What that means is that when the pins were graded, they gave themselves criteria to class the pins. The "plus" pins may vary as much as 0.0002 on the plus side of nominal, and the "minus" pins may vary up to 0.0002 under nominal. Many people assume that the pin IS actually 0.0002 under or over, but that is not the case. Most of the pins I have checked are fairly close to nominal size, so some wear is expected and allowed to occur and yet the pins stay within the original limits.
Think of them as, for example, the 0.2500 pin MUST measure from 0.2500 to 0.2502 if it it is plus pin, and 0.2500 to 0.2498 if it is a minus. As you can see, there is a possibility that the pin could be, for all practical purposes, 0.2500 in either or both cases, but yet be acceptable as a minus or plus pin.
In actual use, always remember that there must be a certain amount of clearance for hand fitting the pins. You will not put a 0.2500 (actual) pin in a 0.2500 hole by hand. That's zero clearance and is considered a light press fit. Common practice is to drop 0.001 and say that if the 0.2500 will not go, but the 0.2490 goes easily, the hole is likely acceptable if specified with a 0.2500 minimum. Same rule applies on the high tolerance. If the high is specified to be 0.2500, the 0.2500 pin should not go. Now, having said that, IF the pin actually measures (as a minus pin) at it's own low tolerance (0.2498) it might be possible to get it in the hole if the hole is actually 0.02500. However, that would require a rare set of conditions that includes a very good surface finish in the hole, and the hole being very straight, and likely not very deep (or material not very thick). Most inspections are done with pins 0.001 under the required limits. Of course, in the event you have a +/- 0.005 tolerance, there certainly wouldn't be a problem if you used (on our 0.2500 example) a 0.2450 pin as a go, and a 0.2540 pin as a no go. The operator or machinist you are inspecting would be within his rights to balk a little on the use of a 0.2450 pin as a "Go", and arguably the hole could be on the very low end of tolerance and the pin would not go, but as good practice, we generally would not want the hole to hug either end of the tolerance. He should not have used a tool that produced a hole that close to the low tolerance. And it might be that the 0.2540 pin would go into a 0.2549 hole....true, and another argument could result. But in QC theory, the closer you get to the shipping dock, the looser the tolerances are. As a final inspector, you have the true high and low pins. On the machine, the operator should have been shooting for a +/-0.003, for example. That allows an error at the machine which still allows a chance for making an acceptable part. This mode of operation is subject to question in some cases, and lots of people do not like it, calling it unfair. But it is done in places.
I hope I did not merely confuse you. I'm pretty opinionated in QA/QC matters, having worked it for a few years.
