. If you attack one, the child will follow and attack your character relentlessly, eventually killing him. That's the closest thing I've seen to what I'm talking about, but that's still not it. That's basically a scripted event, a simple action command ("Kill the PC"). If it had been a change in AI, the children's perception of your character would change, but a number of subsequent events would be possible due to different personalities, different environmental circumstances, and other variables which combine on-the-fly.
ITEMS
I've long waited for contextual items. Item reward systems can be a lot of fun. I think they can be even more fun if some items are not what they appear to be, and if some items receive bonus powers in particular situations.
A broad item reward system has never been combined with a contextual combat system. It would be ridiculous to attempt as many attack animations as Assassin's Creed had for every weapon in a Diablo-type game. But breaking weapons down into a small number of classes (sword, axe, blunt, etc) would make this combo possible. In MMOs, it can be fun seeing how different characters' combat actions mix, but how much more so if there was significant variety within each character's optimal repertoire?
We've seen weapons and armor with effects or resistances that make them optimal for engagement with particular enemies. That's not the sort of contextual variety I'm praising here. In those cases, each item offers a single animation and three degrees of difficulty dependent on the enemy (ineffectual, average effect, superior effect). What I hope to see is different animations and other visible/audible effect variations independent of enemy type. Beyond mere show, I'm hoping for variety in type within one weapon.
When a player uses a "Frostbane" flaming sword against arctic inhabitants, he expects the sword to be more effectual and, though he might take some mild pleasure from pride in being smartly equipped, the experience is basically the same as using the weapon in any other scenario. On the other hand, imagine that the sword exhibits an effect type when fighting those arctic creatures that it doesn't exhibit elsewhere. What if the sword looked rather plain most of the time, but it glows hot as the player journeys further into the cold setting? What if fire suddenly erupted from the blade and was hurled at the enemy the blade was destined to fight?
Experiences like that don't just make players grin and think "cool...". Experiences like that make us wide-eyed, exhilirated, and we tell our friends about it days or even weeks later. It's more impressive, and it's feasible.
SKILLS
One of the cooler concepts I've ever read was something the Darkfall devs suggested about their Alfar race early in production (I don't know if the game still includes this scenario). The Alfar are ruled their god in physical form as a king, and that god is literally insane. All of their magic comes from him. So the spellcasters of that race have unpredictable magic. As the god's mood and sanity shift from one moment to the next, so does the potency of their spells. Sometimes their magic will sputter and fail, while at other times it is excessively powerful and brilliant.
While that's an extreme example, I think there's a lot of hope for similar variety within any particular skill. Variety can keep games fresh and replayable. It would be great if a skill gained early in a game could still surprise its user near the end.
How else might context be used to make games more fun?
Assassin's Creed is a great example. A single player-command during combat (block + timed prevention strike = counterattack) can result in one of at least half-a-dozen moves. Which counterattack is seen depends on the relative position of defender-to-attacker and perhaps timing as well. Climbing is also contextual in Assassin's Creed. Simply direct Altair's movement and he'll use whatever actions are necessary to climb or leap that way. It's simple, fun, and occasionally surprising.
Fable 2 apparently makes spell-casting and ranged combat contextual as well. I'm anxious to see how that will work, but it also has me looking ahead to games of next year and beyond. How else might contextual powers be applied?
STORY
Game events are often activated by invisible triggers placed in the environment. When the player crosses a particular threshold or interacts with a particular object, a scripted event is set in motion. Triggers could also be used to activate variations in AI. When an AI character witnesses a particular event, crosses a particular threshold or such, then a change in personality occurs.
In Neverwinter Nights, child NPCs are invincible. If you attack one, the child will follow and attack your character relentlessly, eventually killing him. That's the closest thing I've seen to what I'm talking about, but that's still not it. That's basically a scripted event, a simple action command ("Kill the PC"). If it had been a change in AI, the children's perception of your character would change, but a number of subsequent events would be possible due to different personalities, different environmental circumstances, and other variables which combine on-the-fly.
ITEMS
I've long waited for contextual items. Item reward systems can be a lot of fun. I think they can be even more fun if some items are not what they appear to be, and if some items receive bonus powers in particular situations.
