Posted: Wed May 09, 2007 7:36 pm
AAlex,
Interesting comments. When I first got my pistol, I changed my sights a lot. But that was because it wasn't correctly sighted in when I bought it.
Now, as a practical matter, the only time I change my sights is when I change from shooting one-handed to two-handed, or when I change the distance at which I'm shooting (wind isn't a factor where I shoot).
I find that when I'm shooting well (for me), my shots almost always seem to go where I think they went. Lately, I've been having an occasional problem with flinching. When I throw a shot low-right, I pretty much know where to look for the hole. When I throw a shot to the left, it's almost always a surprise.
When I get tired, or my NPA is off, I have to exert more effort to put the shot where I want, and my groups open up. But, when I practice, I shoot 50 or more shots at the same target, and the center of my groups hardly varies at all, from target-to-target, from day-to-day, or from week-to-week.
My point is, I know from experience that whatever temporary flaw in my technique caused me to throw a couple of shots in a row in a certain direction, it is temporary, and will even itself out over the course of a match, or a 50-shot practice string.
If I adjusted my sights every time I threw a few shots to the left, then a few shots later, I'd have to undo the adjustment when I started shooting to the right. If I were to shoot what would have been a 10 at that point, it would end up as a 7 to the right. And if I shot what would have been a 7 to the right, it would turn in to a 3 to the right. My scores would suffer considerably.
There is no point in chasing transitory, random variations in shot placement -- and yes, they are random unless you can tell for certain what caused them. This strategy only reduces the effective accuracy of your pistol. And my experience tells me that longer term shifts in POI just simply don't happen.
Shot placement does indeed follow a 2D normal distribution. And triggering errors don't cause fat tails -- they occur randomly also and they just make the variance larger. They do, however, result in a fat-tailed distribution of the size of your groups. So, for accuracy testing, simply shoot a relatively large number of shots and drop some percentage of the outliers. This dramatically reduces the variation in the size of your groups. In fact, the military drops the worst 50% of the shots, which results in a very stable measurement of group size/accuracy with a relatively small number of shots. As a practical matter, this isn't feasible at short range with something like an air pistol, because it isn't possible to distinguish the worst 50% of the shots.
There are far too many variables that go into balancing on two feet, holding a pistol at arm's length, controling your breath, minimizing your wobble, etc. for anyone to be concious of all of them at the same time. So, while in some philosophical sense, nothing about shooting is truly random, as a practical matter, it makes a lot more sense to approach it with a probabilistic perspective.
Regards,
Al B.
Interesting comments. When I first got my pistol, I changed my sights a lot. But that was because it wasn't correctly sighted in when I bought it.
Now, as a practical matter, the only time I change my sights is when I change from shooting one-handed to two-handed, or when I change the distance at which I'm shooting (wind isn't a factor where I shoot).
I find that when I'm shooting well (for me), my shots almost always seem to go where I think they went. Lately, I've been having an occasional problem with flinching. When I throw a shot low-right, I pretty much know where to look for the hole. When I throw a shot to the left, it's almost always a surprise.
When I get tired, or my NPA is off, I have to exert more effort to put the shot where I want, and my groups open up. But, when I practice, I shoot 50 or more shots at the same target, and the center of my groups hardly varies at all, from target-to-target, from day-to-day, or from week-to-week.
My point is, I know from experience that whatever temporary flaw in my technique caused me to throw a couple of shots in a row in a certain direction, it is temporary, and will even itself out over the course of a match, or a 50-shot practice string.
If I adjusted my sights every time I threw a few shots to the left, then a few shots later, I'd have to undo the adjustment when I started shooting to the right. If I were to shoot what would have been a 10 at that point, it would end up as a 7 to the right. And if I shot what would have been a 7 to the right, it would turn in to a 3 to the right. My scores would suffer considerably.
There is no point in chasing transitory, random variations in shot placement -- and yes, they are random unless you can tell for certain what caused them. This strategy only reduces the effective accuracy of your pistol. And my experience tells me that longer term shifts in POI just simply don't happen.
Shot placement does indeed follow a 2D normal distribution. And triggering errors don't cause fat tails -- they occur randomly also and they just make the variance larger. They do, however, result in a fat-tailed distribution of the size of your groups. So, for accuracy testing, simply shoot a relatively large number of shots and drop some percentage of the outliers. This dramatically reduces the variation in the size of your groups. In fact, the military drops the worst 50% of the shots, which results in a very stable measurement of group size/accuracy with a relatively small number of shots. As a practical matter, this isn't feasible at short range with something like an air pistol, because it isn't possible to distinguish the worst 50% of the shots.
There are far too many variables that go into balancing on two feet, holding a pistol at arm's length, controling your breath, minimizing your wobble, etc. for anyone to be concious of all of them at the same time. So, while in some philosophical sense, nothing about shooting is truly random, as a practical matter, it makes a lot more sense to approach it with a probabilistic perspective.
Regards,
Al B.