Like Garry Kasparov, Roger Federer may one day have to compete against a massive supercomputer. Technology is being developed to use computer models of tennis and other sports to formulate the best playing strategy.

The idea is to create virtual athletes with software that tracks actual player movements in conventional video footage. From this data, the program will attempt to recognize individual tactics, as well as overall styles of play.

Users scoping out an opponent will be able to try out different counter-tactics without breaking a sweat.

Tennis coaches, for instance, might be interested to know what would happen if their star player came to the net more against an adversary who tends to stay back on the baseline. Or a doubles team might want to see how a different formation would work against their toughest competitor.

In the past, computer scientists considered sports tactics too complicated to be modeled, but recent developments have improved the tracking of multiple objects/athletes that are interacting with and responding to each other.

If this sounds like research for a video game, it certainly could be.

Ahmed Shihab from Kingston University in London, who leads the project, said, "As well as helping specialized sports training, the technology we are developing could have benefits in fields such as realistic computer gaming, virtual reality and surveillance, which also involve coordinated human activity."

After getting their feet wet with tennis, the researchers plan to move on to more complex sports, like basketball. Computer programmers might then be able to simulate Hack-a-Shaq.