One of the factors affecting group sizes is the design of the pellet itself. Pellet shape and design affect more than just the drag and velocity drop, they also affect how it flies through the air and how it reacts to different stimuli. Suppose you are the perfect shooter using the perfect gun. Each shot you fire you control the recoil exactly the same as the last one and you aim the gun at exactly the same point. The gun also behaves in exactly the same way every shot. You still will not get every pellet flying through the same hole in the target due to small variations in the pellet size, construction and shape. The pellet design will control how the pellet reacts to these small variations and the size of the resulting group. When a pellet leaves the barrel of a gun it will not leave perfectly. The pellet may be pointing at a small angle to the line of the barrel due to being for example slightly asymmetric, known as a yaw angle, or, more commonly, the pellet may be pointing in the line of the barrel as it leaves but will immediately start to change the angle in which it is pointing due to many different things, both in the gun and on the pellet outside the gun. The speed at which the angle changes is known as a yaw rate. How the pellet reacts to the yaw or yaw rate is what is decided by its design and what decides the resulting group size. There are two ways of achieving a small group size with a pellet. One is to make your pellet to a very exact size and shape and concentrate on manufacturing each and every pellet to that exact size and shape. This will give a very consistent launch from a rifle provided the pellet is a good match to the barrel. Matching the pellet and barrel dimensions coupled with exacting tolerances in the pellet shape, centre of gravity etc. will produce very small yaw angles and rates, which will mean that each pellet will fly in the same repeatable way. The problem with this method is that as dies wear or if the pellet dimensions are not an exact match to a barrel yaw angles and rates will increase and with them the group sizes. The second method is to produce a pellet design which concentrates on reducing the effect of the yaw angle and rate. In this case, instead of trying to minimise the yaw angle and rates, you accept that they will exist and minimise their effects on the pellet trajectory, you make the pellet more tolerant of errors. For example, suppose you had two pellets which both came out of the barrel with the same yaw angle or yaw rate, the pellet designed to be more tolerant of any type of yaw would produce the smaller group. There will still be some benefits from exact manufacture and barrel pellet matching but the benefits will be reduced. The advantage is that with the yaw tolerant design you potentially reduce the need for pellet barrel matching to achieve a certain group size. Using a pellet designed by the second method does not mean the ultimate performance will be improved compared to one designed by the first method. If there is no pellet yaw angle or rate then there is nothing to improve upon until the dies start to wear etc. There is also the danger that, in making a pellet which is more tolerant of launch errors, you may well increase the pellet yaw angles and rates as it leaves the barrel. In such a case the group size may be bigger or no better than that from a pellet which has not been optimised in the same way. The conventional diabolo pellet shape is not a good ballistic shape from the point of view of minimising the effects of yaw angles and yaw rates. Small group sizes are achieved through precise manufacture and barrel pellet matching. As soon as there are relatively small amounts of yaw in any form the group sizes will start to grow. The stability problems associated with pellets at high speeds and longer ranges are due to a different set of design properties, linked to but not solely caused by the same problems as the sensitivity to yaw angles and rates.