Options
PCGS Made a Big Mistake
gemtone65
Posts: 963 ✭✭✭
...when it designed its grading scale for gem toned coins. It effectively decided to ignore 3 gem grades: 68, 69, and 70. Moreover, most of the gem coins it does grade are heavily clustered in the MS65 category. So, when you send a gem coin in for grading, you can expect to receive a grade that implies a very wide variance in quality, and hence you're really not getting your money's worth from this entire process.
Let me demonstrate using mint state 1881-s Morgan dollars currently graded MS65 and better as an example. Suppose you had a bag of such coins, randomly put together from other bags. You could expect to find the following distribution of coins by grade:
Grade Number
65 786
66 188
67 24
68 2
----------
1000
Imagine instead that a somewhat different grading scale had been adopted. I suggest that one such better scale could have been created by moving the top 20% of the MS64's into MS65, the top 30% of MS65's become MS66, the top 40% of MS66's become MS67's, and the top 50% of the MS68's become MS69's. Following that adjustment, the distribution of gem common date Morgan dollars in a bag would look something like this:
Grade Number
65 644
66 276
67 71
68 8
69 1
----------
1000
I would argue that this scale, and others like it, provide much more helpful information about the relative technical quality of a coin compared to the current truncated grading scale. But, so what you may ask? It's too late to do anything about it, even in the unlikely event that PCGS would admit that my proposal would have been an improvemnt over the status quo, right?
Well, perhaps not. I can't help wondering whether the NGC* program isn't a way, perhaps inadvertently, to correct the original problem, although it is being described as something a little different than a measure of technical quality. Nevertheless, maybe that approach, which incorpoates eye appeal into the mix, is even a better alternative than the pure technical option offered here. Unfortunately, one can presume that given that NGC introduced this innovation first, PCGS will have a hard time following suite and thereby implicitly admitting that NGC beat it to the punch.
Let me demonstrate using mint state 1881-s Morgan dollars currently graded MS65 and better as an example. Suppose you had a bag of such coins, randomly put together from other bags. You could expect to find the following distribution of coins by grade:
Grade Number
65 786
66 188
67 24
68 2
----------
1000
Imagine instead that a somewhat different grading scale had been adopted. I suggest that one such better scale could have been created by moving the top 20% of the MS64's into MS65, the top 30% of MS65's become MS66, the top 40% of MS66's become MS67's, and the top 50% of the MS68's become MS69's. Following that adjustment, the distribution of gem common date Morgan dollars in a bag would look something like this:
Grade Number
65 644
66 276
67 71
68 8
69 1
----------
1000
I would argue that this scale, and others like it, provide much more helpful information about the relative technical quality of a coin compared to the current truncated grading scale. But, so what you may ask? It's too late to do anything about it, even in the unlikely event that PCGS would admit that my proposal would have been an improvemnt over the status quo, right?
Well, perhaps not. I can't help wondering whether the NGC* program isn't a way, perhaps inadvertently, to correct the original problem, although it is being described as something a little different than a measure of technical quality. Nevertheless, maybe that approach, which incorpoates eye appeal into the mix, is even a better alternative than the pure technical option offered here. Unfortunately, one can presume that given that NGC introduced this innovation first, PCGS will have a hard time following suite and thereby implicitly admitting that NGC beat it to the punch.
0
Comments
That's my opinion too.
2 Cam-Slams!
1 Russ POTD!
ie, a better distribution for your bag of 1000 is
60 150
61 500
62 150
63 100
64 60
65 25
66 10
67 4
68 1
69 0
K S
2 Cam-Slams!
1 Russ POTD!
Think of all the coins that would have to be redone at tremendous expense to the collector and immense profit to the grading companies and.........um..........oh oh........
Grade Frequency Probability
60-----------120------------------0.120
61-----------195-------------------0.195
62-----------275-------------------0.275
63-----------195-------------------0.195
64-----------120-------------------0.120
65-----------70-------------------0.070
66-----------15.90----------------0.01590
67-----------6--------------------0.006
68-----------2-------------------0.002
69-----------1-------------------0.001
70-----------0.10----------------0.0001
However, assuming 1,000 coins per bag, whether dollars, or something else, I assume there will be at least one MS70 coin per every 10 bags. Even that may be too high.
C'mon.
That's like scaling a test in school. If everyone flunks, they flunk. You shouldn't skew the numbers after the fact.
Sure, grading is going to be argued from now until eternity. Could it be possible that the 68's, 9's and 70's just don't exist in the real world? That the bulk of the coins fall as they have?
I do think this is a very good topic.
LSCC#1864
Ebay Stuff
In school, the scales are skewed to the right, not left.