A broad item reward system has never been combined with a contextual combat system. It would be ridiculous to attempt as many attack animations as Assassin's Creed had for every weapon in a Diablo-type game. But breaking weapons down into a small number of classes (sword, axe, blunt, etc) would make this combo possible. In MMOs, it can be fun seeing how different characters' combat actions mix, but how much more so if there was significant variety within each character's optimal repertoire?
We've seen weapons and armor with effects or resistances that make them optimal for engagement with particular enemies. That's not the sort of contextual variety I'm praising here. In those cases, each item offers a single animation and three degrees of difficulty dependent on the enemy (ineffectual, average effect, superior effect). What I hope to see is different animations and other visible/audible effect variations independent of enemy type. Beyond mere show, I'm hoping for variety in type within one weapon.
When a player uses a "Frostbane" flaming sword against arctic inhabitants, he expects the sword to be more effectual and, though he might take some mild pleasure from pride in being smartly equipped, the experience is basically the same as using the weapon in any other scenario. On the other hand, imagine that the sword exhibits an effect type when fighting those arctic creatures that it doesn't exhibit elsewhere. What if the sword looked rather plain most of the time, but it glows hot as the player journeys further into the cold setting? What if fire suddenly erupted from the blade and was hurled at the enemy the blade was destined to fight?
Experiences like that don't just make players grin and think "cool...". Experiences like that make us wide-eyed, exhilirated, and we tell our friends about it days or even weeks later. It's more impressive, and it's feasible.
SKILLS
One of the cooler concepts I've ever read was something the Darkfall devs suggested about their Alfar race early in production (I don't know if the game still includes this scenario). The Alfar are ruled their god in physical form as a king, and that god is literally insane. All of their magic comes from him. So the spellcasters of that race have unpredictable magic. As the god's mood and sanity shift from one moment to the next, so does the potency of their spells. Sometimes their magic will sputter and fail, while at other times it is excessively powerful and brilliant.
While that's an extreme example, I think there's a lot of hope for similar variety within any particular skill. Variety can keep games fresh and replayable. It would be great if a skill gained early in a game could still surprise its user near the end.
How else might context be used to make games more fun?
Thursday, June 26, 2008
coopting level design for AI
You might enjoy watching the NOVA special on the DARPA Grand Challenge, a desert race for robots. I was particularly impressed by Stanley, a robot by Stanford's AI Lab led by Sebastian Thrun.
Stanley combines laser input with a video camera's data. The laser scans terrain to detect flatness. Next, the AI compares that data to the video image. It guesses that the video-captured landscape beyond the laser's range will roughly match the landscape within the laser's range. The AI focuses on colors in the video image. So, Stanley assumes, if flattest terrain coincides with a general color (the color of the road), then wherever that color is seen in the video image is more likely to be flat terrain (useable road) as well. The laser data is the core data, but the camera data enables the AI to predict future input.
Needless to say, going to the link above and watching the show will clarify my description.
Making art part of AI
It occurred to me that similar systems could be used to improve game AI... only game designers have the advantage of being able to design environments to fit. Rather than design pathfinding programs to work in any game setting, why not work cooperatively with artists and level designers so that environments can directly inform the AI?
Stanley had to actively discern road by comparing two types of data, and with limited certainty. In a game, certain colors and textures can be exclusive to particular objects, thereby acting as perfectly accurate qualifiers for AI.
Script triggers and invisible rails for object movement are commonly seen in games, of course. But what I'm proposing (and perhaps this method is already used -- I'm not a programmer) is data included in environment/level planning that is informative, rather than determinative.
Such cues might take the form of marker objects, hues, line or object patterns, textures, or combinations. And they could be used for anything from pathfinding to personality simulations (ex: different game AI could prefer different terrain or abundance of cover).
Stanley combines laser input with a video camera's data. The laser scans terrain to detect flatness. Next, the AI compares that data to the video image. It guesses that the video-captured landscape beyond the laser's range will roughly match the landscape within the laser's range. The AI focuses on colors in the video image. So, Stanley assumes, if flattest terrain coincides with a general color (the color of the road), then wherever that color is seen in the video image is more likely to be flat terrain (useable road) as well. The laser data is the core data, but the camera data enables the AI to predict future input.
