Should Israel or the United States bomb Iran? In newspapers and magazines, in blogs and tweets, on television pundit panels, the answer is clear and emphatic.

Yes, say many.

Iran has a long history of supporting terrorism and destabilization. A nuclear Iran would immunize itself from attack. It would become far more aggressive, using its nuclear arsenal to bully others into submission. Feeling threatened, Iran's neighbours would seek weapons of their own and nuclear proliferation would come to the world's most dangerous region. And given the religious fanaticism of Iran's leaders, and their obsessive hatred of Israel, Iran might give nuclear weapons to terrorists, to use against Israel, or attack Israel directly, seeing the annihilation that would follow as martyrdom.

Let the bombing commence.

No, say others.

Attacking Iran would only set back Iran's nuclear program a few years while making Iran's leaders all the more determined to get nuclear weapons. Future crises would arise, and more attacks would be needed, with each being more difficult as the Iranians learned to defend their facilities from airstrikes. Meanwhile, an attack would cause a surge of nationalism, strengthening the regime and undermining the opposition. Diplomatic options can still work. And even if they were to fail, and Iran were to get nuclear weapons, history shows it would not lead to proliferation and Iran would be unable to use the weapons to expand its regional power or destabilize neighbours. Nor would Iran actually use the weapons in an attack, either directly or via terrorists.

Bombing would be a terrible mistake.

So who's right?

In past columns, I've discussed some history that suggests the nuclear fears of hawks are overblown. But that doesn't mean I think the doves in the media are necessarily correct.

In fact, I think all the ultra-confident pundits - people such as Niall Ferguson and Tucker "Iran deserves to be annihilated" Carlson - are profoundly misguided. Hawks and doves alike. Their main contribution to public discourse is to provide textbook examples of how not to analyse difficult problems.

Look at all those claims, above, from both sides, about what will and will not happen. None is absolutely impossible. None is certain. That leaves probability judgments.

For each claim, a serious analyst has to ask "in event of X, how likely is Y?" Any answer between zero and 100 per cent is possible.

But there's more to risk than probability. The relative weight of various outcomes has to be considered. Remember the old formula: "risk = probability X consequence."

And notice that there are risks to attacking and risks to not attacking. Having clarified what the risks are on both sides, they must be weighed against each other. And a final conclusion drawn.

A serious debate would involve each side explaining the reasoning behind its judgments while trying to find flaws in the other side's. It would be complex (note that my summary of the arguments for and against attacking is greatly simplified). It would be nuanced. And above all, it would respect the ineradicable uncertainties involved.

Intelligence is sometimes wrong. Judgment can be skewed by psychological biases. What Don Rumsfeld famously called "unknown unknowns" can blow away calculations like dust in the wind.

This is not the realm of simple arithmetic. There may be better and worse answers. But no answer is correct.

Wisdom consists of thinking carefully and making the decision, knowing how easily error and uncertainty may lead to failure. Only fools are certain.

Of course we seldom think of decision-making this way. In part, that's because we look at the outcome of the decision and use it as proof of the decision's correctness. If the outcome is good, it was a good decision. If not, it wasn't.

As simple and right as that may sound, it's very misleading. Consider Barack Obama's decision to authorize the strike that located and killed Osama bin Laden.

Most people would say it was the correct decision. Obviously. The mission was a total success.

But Obama was told by the CIA that there was only a 60 to 80 per cent chance that bin Laden would be found in the compound. And he was told by the military that countless things could go wrong. The mission could turn into an international incident, a debacle, a humiliation. The end of Obama's shot at re-election.

Obama asked his advisers what they would do. "Every single person in that room hedged their bet except (CIA director) Leon Panetta," recalled vicepresident Joe Biden. "Leon said go. Everyone else said, 49, 51 (per cent)." Biden himself told the president he shouldn't do it.

Obama did do it. And it was a success. But it could easily have ended differently and Obama's decision, which now looks so right, would have looked very wrong.

Doubt that? In April, 1980, president Jimmy Carter ordered the newly formed Delta Force to carry out "Operation Eagle Claw" - the helicopter-borne rescue of 52 Americans held hostage in the U.S. embassy in Tehran.

The plan called for six helicopters. To provide margin for error, eight were sent in. But a sand cloud forced one helicopter to turn back and another to crash. Further setbacks were suffered. Some seemed trivial but cumulatively they devastated the mission. After "Operation Eagle Claw" was officially aborted, a helicopter crashed into a transport plane, killing eight servicemen.

The debacle cemented Carter's image as a bumbler and contributed to his defeat in the presidential election seven months later.

And yet, it could so easily have turned out differently. The wind could have shifted slightly. The sand cloud could have been just a little further away. The mission could have succeeded. And Carter could have been hailed as a courageous leader.

But the wind blew as it did.

In this tragic world, the wise statesman thinks very carefully, makes the decision, and holds his breath. Only fools are certain.