As Islamic State knows, to change is to survive

This is true in a way, and success in one aspect - the protracted war which has pushed IS out of many cities and towns in Syria and Iraq - cannot be denied.

IS directly holds less territory now than at any time after its explosive capture of swathes of Iraq in 2014. But, as analysts such as Hassan Hassan point out, this does not mean IS is done.

An experienced terror organisation with a long history of adaptation, IS' leaders know, and their predecessors knew that to change is to survive. The timing is in dispute, but it is clear now that IS has either begun or completed its transition from semi-state to insurgency.

The current campaign against the terror group can still see some successes, but these will decrease in significance as time goes on. And this campaign does not have the tools or likely the will to counter an insurgent IS.

American patience, especially under Donald Trump, is not endless; and amid the general defeat of IS, the coalition ranged against it has already experienced notable infighting.

Now is the right time to begin a serious assessment of the anti-IS war, especially its American component. Journalists, perceiving a change in atmosphere, are beginning to write summative-seeming pieces about the campaign.

Now is the right time to begin a serious assessment of the anti-IS war

One of these efforts is a remarkable essay published in The New York Times Magazine by Azmat Khan and Anand Gopal. It takes for its subject the civilian casualties of American and coalition airstrikes, many of whom have been left out of official accounts and projections. These are the uncounted.

Building on extensive fieldwork in Iraq and Syria and the work of monitoring organisations such as Airwars, a London-based operation which tracks "collateral damage", the authors estimate that the number of civilian casualties of the anti-IS aerial campaign is likely five or six times higher than the official coalition estimates - and possibly greater still.

Read more: The 'Islamic State' brand is over

The authors attribute this rate to a mixture of intelligence failures, practical problems and a December 2016 "rule change [which] allowed more ground commanders to call in strikes, possibly contributing to a sharp increase in the death toll".

Other sources attribute a rise in estimated casualties this year to a loosening of the US campaign's rules of engagement under president Trump, something which Trump himself has claimed, albeit in a different context.

The number of civilian casualties of the anti-IS aerial campaign is likely five or six times higher than the official coalition estimates

But in truth, the factors which caused the civilian deaths featured in the Times piece were present in the anti-IS campaign from the beginning.

Under Barack Obama, the United States' effort to "contain" and defeat IS was shaped. It has remained in a similar shape since, despite presidential changes of emphasis.

That shape included a heavy aerial component supporting local "partner forces", whose job it was to do the hard fighting on the ground.

This strategy was always going to cause inconsistency in the fight against IS and turmoil after the group's removal from urban strongholds.

Not all partner forces are equal; and none of them are as well-equipped and well-trained as an American or NATO force would have been. And expecting others to fight and die on the Americans' behalf, while aiding these forces with overwhelming artillery and air power, has its own problems.

In Iraq, federal forces faced extremely heavy casualties - often in difficult to replace elite units - when retaking Mosul from IS. No wonder the Iraqi state sought to bring as much American firepower to bear on IS as possible; and no wonder, in other instances, the Iraqi state did not mind having the Popular Mobilisation Units (PMF) - militias, many of which are Iran-backed and openly sectarian - do the work of fighting IS and clearing settlements.

The case of the PMF also exposes a problem with the selection of partner forces. Due to the reality of the Syrian war in particular, but also the state of Iraq after years of internal fighting and Iranian power building within the country, the number of allies the United States had to choose from was limited.

And another resource was limited, too: Time. The Americans did not want to hang around. Both Obama and Trump pledged to end wars, not start them. They each hoped to beat IS extremely fast, and this set the tone.

Some partner forces were ill-chosen. Giving the PMF tacit American support was a grave mistake. So too, in Syria, was expecting the Kurdish-dominated Syrian Democratic Forces (SDF), to be a go-to American proxy, fighting in the north of the country and in liberating Raqqa - where either local Sunni Arab partners could have been enlisted or the forces of the Turkish-backed Operation Euphrates Shield offensive could have been better employed.

The American desire for speed, and the fact of having less than perfect proxy forces doing the majority of the fighting, did not help.

Expecting others to fight and die on the Americans' behalf, while aiding these forces with overwhelming artillery and air power, has its own problems

It meant that the Mosul offensive was ostensibly meant to take no time at all, and when it dragged bloodily on, and the pace of civilian-killing air strikes was significantly increased; it meant that the Raqqa offensive saw hundreds of air strikes hurled at a comparative small number of IS fighters inside a shattered city, with predictable consequences for the civilian population.

This is not to say that the campaign was misconceived or an outright error. Someone had to defeat IS, after all.

But this does not mean that the anti-IS campaign should escape criticism. It has been deeply flawed and some those flaws directly cost the lives of hundreds, likely thousands, of Syrian and Iraqi civilians. And as IS begins its insurgent stage, the way the United States selected and rewarded its partner forces will matter – and possibly set the tone of future conflict.

Despite the American attempt to fight this campaign at a distance and bring it to a close as fast as possible, its effects will remain, for years to come. And IS is not going anywhere. Let us hope American policymakers can learn from the many mistakes made in fighting it this time. They or their successors will certainly need to learn now, for use in the not-too-distant future.

James Snell is a writer whose work has appeared in numerous international publications including The Telegraph, Prospect, National Review, NOW News, Middle East Eye and History Today.

Follow him on Twitter: @James_P_Snell

Opinions expressed in this article remain those of the author and do not necessarily represent those of The New Arab, its editorial board or staff.