I disagree with many of the so-called "truths" of professional grading services stated here. It's one thing to be a professional grader and push out product when you have no financial stake (other than your job!). But it's clear to me that former professional graders buying coins as dealers spend a lot more than 10 seconds a coin when it initially grabs them to look closer. Then they go in for a look with 5X....almost never unaided...esp when it comes to higher end coins for the grade. They are searching for all possible flaws that could prevent an upgrade or a premium price....something the TPG's really don't have time to do. Sure, sometimes the price of the coin doesn't require even a 5 sec look, just a glance on the slab grade. But for something not in the normal "I'll buy it at that price" range... it warrants a closer inspection. I'm not talking MS65 common Walkers, Saints or Morgans here but maybe a gem seated dime or half.
so take 2 top-notch graders and they will agree 80% of the time take 3 top-notch graders and 2 will be the same 97% of the time
Theoretically and mathematically speaking that would be correct. Even 3 graders at 70% accuracy should net you 97% consistency as a group. But clearly we don't have 97% consistency...at least not in UNC/PF type coins. That would indicate something is not working to the model (which is obvious since I feel type coin grading is about 60-80% consistent at the TPG's...that would indicate one grader or at least 2 graders not working well together or not in agreement). I've had shipments unc of seated coins come back 80% misgraded such that the next time through nearly all change a grade. That's not consistency no matter how you cut it.
Since a top grader can net 80% by himself it makes no sense that we can't get 85-95% consistency on US type...but we don't. I'll leave the sources to you as I have my own thoughts to the reasons.
The bottom line to me is the finalizer. One top notch finalizer at 80-85% accuracy should be able to keep the dreck or hack jobs from getting out just by himself. When the other graders blow the grade, send the coins back to them. Start it over again.
Grading on type coins was better (more consistent) in the 1980's. The expertise of classic coin graders was better and there were more of them grading. We definitely do not have that level of experience any more with maybe the exception of the finalizers. It has to hurt (or help depending if you are buyer or seller). And this has to be expected considering that the vast majority of the coins being graded these days are 20th century or moderns. That was not the case in 1989 when most coins submitted were classics (or pre-1930 or even pre-1916).
Comments
so take 2 top-notch graders and they will agree 80% of the time
take 3 top-notch graders and 2 will be the same 97% of the time
Theoretically and mathematically speaking that would be correct.
Even 3 graders at 70% accuracy should net you 97% consistency as a group. But clearly we don't have 97% consistency...at least not in UNC/PF type coins. That would indicate something is not working to the model (which is obvious since I feel type coin grading is about 60-80% consistent at the TPG's...that would indicate one grader or at least 2 graders not working well together or not in agreement).
I've had shipments unc of seated coins come back 80% misgraded such that the next time through nearly all change a grade. That's not consistency no matter how you cut it.
Since a top grader can net 80% by himself it makes no sense that we can't get 85-95% consistency on US type...but we don't. I'll leave the sources to you as I have my own thoughts to the reasons.
The bottom line to me is the finalizer. One top notch finalizer at 80-85% accuracy should be able to keep the dreck or hack jobs from getting out just by himself. When the other graders blow the grade, send the coins back to them. Start it over again.
Grading on type coins was better (more consistent) in the 1980's.
The expertise of classic coin graders was better and there were more of them grading. We definitely do not have that level of experience any more with maybe the exception of the finalizers.
It has to hurt (or help depending if you are buyer or seller). And this has to be expected considering that the vast majority of the coins being graded these days are 20th century or moderns. That was not the case in 1989 when most coins submitted were classics (or pre-1930 or even pre-1916).
roadrunner