Speaking of everyone in class flunking, I had a class with three other students. Four of us total from start till end. On the first exam, we all flunked. I got the highest "F". I was very proud.
Anyway, he ended up giving our last two exams as take home exams, one week each.
The final exam had one question. My answer turned out to be 16 pages, give or take a sentence or two.
My final grade was a "C+". Highest grade in the class was a "B-".
The class was Petrology.
Anyway.......back to our regularly scheduled thread.
<< <i>You mean like gymnastics where on a 10 point scale all the scores are bunched between 9.7 and 10? Why not use all of the scale? >>
It only looks like all the scores are between 9.7 and 10 because that's what they show on TV. When you're watching a televised event, and especially the Olympics, you're seeing the best of the best. In the entire scope of gymnastics, there's plenty of lower scores out there.
I'm guessing you would have to burn down the gym or fly into the crowd and injure spectators in order to get a 1.
2 Cam-Slams!
1 Russ POTD!
Remember the full page adds pcgs took out in coin world when they first got started ? It would be great if someone could find and post one of those old adds here. Our grading standards will be the same tommorrow and 20 years from now or something like that. I feel sort of like that promise has already been partially broken with the new designations every once in a while that weren't listed on the holders back in the 80's. I think the full horn buffalo nickle is inevitable and I am suprised it hasn't already happened and I do feel that designation for full strike should be considered as part of the grade, so in that sense things are still changing.
Les
Doggedly collecting coins of the Central American Republic.
Visit the Society of US Pattern Collectors at USPatterns.com.
You are correct. I forgot about commems?
Actually, the scale is fine the way it is. I was just "correcting" DorkKarl's original scale.
I have no problems with the way PCGS grades, unless I get a really sucky grade. Then of course, I have to go back and look at it when it comes back and then find the problem and then kick myself for sending it in.
That one called SGS even has taken it to the extreme of 95%+ of their MS coins grade MS70
From what you describe, it sounds like what NGC is doing
As a disclaimer - all grading services occasionally grade too high or too low - some much more often than others
The Ludlow Brilliant Collection (1938-64)
<< <i>But, if we start grading on a "curve" or statistical probabilities, you are opening a huge can of worms that will lack easy answers down the road ... each coin is unique and deserves to be graded as such instead of filling a perceived slot in a mathematical model ... >>
Not only this, but this scale presupposes a fixed population with all pieces known. Not possible. Preservation, hoards, doctors, etc. can ruin any scale. Who is to say what each coin year should bring? The curve for an 1880-S morgan would sure look different from a 1921-S morgan. Or so we suppose. Just not possible to know enough to declare how the distribution should appear.
roadrunner
gents, a coin is what it is, and there are very few MS68 and 69 Morgans not because of quotas, but because these big heavy coins are RARE in those grades, they are from before the days of coins coming from the mint in plastic capsules.
I think the whole premise of grading being rigged is flawed.
Liberty: Parent of Science & Industry
<< <i>Actually, the scale is fine the way it is. I was just "correcting" DorkKarl's original scale >>
actually rampage, i like your distribution better.
<< <i>gents, a coin is what it is, and there are very few MS68 and 69 Morgans not because of quotas, but because these big heavy coins are RARE in those grades, they are from before the days of coins coming from the mint in plastic capsules. >>
the point your missing, is that "choice" should mean "above average". if that's true, how can the CHOICE grade also BE the average grade, ie the grade most of a particular issue shows up in? the average grade should be ms-62, w/ ms-61 worse than average, and ms-60 a lot worse than average (but not a crappy slider), and ms-63 really should mean CHOICE as in ABOVE AVERAGE. whatever the average grade is, the choice grade should be more difficult to obtain. it's stupid when the population of a morgan dollar in ms-63 outnumbers the same issue in ms-62 by a ratio of 4-1.
K S
<< <i>the point your missing, is that "choice" should mean "above average". if that's true, how can the CHOICE grade also BE the average grade, ie the grade most of a particular issue shows up in? the average grade should be ms-62, w/ ms-61 worse than average, and ms-60 a lot worse than average (but not a crappy slider), and ms-63 really should mean CHOICE as in ABOVE AVERAGE. whatever the average grade is, the choice grade should be more difficult to obtain. it's stupid when the population of a morgan dollar in ms-63 outnumbers the same issue in ms-62 by a ratio of 4-1. >>
Choice or better coins are submitted more often and therefore there is a dropoff at lower MS grades.
I aggree with the "A COIN IS THE GRADE IT IS" crowd. All different coins have different POP situations for all kinds of different reasons. It does seem that the services change standards often by tightening and loosening here and there as they see fit. Many here would say it's "Market Driven" whatever that means.
Live Long and Prospect.