I often read people boasting of 2" or 3" patterns at 25 yards with handguns. When people compare such patterns, are they talking about from a free standing position, such as the Isosceles and Weaver stances, thus testing both their skill and the handgun's accuracy? Or are they talking about shooting with a braced or secured handgun, thus testing just the gun's accuracy? If it's free standing I'm impressed, and have a lot more practicing to do.
The fact that most such reports omit information critical to assess such a group's significance is telling. With a small sample size (eg, 5 shots), one can generate a tight group every so often on chance alone. Anyone who, say, shoots five 5-shot groups then reports only the tightest is deluding mostly himself. Without full disclosure results are meaningless.
Before I obtained my carry pistols I was mostly a handgun plinker, so not very concerned about precision. So when precision became relevant to me, I had in my mind the rifle benchrest gold standard of 1 minute of angle (moa), or a 1.047-in, 5-shot group at 100 yd. Well, I was, of course, disappointed with my off-hand pistol performance by that standard. That sent me scouring for relevant pistol performance standards.
The gold standard for semiautos seems to be a 2-in group at 50 yd (1-in at 25 yd), or 3.8 moa; but, this is from a Ransom rest. You can buy a custom pistol guaranteed to 1.5-in at 50 yd (0.75-in at 25 yd), or 2.9 moa. But a Ransom rest tells us about the gun, not the entire system -- gun and shooter. For defensive purposes, the shooter is generally the critical factor in determining precision, as most guns are much more precise than we are.
I collected varoious precision benchmarks on a spreadsheet, which I seem to have organized out of existence. As I recall, I concluded that a reasonable precision goal for off-hand shooting is 30 moa, or 7.9-in at 25 yd. I think I derived this from the size of the thoracic -0 portion of the IDPA target and an upper bound distance for that target. This would be an ideal goal for rapid fire, so a reasonable goal for slow fire at the range should be tighter.