Killing by Numbers; Part 1

Statistical reanalysis of The Troubles



Over the course of the past 2 years, myself and my fellow ISOT ‘admin’ have re-examined approximately 75-80% of all Troubles related deaths, those attributable to either Loyalist or republican armed groups.

Our objectives were relatively simple-

1: To ascertain who exactly both sets of ‘combatants’ regarded as being “legitimate targets“.

2: To calculate how many of the aforementioned “legitimate targets” each armed group actually killed.

3: To calculate, or in some instances estimate, how many “legitimate targets” were available over the course of the conflict.

4: To calculate, as a percentage, the total number of “legitimate targetsactually killed by each ‘side’


At first glance, this seems a straightforward, if somewhat time consuming, task. In reality however, it is nowhere near as simple as it sounds. Take, for instance, the question of just who was regarded as a “legitimate target”. At first glance it seems easy to answer but it is, in fact, anything but.

At different stages of the conflict, different groups of people were regarded as “legitimate targets” by various other groups of people who wished them harm. Whilst at other times such people were left alone by those who had previously, or who would in future, seek to kill or seriously injure them.

One also has certain ethical issues to address when undertaking such a study. Often times armed groups (on both sides) would claim that a murder victim was a “legitimate target” because they, the killers, believed their victim to have been a so-called “legitimate target”.

Does the researcher take all such claims at face value? (where proof of such claims even exists). Should such claims be disregarded in instances where they later proved to be unsubstantiated, or unproven? Should such claims be accepted only where there is some evidence of their veracity, however flimsy, to back them up?

People delivering milk to their local police station were considered “legitimate targets” by Irish republican extremists.


It was clear from the beginning that clear parameters would have to be set and that certain rules would have to be applied. Therefore we agreed to the following-

a) All claims, with regard to the status of victims, would be taken into account only where there was some evidence to substantiate them, even if such evidence was suspect or circumstantial.

b) Certain groups of victims would not be regarded as “legitimate targets” in the usual way, but would not be included as “innocent victims” either*

[*this applies only to those groups/agencies who were regarded by a particular organisation as “legitimate targets” only periodically, or only under certain circumstances, those members of armed groups killed by their own actions or the actions of their comrades, those, both civilian and members of armed groups, who were killed because they were alleged informers, or those who were killed attempting to prevent an attack]

Once the number of potential targets was known (or reasonably estimated) we set out to ascertain the actual number of persons killed who fell into the “legitimate targets” category and to then arrive at a percentage.

For example, if Group A regarded the British Army in Northern Ireland as “legitimate targets”, we would calculate how many Army personnel, including former soldiers, Group A had killed; let us say, purely hypothetically, that number was 200. We would then try to arrive at a figure for potential targets, that is; the number of British Army personnel in Northern Ireland during the years within which Group A was active.

If, for example, Group A had begun their campaign in 1975 and ended it in 1998, we would need to calculate (or reasonably estimate) the number of Army personnel either serving, or resident, in Northern Ireland during those 23 years.

We know that during the entire length of ‘Operation Banner’ 300,000 British troops were deployed in Ulster. If we subtract the number who served here before 1975 and after 1998, then we arrive at a figure of approximately 200,000 (this is a high estimate since troop numbers were at their highest between about 1972 and 1994). It is then easy to work out that, from 200,000 potential targets, Group A killed 200, which is 0.1%.

We see, therefore, that our hypothetical Group A killed only 0.1% of the “legitimate targets” available to them.

Of course, no armed group targeted only one section of the Security Forces, or only one opposing group. Therefore, it is necessary to add together all potential “legitimate targets”, then to subtract the number actually killed in order to arrive at a final number and work out a percentage.

Parts 2 and 3



Parts 1 and 2 will focus on Irish republican groups which were active during at least part of the Conflict, namely- PIRA, OIRA, INLA and IPLO. We will not include CESA, since in the years that group was active, they were primarily a ‘vigilante group’ who carried out only a small number of offensive operations.



We will not include other republican “micro organisations“, such as Soar Éire, since they were responsible for only a handful of deaths and most of those were in the Irish Republic.

The so-called Republican Action Force and Catholic Reaction Force will be regarded as “flags of convenience”, used mainly by the Provisional IRA and INLA/IRSP to claim indiscriminate sectarian attacks, such as the Darkley massacre.



Parts 3 and 4



Parts 2 and 3 will focus on Loyalist armed groups and will only include those organisations which were active during at least part of the Conflict (1969-1998). Groups such as the PAF will be included but will be considered to be a part of the Ulster Volunteer Force.

The Ulster Defence Association and the Ulster Freedom Fighters will be considered as a single organisation.

Part 2 will be published next week.