[an error occurred while processing this directive]
BBC News
watch One-Minute World News
Last Updated: Tuesday, 28 August 2007, 10:01 GMT 11:01 UK
State of Play: Man versus machine
By Margaret Robertson
Video game consultant and writer

Bot that plays Guitar Hero
Researcher Kevin Herron has built a bot that plays Guitar Hero

Computers playing computer games may sound odd, but it's an example of the ongoing man versus machine debate.

To the outside observer, all games can seem pretty pointless. But even to gamers who know they're not, there's a whole subset of pointlessness.

To some, it's investing in a subtle, taxing game like Virtua Fighter and never doing more than bashing random buttons in the hope of a win.

For others it's buying a guide book to a Final Fantasy game and blindly following its instructions as you play, rather than the story or your own sense of adventure.

These ways of playing - staunchly defended by some - do seem to rob gaming of nearly everything it has to offer.

But there is an even greater level of pointless play that baffles almost everyone. These are the systems devised by people who are more interested in getting machines to play for them than in playing themselves.

It's a vibrant, but idiosyncratic, world, from the man who's hoping his genetic algorithms will eventually evolve into Dr Mario masters, to the frustrated Guitar Hero who built a bot - complete with a camera for eyes and virtual fingers - to complete the hardest tracks for him.

It's a little surreal, but watching both of these systems play is strangely compelling, and there's an odd sense of vindication to be found in their triumphs. It's literally a case of seeing something beaten at its own game.

So what's going on? Surely nothing could be more pointless than watching a machine play a machine.

Gamers have to put their faith in their heroes every time they play

The reality is that it's not as alien an idea as it might at first sound. Gamers are used to watching games try to beat themselves.

Arcade games have always traditionally had "attract modes" that demonstrated how the game should be played, and ever since it's been possible to see games play themselves.

Even stranger, most gamers are used to games which do their best to help you win while they try all-out to beat you.

Many gun-based games have an auto-aim option, which means the game is busy processing how best to swing your reticule over the target at exactly the same time it's trying to process how to get enemies to pop up in corners you didn't expect.

Puzzle games take a few precious cycles out from relentlessly crunching the numbers on winning strategies to offer you helpful tips. And, we, weirdly, are totally at ease with this schizophrenia.

Why on earth would we trust a machine that seems to be playing both sides against the middle?

Why don't we ever wonder if the game is using a slightly dumber AI to proffer tips than it is to guide its own moves?

Mario
Genetic algorithms are being used to "teach" computer to play Nintendo games

Can we be sure the so-called auto-aim isn't doing the software equivalent of jogging our elbow every time we shoot?

We can, but only because there's an instinctive understanding that - despite all the indications to the contrary - the games aren't trying to win.

They aren't machines determined to beat you, they're machines determined to entertain you. And it's reasonable to put your trust in them, because it's not them you're really trusting.

It's tempting to see games as the front-line of the great man versus machine debate. As the quest to build computers than can outsmart humans marches on, games remain the high-profile acid test.

A decade on from Kasparov's famous defeat by Deep Blue, computer programs such as Rybka can routinely beat players at Grand Master level.

Human intelligence, refusing to be put on the back foot, is taking refuge in creative ingenuity

Human intelligence, refusing to be put on the back foot, is taking refuge in creative ingenuity, inventing games like Arimaa, which are specifically designed to be easy for humans but hard for artificial intelligence.

But for most people, and for most games, that sense of pitting the human against the artificial isn't what you're aware of.

When you play, what you're doing is going up against the men behind the machine.

That's why watching bots and algorithms get the best of a game is so satisfying. For years, the machines that game developers build have been getting the better of us. Now our machines can, if only occasionally, get the better of them.

But at its best, gaming isn't about our machines or their machines.

At its best, the machine disappears, becoming instead an invisible conduit between you and these dozens of men and women who've spent years bending their immense imaginations and arcane skills to creating something designed purely to make you happy.

You put your trust in their instincts, and in their sincerity in trying to build an extraordinary, satisfying experience for you.

It's one of the ways in which game players have a much more intense relationship with game creators than is true of consumers of most other artforms.

Directors, writers or musicians can let you down, by making something you don't like, but they can rarely betray your trust.

Gamers have to put their faith in their heroes every time they play.

It's perhaps part of what enables game companies to engender the kind of fanatical loyalty that Nintendo does amongst its fans, and what accounts for the viciousness of the vitriol that unpopular products can produce.

Indeed, when you read the kind of hate-mail some studios are subject to, you have to wonder if they wouldn't prefer to be making games for bots and genetic algorithms. You rather suspect they might have better manners.



RELATED INTERNET LINKS
The BBC is not responsible for the content of external internet sites



FEATURES, VIEWS, ANALYSIS
Has China's housing bubble burst?
How the world's oldest clove tree defied an empire
Why Royal Ballet principal Sergei Polunin quit

PRODUCTS & SERVICES

Americas Africa Europe Middle East South Asia Asia Pacific