Options
Statistical analysis of 1909 VDB RD grades between PCGS and NGC
MikeInFL
Posts: 10,188 ✭✭✭✭
Everyone says that NGC is looser in their grading. A recent purchase (and a post on another forum) got me to thinking about how much looser NGC is than PCGS.
For your review, I submit the following statstical analysis of grades of a coin I recently purchased, a 1909 VDB Lincoln (mine happens to be a MS 66):
Making the assumption that PCGS and NGC both get the same distribution of MS coins to grade and the same percentage of resubmissions (this may be a big assumption), then the following can be said:
For the coin in question, an MS 66, NGC graded 28.232% of the coins MS 66, whereas PCGS only graded 21.163% of their coins MS 66.
Again, assuming NGC and PCGS get an even distribution of coins, NGC is looser on this coin in particular.
Overall, on average, NGC grades 0.184 points higher. Again, on average NGC is looser.
It is interesting to note that for the next grade higher, MS 67, that PCGS actually grades a higher percentage of their coins in this grade than NGC. I am at a loss to explain this anamoly. My only WAG is that crackouts are to blame here or PCGS gets a higher percentage of these coins because of the price delta in coins graded the same.
Thoughts?
For your review, I submit the following statstical analysis of grades of a coin I recently purchased, a 1909 VDB Lincoln (mine happens to be a MS 66):
Making the assumption that PCGS and NGC both get the same distribution of MS coins to grade and the same percentage of resubmissions (this may be a big assumption), then the following can be said:
For the coin in question, an MS 66, NGC graded 28.232% of the coins MS 66, whereas PCGS only graded 21.163% of their coins MS 66.
Again, assuming NGC and PCGS get an even distribution of coins, NGC is looser on this coin in particular.
Overall, on average, NGC grades 0.184 points higher. Again, on average NGC is looser.
It is interesting to note that for the next grade higher, MS 67, that PCGS actually grades a higher percentage of their coins in this grade than NGC. I am at a loss to explain this anamoly. My only WAG is that crackouts are to blame here or PCGS gets a higher percentage of these coins because of the price delta in coins graded the same.
Thoughts?
Collector of Large Cents, US Type, and modern pocket change.
0
Comments
Therefore, people send MORE highend coins to PCGS for grading.
So, I would think that PCGS sees more of the highend coins and grades from that.
I've been told I tolerate fools poorly...that may explain things if I have a problem with you. Current ebay items - Nothing at the moment
<< <i>People can sell a highend PCGS coin for more $$$ than the same grade NGC coin.
Therefore, people send MORE highend coins to PCGS for grading.
So, I would think that PCGS sees more of the highend coins and grades from that. >>
Interesting point. So if I follow you logic, you feel that the delta is actually larger than 0.184 points. Interesting....
I've been told I tolerate fools poorly...that may explain things if I have a problem with you. Current ebay items - Nothing at the moment
<< <i>People can sell a highend PCGS coin for more $$$ than the same grade NGC coin. Therefore, people send MORE highend coins to PCGS for grading. >>
Not necessarily. If submitters think NGC is looser, they might just as easily send their high end coins to NGC in the hopes of getting higher grades.
<< <i>Making the assumption that PCGS and NGC both get the same distribution of MS coins to grade and the same percentage of resubmissions (this may be a big assumption), then the following can be said... >>
I have seen other posters make such an assumption. However, while I might be in the small minority here, I think that such an assumption cannot be made fairly.
The only way to get (anywhere near) meaningful results and comparisons would be to send the identical group of coins to both PCGS and NGC. Even then, the results would only tell us how each company graded such coins at that particular time.
<< <i>If you have a highend 1909 VDB RD (say, MS66 or MS67) where are YOU going to send it if you want to resell it for the most money? >>
Bochiman,
If it were a lock grade, I'd send it to PCGS. However, if the coin were a liner 66/67 and I didn't think it would go into PCGS plastic, then I'd send it to NGC -- which is kind of the point that Mark makes in his post.
Take care...Mike
<< <i>I have seen other posters make such an assumption. However, while I might be in the small minority here, I think that such an assumption cannot be made fairly. >>
Mark,
Why not?
