Those who are under attack naturally focus on the most immediate threat. This week, we are all focused on the bombing of the Manchester Arena, the bloodiest example of extremist violence in the UK for 12 years. There is shock, grief and of course fear.

That fear is understandable. It is highly improbable Salman Abedi, the 23-year-old university drop-out who killed 22 people and injured scores more, could not have built the lethal device or devices that wrought such havoc alone.

This explains the raising of the threat level in the UK to its highest level — that which indicates an attack is thought to be imminent. The concern is that Abedi’s accomplices will try to strike before they are caught in the police dragnet. The threat level has only been this high twice before: after the discovery of a plot to bring down a dozen transatlantic airplanes, which was being run from an apartment in east London in 2006; and following the Glasgow airport attack a year later.

Each time it was lowered again after a few days. This time it could rest at this maximum level until the election on June 8.

There is no doubt that the UK is now facing a significant danger. Polls have historically encouraged militants to seek extra publicity by attacking during electoral campaigns. The Muslim holy month of Ramadan, which starts on Friday, has also often seen a spike in terrorist activities as Isis and others tell their followers their efforts will be rewarded more generously in paradise if they die then.

There is also the steady stream of military defeats for a crumbling Isis in its Middle Eastern strongholds which have convinced some of the “foreign fighters” who had joined the group to return home. Nearly 900 Britons are thought to have travelled to Syria to join Isis.

Only a third are thought to have died, so many are unaccounted for. Experts have long predicted that the decline of the group would lead to a spike in violence in the West, and this may now be occurring.

One reason is that the slow but steady progress made by the military campaign against Isis in the Levant may also have convinced some aspirant jihadis not to travel to the Middle East at all. This too is a problem.

Some may seek to execute their violent ambitions closer to home. Others may exploit other connections to travel to conflict zones. It appears that Abedi travelled frequently between the UK and Libya, where his family was from and where many members still live. He may have returned from north Africa only shortly before his attack. Or he may even have used contacts in Libya to reach Syria. Isis and al-Qaeda are active in both countries.

He would not have been the first to follow the Europe-Libya-Syria route. A series of French militants have reached Isis in this way, including several who went on to play significant roles in the external operations of the group.

Abedi’s foreign travel will underline the failure of the security services to identify him as a threat. In so many aspects, he fits a classic profile of the contemporary attacker — male, 23, of immigrant origins, with friends who apparently include known militant organisers, not wealthy but not destitute. He has been reported to have been involved in gangs — which is typical of many recruits to Isis — and to have developed a recent and abrupt interest in extremist strands of Islam.

The reason he wasn’t picked up as an imminent threat was that there are simply thousands of people like him. Security services have learned that profiling doesn’t work, because the cohorts generated are too large and because pathways into violence are so unique and unpredictable. Abedi was seen, like Khalid Masood, the Westminster attacker, as peripheral.

The problem is that engagement in extremism is dynamic, and someone who is known one day as marginal can become central almost overnight.

Another problem is the rapid evolution and variety of links between militants and overseas groups. The vocabulary here — “lone wolves”, homegrowns, directed/inspired — is clearly inadequate. Nothing captures the diversity of ways individuals interact with ideologies and organisations. This too complicates the task of the spooks.

None of this is reassuring, but the long view helps us be more sanguine about the future. Yes, the UK has suffered two murderous attacks in just two months, and two major spikes in violence in 2004 to 2006 and now.

But the total deaths from Islamist militant violence are still less than 100 and our security services, and the police, have been successful at stopping the vast majority of plots.

There will no doubt be further attacks in the years to come, but the biggest risk remains over-reaction to a threat which is certainly concerning but which still remains far from an existential danger.