PCGS MS-66.1 Or MS-66.9 ?

Here's another one for you. I own a 1878 8TF that appears to be a gorgeous Gem, MS66 or MS67, but on close examination you can see minute rubbing or wear if you like on the breast feathers. What you would call a real slider. There is no wear on the Obverse whatsoever and no discerenable bagmarks or hairlines. Absolutely sharp srike and beautiful frosty luster. This coin has not been polished, whizzed, cleaned, etc. I've showed it to many reputable dealers and fellow collectors and one thing we all agree on is it is impossible to put a grade on. I paid MS63 price for it because of it's beauty. Now, I could send this off for TPG grading and get a AU58+ or CAC bean and it sill would not be a correct valuation or evaluation of this coin.

The Sheldon Scale is already too granular at 70 grades. The + designation nearly doubles it to 139 (as I do not believe they give out a MS70+).

There needs to be some sort of more objective and quantifiable overhaul and the Librarian in me is crying out to do something...! -- But the Sheldon Scale is already pretty ensconced in the hobby and changing it now would be next to impossible...

...unless you're PCGS and decide to add a +.

Good grief.

There's the rub. If there is a 5-10x multiple in price between grades, there is a legitimate demand for more "precision" in grading.

The problem is, that doesn't exist. Scales can be accurate, precise or both, but just because they are one doesn't mean the other follows.

You can buy a scale that is precise to 0.001 g, but if a 100 g weight reads 105.072 g and a 100.1 g weight reads 105.172, it's precise, and even consistent, but it's certainly not accurate. Another scale may only weigh to the nearest gram, but hit both those weights at 100 g every time, so it's accurate, but not real precise.

With grading, anyone here can grade. A six-year-old can grade. The question is repeatability (accuracy) and consistency (precision).

Given 100 coins, even the kid can arrange them from worst to best. The same grader may be able to jumble and re-sort them in substantially the same order. That's the first hurdle.

The second jump is to say #16 is good, but #17 is vg and #94 is ms64. Or maybe they were a couple rolls of unc cents and #16 is ms61, but #17 is ms62 and #94 is ms67.

Enter grader 2, who can do the same line-up, etc. That's easy. How likely is he to put them in the same order? Maybe plus of minus 10, but perhaps substantially differently.

Hurdle 3 is for grader 2 to look at grader 1's order, and decide where to make the grade breaks. Not easy, especially if he'd have rated certain coins significantly higher or lower.

Finally, both graders have to match some sort of standard, as well as each other.

So we have two (million) graders, who will not be able to order the same group in the same order, and even if they could, couldn't match each other's order, much less agree on matching the standard.

So back to precision not existing. There are going to be coins all along the scale where reasonable graders can't even agree on g vs vg, or au vs ms. If one says ms64 and the other says ms66 (and we've seen wider ranges on "grade my coin" posts), exactly what does ms65.23 mean? Purty.

But if you think ms65.23 is a bit more precise than ms64 or ms66, you're only fooling yourself.

While poking around on ebay, I saw evidence that buyers look for jellybeans. Slobbed coins had one or two beans with stuff like "choice" or "nice" on them.

At which point I'd like to smack someone, but hey, why not?

That's because AU58+ or CAC is the accurate grade. That has nothing to do with value. I've bought AU coins for MS prices, and vice versa.