Needless to say, going to the link above and watching the show will clarify my description.
Making art part of AI
It occurred to me that similar systems could be used to improve game AI... only game designers have the advantage of being able to design environments to fit. Rather than design pathfinding programs to work in any game setting, why not work cooperatively with artists and level designers so that environments can directly inform the AI?
Stanley had to actively discern road by comparing two types of data, and with limited certainty. In a game, certain colors and textures can be exclusive to particular objects, thereby acting as perfectly accurate qualifiers for AI.
Script triggers and invisible rails for object movement are commonly seen in games, of course. But what I'm proposing (and perhaps this method is already used -- I'm not a programmer) is data included in environment/level planning that is informative, rather than determinative.
Such cues might take the form of marker objects, hues, line or object patterns, textures, or combinations. And they could be used for anything from pathfinding to personality simulations (ex: different game AI could prefer different terrain or abundance of cover).
Wednesday, May 21, 2008
AI: beyond logic
Human beings are not logical. Or, rather, we're not only logical. Every one of us makes choices that depart from what we consciously perceive to be the best, most rational options. A person who exhibits no emotion is called robotic, because emotion is also an essential aspect of humanity. Some might even go further and say that the will is neither logic nor emotion, but something else.
My own view is that emotions are based in a sort of logic. Happiness is a response to communion with the things we love; togetherness. Sadness is a response to being apart from those objects. Anger is a response to injustice. And other emotions are generally some mixture of basic emotions like those, filtered through our own personal perceptions of what is truth, what is love, etc.
Regardless of what one theorizes emotions are, we are incapable of completely predicting human behavior. We can't even fully predict the behavior of many lesser animals.
Viewers of horror films often remark that particular characters are unbelievable because those characters act irrationally. I don't think the audience truly expects an absence of impediments to logic, such as fear and anger (or even varying degrees of intelligence). Instead, the audience is responding to the characters being too neatly fashioned... too predictable. Of course, she's just going to hide and cry until she's found and slaughtered! That's what they always do, right? Audiences generally expect characters to have half a brain, but the actions of depthful characters always contain surprises. A truly depthful character is not completely knowable, because no human being is completely knowable.
So where is all of this leading? When depth is a goal, AI should include self-contained variables. A character's actions should not be determined solely by a "personality" type plus environmental circumstances. Within that personality, there should be a variable range.
The same person placed in the same basic circumstance will often respond differently. In reality, this doesn't happen willy-nilly; it's in response to a change in internal circumstances (distraction, drowsiness, recurring thoughts due to individual values, new ideas conflicting with conditioning, etc). But trying to mimic the endless intricacies of a human personality is a lost cause. So a simulation meant to give the illusion of true depth should contain internal dice... doing for personality what the Euphoria engine does for physics, albeit accomplishing that illusion in a very different way.
Minor variables can have big consequences, as I'm sure AI programmers are very aware. If an enemy hesitates in combat for just a second, all sorts of things could happen in that second. The player might gain an advantage. The enemy's companions could arrive and start shooting. The wind or waves could change, pushing a vehicle slightly off its intended course. That moment's hesitation doesn't have to be exactly described by the programmers or animators, because the player's imagination will fill in the blanks. Perhaps it was fear. Perhaps it was a sudden realization. Perhaps it was his back going out of alignment (speaking of simulation... the problems of getting old!).
One thing game developers could learn from old movies (40s and 50s) is the benefits of innuendo and incomplete information. Scenes are most powerful when not everything is spelled out for the audience. That includes combat scenes.
Anyway, I think the future of storytelling in games lies largely in the development of better AIs. This is one way.
My own view is that emotions are based in a sort of logic. Happiness is a response to communion with the things we love; togetherness. Sadness is a response to being apart from those objects. Anger is a response to injustice. And other emotions are generally some mixture of basic emotions like those, filtered through our own personal perceptions of what is truth, what is love, etc.
Regardless of what one theorizes emotions are, we are incapable of completely predicting human behavior. We can't even fully predict the behavior of many lesser animals.
