Small arms, small data : small arms shooting accuracy and the small data problem

This paper explores how small-arms aiming errors appear to vary with range. In particular it investigates the question of whether a “proximity effect” exists, reducing accuracy at close range.

Data on small-arms hitting rates is sparse. The paper analyses nineteen sources of data, giving 83 data points; combat data is augmented with data from police shootings, range trials, OR models and qualification scores. The paper explains the assumptions made in order to compensate for gaps in the data, such as target size.

Data points are reduced to a common basis of angular error, which would produce the observed hitting rate if shooting at a visible static rectangular target at the stated known range. This subsumes all errors normally included in ballistic error budgets, plus the uncertainty of target location if firing at targets not clearly visible.

The data available indicates that the accuracy of small-arms fire decreases with proximity to the target, so that the hitting rate does not increase as much as would otherwise be expected at closer ranges. The effect seems to apply across different types of data source, weapon, and fire, up to about 100 metres.

The paper discusses possible explanations for the effect: prevalence of close terrain in the combat sample, restricted visibility, targets reducing their exposure time close to the enemy, uncertainty as to true target position, psychological stress due to enemy proximity, and a transition to pointed rather than aimed shooting. Each may be a contributory factor, but it is tentatively concluded that a hastier shooting style arising from psychological stress provides most of the explanation.

Finally the paper suggests some directions for future work, the most important of which is clearly to add to the data available on this subject.