Sharing major international events is the fashion this year. Next week, the finals of the UEFA European Football Championship will kick off in two host countries: Poland and Ukraine. Meanwhile, scientists are digesting another split decision, on where to build the world's most powerful radio telescope.

The battle to host the Square Kilometre Array (SKA) was between South Africa and a joint Australian/New Zealand bid. Both sides fervently wanted the title, and rhetoric ran high in the weeks before the negotiation. But like many highly anticipated matches, it fell short and ended as a draw. Under the deal, half of the SKA will end up in Australia and the other half in South Africa (see page 555).

Despite some weak justifications from the political leadership of the project, there is no compelling scientific reason to build the SKA in two places. Indeed, the law of parsimony should lead any scientist to conclude that putting the SKA in a single location would be much better than putting it in two locations six time zones apart.

Scientists aren't protesting too much, yet. They understand that large-scale experiments follow rules other than those of reason. Take, for example, ITER, the massive fusion experiment now under construction near Cadarache in the south of France. In a logical world, that complex machine would be designed, commissioned and assembled by a single directorate near the site, but instead it is being run through a baffling system of contracting and subcontracting. And the world's most powerful particle accelerator, the Large Hadron Collider (LHC) at CERN near Geneva, has expensive data-analysis centres all over the world.

This added complexity reflects the human side to these large projects. In the case of ITER, the procurement system means that governments can ensure there is work for domestic labs and companies. Distributed computing at the LHC lets researchers who have contributed to its construction participate from their home countries.

But adding another layer of complication to an already sophisticated project adds risk and cost. ITER's costs have doubled, largely due to its inefficient structure, and procurement is running behind schedule. The International Space Station (ISS), is currently looking at ways to make cuts — perhaps by trimming its four, redundant operations centres.

For the SKA, administering two remote locations will be more difficult than one, and the extra roads, networking, computing and power requirements will surely raise the price. The split site might also threaten the cohesion of the project: it now looks much more like two telescopes than one.

But splitting the site does have potential benefits. For one thing, the redundancy created by international collaboration can come in handy. The ISS, for example, continues to be serviced by a slew of vehicles from its different partners, even though the US space shuttle no longer flies there. There is also a perverse financial advantage — multiple partners are less likely to cancel an over-budget project than is a single government. But the greatest benefit is human: a more complex project draws in more people from more places and gives them an opportunity to participate.

It is easy to view fights over projects such as the SKA as sport. Participants instinctively seek winners and losers, and want to walk away with a trophy. But an international project is far more complicated. In the case of the SKA, both bidders can walk away with their egos intact, and the project has held together. However unsatisfying the result might seem, it was probably the best way to see that the telescope got built. It is, as the football pundits will say next week, a game of two halves.