<< <i>The only way to get (anywhere near) meaningful results and comparisons would be to send the identical group of coins to both PCGS and NGC. Even then, the results would only tell us how each company graded such coins at that particular time. >>
Agreed. To your knowledge, has anyone tried such an experiment? I'd be very interested in the results....Mike
<< <i>A .184 delta is not statistically significant, given the volume and time frame (so says the math major). >>
Paigowjohnny,
Given the sample size, over 11,000 coins, I would say the sample size is more than sufficient (so says the engineering/computer science major).
The comment about timeframe is, perhaps, a valid one. Since these coins used in these population reports have been generated over a long period of time (how long?), they don't say much about today's grading. However they do say something about the coins you are likely to find in holders today (on average).
Thanks for the comments...Mike
<< <i>If it were a lock grade, I'd send it to PCGS. However, if the coin were a liner 66/67 and I didn't think it would go into PCGS plastic, then I'd send it to NGC -- which is kind of the point that Mark makes in his post.
Take care...Mike >>
I remember using this broken philosophy. Sent the coin to NGC and it came back a grade lower. Sent it back to PCGS and it was graded as before.
I think you make a huge statistical error in assuming anything, much less that both companies get the same number of submissions/resubmissions.
keep thinking that way will ya....im not going to harp here to much but go check out some of my other posts....and do some reading
food for thought
1) tpg are ranking coins...t or f
2) upgraders are just as good if not better than tpg s are...t or f
3) tuffer ( when wrong ) means more resubmnissions...t or f
4) when a tpg gets more submissions income goes up ...t or f
5) when company income goes up usually the share price goes up...t or f
6) when company income goes up thus net income... stock options go " in the money"...t or f
7) when stcok options go in the money they get exercised...t or f
8) when options get exercised executives make big cash hits...see clct insider stock sales...someone exercise 70 mil thats big money..t or f
9) when pcgs tightens(aka gets greedy) upgraders go across the street for the right grade..t or f
10) when the upgrader`s monster coins go across the street the said monster coins are now in the othwer guys holders..t or f
11) when collectors have their heads up their rear and dont realize this and thus dont buy coins that are across the street their collections are much less than those that do...t or f
12) then in time ( aka gradeflation ) the coins across the street come back to pcgs and the process starts all over again at the next grade up...t or f
13) in a perfect world .......usually ( barring a coin coming out of the woodwork ) the best 20 years ago are the best today only in higher grades and often go across the street to stay ahead of the pack when the pop gets crowded with lesser coins this side of the street...aka gradeflation..t or f
14) this phenomena of having a " glass ceiling " to break through each grade at pcgs makes for dozens and dozens of submissions for the same coins at each grade level...t or f
15) if you buy the coin not the plastic you can build a great collection ..t or f
16) if you only buy one tpg coins your collections will be less than it could or should be....t or f
which by the way i always laff when the guys of yester year collection`s grades sink lower and lower in the registries and todays guys say...geez my collection is better than so and so...yeah right
conclusion
if you answered f to any of these points you need to do new thinking ...or languish with the other clones ...there the ones with their eyes closed
monsterman
ps...now let her rip guys...im all ears !!!
but i do speak the truth!!!!!
out of rockets ...out of bullets...switching to harsh language
<< <i>If you assume that the coins being sent to each service are a relative sample for the over all population, then they are statisically different. >>
Dbldie55,
I'm a bit confused. If the coins were each evenly distributed across MS grades (my assumption), the they should be statistically the same by the law of large numbers (law of averages)....Mike
Mark,
Why not?
Mike, there are three possibilities regarding the quality/distribution of a given coin/issue being submitted to PCGS vs. NGC: 1) PCGS and NGC each receive the same quality/distribution of coins; 2) NGC receives the better quality/distribution; 3) PCGS receives the better quality/distribution.
With three different possibilities involved, simply assuming any one of them to be the case, is bucking the odds, unsubstantiated and therefore unreliable. I'm generally a "show me/prove it and then I'll believe it" it sort of person. Maybe I'm still feeling the lingering effects of law school from long ago.