Viewers of horror films often remark that particular characters are unbelievable because those characters act irrationally. I don't think the audience truly expects an absence of impediments to logic, such as fear and anger (or even varying degrees of intelligence). Instead, the audience is responding to the characters being too neatly fashioned... too predictable. Of course, she's just going to hide and cry until she's found and slaughtered! That's what they always do, right? Audiences generally expect characters to have half a brain, but the actions of depthful characters always contain surprises. A truly depthful character is not completely knowable, because no human being is completely knowable.
So where is all of this leading? When depth is a goal, AI should include self-contained variables. A character's actions should not be determined solely by a "personality" type plus environmental circumstances. Within that personality, there should be a variable range.
The same person placed in the same basic circumstance will often respond differently. In reality, this doesn't happen willy-nilly; it's in response to a change in internal circumstances (distraction, drowsiness, recurring thoughts due to individual values, new ideas conflicting with conditioning, etc). But trying to mimic the endless intricacies of a human personality is a lost cause. So a simulation meant to give the illusion of true depth should contain internal dice... doing for personality what the Euphoria engine does for physics, albeit accomplishing that illusion in a very different way.
Minor variables can have big consequences, as I'm sure AI programmers are very aware. If an enemy hesitates in combat for just a second, all sorts of things could happen in that second. The player might gain an advantage. The enemy's companions could arrive and start shooting. The wind or waves could change, pushing a vehicle slightly off its intended course. That moment's hesitation doesn't have to be exactly described by the programmers or animators, because the player's imagination will fill in the blanks. Perhaps it was fear. Perhaps it was a sudden realization. Perhaps it was his back going out of alignment (speaking of simulation... the problems of getting old!).
One thing game developers could learn from old movies (40s and 50s) is the benefits of innuendo and incomplete information. Scenes are most powerful when not everything is spelled out for the audience. That includes combat scenes.
Anyway, I think the future of storytelling in games lies largely in the development of better AIs. This is one way.
Wednesday, April 16, 2008
we need bots
There's a good post about opponent bots over at Gaming Verdict.
This is one of those industry changes that has slipped under the radar, I think. My friends and I used to spend a ridiculous amount of time playing with bots in Perfect Dark on the N64. Super Smash Bros. is another great example. I'm always disappointed when games that could include bots don't.
COMPETITION
Gaming Verdict mentions the steep learning curve of gaming against human opponents. Just the other day, I was introducing a friend to Call of Duty 4 multiplayer; the game didn't take. He told me that it definitely seemed awesome, but someone like him, for whom games are only occasional recreation and whose reflexes are no longer lightning-quick, simply can't compete. And competing split-screen, just me versus him, is nothing like the gleeful chaos of our Perfect Dark experience.
All games have that problem with competition... not just FPS fragfests that require split-second reactions. I'm a fervent and lifelong gamer, but even I don't try to compete online with my favorite RTS game, Battle for Middle Earth 2. The several family members and friends with whom I've played that game prefer to team up against bots in skirmishes or campaigns.
Michael de Plater, the creative director of Tom Clancy's Endwar, has cited that upcoming game's use of voice commands as a way to counteract the advantage of knowing macros and complex controls, but difficulty of control isn't half the problem (in fairness, de Plater doesn't suggest that it is). Most competitive online gaming networks feel as if professional, college, and high school sports conferences were combined into one. More than a few factors separate the amateur from the professional.
CoD4 admirably matches pros with amateurs in team matches, but I think the amateurs more often feel more useless and outmatched than like they're contributing to the team. That's how my friend felt. That's how I feel sometimes... especially when another player clearly knows the map, and I don't. The greatest sense of accomplishment comes from winning; nearly losing, but winning. Losing occasionally or even half the time is no problem. Losing almost every time is not fun for most people (unless it's to a pretty lady, of course). Relatively few have the personality type that views nine losses as an acceptable price for one win, when he or she is only competing for entertainment. One can view his or her own mild contributions as meaningful contributions, but that is, perhaps, not an automatic or particularly common view.
The skill levels and playstyles of players are hard to quantify, so matchmaking systems only help so much. In fact, matchmaking is downright useless when players alternate between play-options. A player might be good with some weapons/units, but not others; with some factions, but not with others; on some maps or scenarios/modes, but not others. And suppose I've improved my skills in single-player mode since the last time I played multiplayer. Should I now have to begin facing easy human opponents, because I haven't yet proven myself to matchmaking system?
