With a springer, less than 1" is good going.
The reality is that many people shoot "groups" that are way too small in number of shots and then claim nonsensical group sizes. What they have is a small sample, and the smaller the sample the smaller the size of the "group" and the outlandish accuracy claims that follow. It's all to do with a misunderstanding of statistics and randomness.
Personally think any group of less than 10 shots is pretty worthless as an accuracy test, and as Brian says, or a modification thereof, flyers should only be counted as such if it is very evident that something seriously wrong has happened, i.e. a pulled shot, perhaps a gust of wind out of no-where, etc, not as a convenient means of condensing a realistic group size into something from the realms of fantasy.
**WANTED**: WEBLEY PATRIOT MUZZLE END; Any Diana/Original mod.50 parts, especially OPEN SIGHTS