The idea that grading is by shooting-only methods doesn't explain why some lots that are ostensibly the best off the production line prove in practice that they aren't -- either over the chrony or on the target. It also fails to explain why lesser graded lots can in practice be much better than they should -- both over the chrony and on the target.
Sample size is based on statistics, right?
Statistics is mathematics based on the accumulation of data from previous experiences.
Sample sizes are based on the levels of confidence desired.
the higher the level of confidence, the larger the sample size needs to be.
Lot grading through statistical sampling does not guarantee the quality of entire batch.
Using the calculator linked above, note the differences in required sample size...
25,000 cartridges 80% +/- 5% level of confidence ...163 cartridges to test
25,000 cartridges 80% +/- 1% level of confidence...3520 cartridges to test
25,000 cartridges 95% +/- 5% level of confidence ...379 cartridges to test
25,000 cartridges 95% +/- 1% level of confidence...6939 cartridges to test
The reason for poor quality cartridges from factory tested/graded lots
is going to be due to the confidence level used to set the parameters.
Confidence level is not a guarantee, only a calculation based on results from previous testings.
To obtain a 95% level of confidence with 0% level of error, you have to shoot all 25,000 cartridges.
Statistics are a tool to assist in quality control, but if the sample size is too small, you can't trust the results.
That's why we get Tenex that spits strays and shows an ES of 73 fps. Testing is expensive.
Do the absolute minimum possible to maintain profits and still be able to claim the ammunition was graded at the factory.
That's why those batches tested at the factory run tunnels have such a variation in results.
If the grading at the factory was effective, X-Act or Tenex wouldn't show 23mm outside to outside
on those computer generated test results. That ain't rifle preference, that's poor quality cartridges.
Spend some time shooting across a chronograph with any brand or label of rimfire,
y'er gonna end up very skeptical with the labeling on those boxes of cartridges.
Hard to argue that the ammunition is high quality, when the chrony shows the variations in muzzle velocity