Furthermore, a person's playstyle and desired degree of investment in the game can change from session to session. A gamer can feel particularly competitive one day and be interested in only leisurely engagement the next (perhaps due to being tired after a long day's work).
Remember that the difference between a professional and amateur isn't just skill and experience, but interest. A lot of people play casual club sports -- not because they're not capable of more intense performance -- because they prefer a different pace, different style and rules, or do not consider the game a high priority.
Finally, though I'm generally a big fan of reducing predictability in games, bots can actually offer a deeper sense of strategy than human opponents precisely because they're more predictable. Half of battle strategy is anticipating your opponent's next move. When you're playing against hundreds or thousands, if not millions, of other human beings, then you rarely face the same opponent more than once or twice. As a result, you can only learn and adjust to the broad strategies that are common among players, and not the specific strategies of individuals.
Online multiplayer games are more tactical than strategic... the player must react to momentary events, rather than plan for organized movements. Bots (particularly, campaigns against bots) enable true strategy.
PAUSE
Another thing I had to explain to my friend about CoD4 multiplayer was why one of his matches was ended abruptly when "the host... left the game". Sure, there are plenty of inconsiderate players who cut a game off just because they didn't like the map or the way the match was going, but there are also players answering telephones or doors, being distracted by kids or pets, or having to unexpectedly end their play-sessions.
As Gaming Verdict points out, the ability to pause is easier with bot-driven games. Games could even be designed to automatically replace a player's character with a bot if that player has to leave the game temporarily. Life is unpredictable. A purely logical software program is more likely to understand that than a stranger.
I realize that developing fun and competitive AI is no quick or easy task. But remember that even simple AI can be fun when combined with the right dynamics (ex: Turok shows that even a dinosaur that simply runs straight for you is tough when there's also a soldier nearby shooting at you).
We miss our bots. Give them back.
This is one of those industry changes that has slipped under the radar, I think. My friends and I used to spend a ridiculous amount of time playing with bots in Perfect Dark on the N64. Super Smash Bros. is another great example. I'm always disappointed when games that could include bots don't.
COMPETITION
Gaming Verdict mentions the steep learning curve of gaming against human opponents. Just the other day, I was introducing a friend to Call of Duty 4 multiplayer; the game didn't take. He told me that it definitely seemed awesome, but someone like him, for whom games are only occasional recreation and whose reflexes are no longer lightning-quick, simply can't compete. And competing split-screen, just me versus him, is nothing like the gleeful chaos of our Perfect Dark experience.
All games have that problem with competition... not just FPS fragfests that require split-second reactions. I'm a fervent and lifelong gamer, but even I don't try to compete online with my favorite RTS game, Battle for Middle Earth 2. The several family members and friends with whom I've played that game prefer to team up against bots in skirmishes or campaigns.
Michael de Plater, the creative director of Tom Clancy's Endwar, has cited that upcoming game's use of voice commands as a way to counteract the advantage of knowing macros and complex controls, but difficulty of control isn't half the problem (in fairness, de Plater doesn't suggest that it is). Most competitive online gaming networks feel as if professional, college, and high school sports conferences were combined into one. More than a few factors separate the amateur from the professional.
CoD4 admirably matches pros with amateurs in team matches, but I think the amateurs more often feel more useless and outmatched than like they're contributing to the team. That's how my friend felt. That's how I feel sometimes... especially when another player clearly knows the map, and I don't. The greatest sense of accomplishment comes from winning; nearly losing, but winning. Losing occasionally or even half the time is no problem. Losing almost every time is not fun for most people (unless it's to a pretty lady, of course). Relatively few have the personality type that views nine losses as an acceptable price for one win, when he or she is only competing for entertainment. One can view his or her own mild contributions as meaningful contributions, but that is, perhaps, not an automatic or particularly common view.
The skill levels and playstyles of players are hard to quantify, so matchmaking systems only help so much. In fact, matchmaking is downright useless when players alternate between play-options. A player might be good with some weapons/units, but not others; with some factions, but not with others; on some maps or scenarios/modes, but not others. And suppose I've improved my skills in single-player mode since the last time I played multiplayer. Should I now have to begin facing easy human opponents, because I haven't yet proven myself to matchmaking system?
Furthermore, a person's playstyle and desired degree of investment in the game can change from session to session. A gamer can feel particularly competitive one day and be interested in only leisurely engagement the next (perhaps due to being tired after a long day's work).
Remember that the difference between a professional and amateur isn't just skill and experience, but interest. A lot of people play casual club sports -- not because they're not capable of more intense performance -- because they prefer a different pace, different style and rules, or do not consider the game a high priority.
Finally, though I'm generally a big fan of reducing predictability in games, bots can actually offer a deeper sense of strategy than human opponents precisely because they're more predictable. Half of battle strategy is anticipating your opponent's next move. When you're playing against hundreds or thousands, if not millions, of other human beings, then you rarely face the same opponent more than once or twice. As a result, you can only learn and adjust to the broad strategies that are common among players, and not the specific strategies of individuals.
Online multiplayer games are more tactical than strategic... the player must react to momentary events, rather than plan for organized movements. Bots (particularly, campaigns against bots) enable true strategy.
PAUSE
Another thing I had to explain to my friend about CoD4 multiplayer was why one of his matches was ended abruptly when "the host... left the game". Sure, there are plenty of inconsiderate players who cut a game off just because they didn't like the map or the way the match was going, but there are also players answering telephones or doors, being distracted by kids or pets, or having to unexpectedly end their play-sessions.
As Gaming Verdict points out, the ability to pause is easier with bot-driven games. Games could even be designed to automatically replace a player's character with a bot if that player has to leave the game temporarily. Life is unpredictable. A purely logical software program is more likely to understand that than a stranger.
I realize that developing fun and competitive AI is no quick or easy task. But remember that even simple AI can be fun when combined with the right dynamics (ex: Turok shows that even a dinosaur that simply runs straight for you is tough when there's also a soldier nearby shooting at you).
We miss our bots. Give them back.
Friday, March 21, 2008
would real A.I. be perceived as real?
A question entered my mind as I played a little football today. As has been pointed out before, an accurate simulation of human intelligence would be capable of mistakes. But do you think gamers can ever be expected to perceive the mistakes of their NPC companions and teammates as acceptable... or even realistic?
Frankly, I doubt it.
When an NFL wide receiver doesn't catch a pass, when a World Cup goalie loses sight of a corner kick, when a basketball player misses a free throw... fans, in my experience, are generally not very forgiving. Most fans don't think "Well, even a professional's going to mess up sometimes... ". No, they think more along the lines of "C'mon, you idiot! Even I could have caught that!". Right?
If most people don't keep in mind that even the best trained and most experienced veterans will mess up and have off days, then how could we possibly expect gamers to accept mistakes from NPCs... even if those mistakes are intentionally included as realistic variables?
Every long-time gamer has had an experience that makes us say "the game cheated". Sometimes, we're right. Games do cheat. But it's often the same sort of response as yelling "Catch the freakin' ball!" while watching our favorite quarterback shake his head in disbelief (even the players have trouble accepting mistakes). We habitually choose to believe the simplest answer, rather than the truth.
If we're so willing to doubt our flesh-and-blood comrades, then NPCs don't stand a chance in hell.
Frankly, I doubt it.
When an NFL wide receiver doesn't catch a pass, when a World Cup goalie loses sight of a corner kick, when a basketball player misses a free throw... fans, in my experience, are generally not very forgiving. Most fans don't think "Well, even a professional's going to mess up sometimes... ". No, they think more along the lines of "C'mon, you idiot! Even I could have caught that!". Right?
If most people don't keep in mind that even the best trained and most experienced veterans will mess up and have off days, then how could we possibly expect gamers to accept mistakes from NPCs... even if those mistakes are intentionally included as realistic variables?
Every long-time gamer has had an experience that makes us say "the game cheated". Sometimes, we're right. Games do cheat. But it's often the same sort of response as yelling "Catch the freakin' ball!" while watching our favorite quarterback shake his head in disbelief (even the players have trouble accepting mistakes). We habitually choose to believe the simplest answer, rather than the truth.
If we're so willing to doubt our flesh-and-blood comrades, then NPCs don't stand a chance in hell.
Subscribe to:
Posts (Atom)