<< <i>mike
keep thinking that way will ya....im not going to harp here to much but go check out some of my other posts....and do some reading
food for thought
...
conclusion
if you answered f to any of these points you need to do new thinking ...or languish with the other clones ...there the ones with their eyes closed
monsterman
ps...now let her rip guys...im all ears !!!
but i do speak the truth!!!!! >>
Monsterman,
A very insightful post. Interestingly enough, I answered "t" to all your quesitons. What do I win?
The point I take away from this is that gradeflation, or more properly, the delta in grades between NGC and PCGS at any one time, would skew the populations as the better coins gravitated to the looser company.
That being said, assuming these deltas flip flop over time, wouldn't this effect be averaged out of the numbers, or at worst skewed to the latest TPG with the looser grading?
You are forgeting migration. There will be more of them "entombed" in NGC holders that don't make the grade at PCGS. If it does make the grade, well it's going to be in a PCGS holder.
I did a similar analysis a couple of years ago with Matte Proofs, showing more in PCGS holders. However, my experience of years of looking at many, many matte proofs is that grade for grade -- they are NOT the same. On average, you will not find similar quality between the services in the same grade, and in most cases it won't even be close.
I can't say this for any series other than Lincolns, becuase that is where most of my experience is.
<< <i><< I have seen other posters make such an assumption. However, while I might be in the small minority here, I think that such an assumption cannot be made fairly. >>
Mark,
Why not?
Mike, there are three possibilities regarding the quality/distribution of a given coin/issue being submitted to PCGS vs. NGC: 1) PCGS and NGC each receive the same quality/distribution of coins; 2) NGC receives the better quality/distribution; 3) PCGS receives the better quality/distribution.
With three different possibilities involved, simply assuming any one of them to be the case, is bucking the odds, unsubstantiated and therefore unreliable. I'm generally a "show me/prove it and then I'll believe it" it sort of person. Maybe I'm still feeling the lingering effects of law school from long ago.
Fair point, Mark. However, as I'm sure you realized, I had to make the assumption (even distribution of coins) in order to do any analysis, for to take your #2 or #3 option would skew the numbers and invalidate any result -- unless I could quantify the skewing, which is practically impossible.
I guess I'm back to the drawing board....
A very insightful post. Interestingly enough, I answered "t" to all your quesitons. What do I win? >>>>
ouch your toast.....move the deltas to the stock market where you will have a better chance getting enron`ed....cpn`ed...worldcom`ed...and the like
>>>>The point I take away from this is that gradeflation, or more properly, the delta in grades between NGC and PCGS at any one time, would skew the populations as the better coins gravitated to the looser company.
you keep saying looser.....LISTEN UP.... UPGRADERS ( AKA MARKET SUPPLER ) WATCH WHAT COINS THAT PCGS IS PUTTING INTO PLASTIC WITH A CERTAIN GRADE....WHEN THEY SEE ONE SIGNIFICANTLY BETTER THAN THAT IN THE SAME OR LOWER GRADE THEY TAKE A SHOT....IF IT COMES BACK AND ITS NOT WHAT THEY THINK IT SHOULD BE THEY TRY IT AGAIN AND AGAIN ...IF THERE IS NO SOAP AND THEY THINK ITS AT THE TOP OF THE HEAP OR SOMEWHERE IN THE BALLPARK THEY GO ACROSS THE STREET...thus if pcgs has their window shut and want to raise submissions and increase the bottomline for wall street the upgraders go across the street where it should get a fairer shake....thus if the coin happens to be indeed the best of it`s series and date when sent to pcgs and they poo poo it...off goes the best of the best into an ngc holder....where smart people buy it....dont worry though...we all know the clones will pay the big bucks...so in time...all the clones will get their shot to pay ther big bucks..........sounds kinda stupid to me but the clones do it all the time....
comprende!!
monsterman
out of rockets ...out of bullets...switching to harsh language
Looking for alot of crap.
>>>in order to do any analysis,
youd be better served if you got a grip on what im telling you as it is the truth
on the other hand
the most important analsis you need to make is where does the coin your looking at, want or have fit in relation to the rest of it`s population....as the best a worth the most....and the 2nd best are worth the 2nd most and so on.....keep in mind take the theoretical approach here......coin # 57 is probably worth the same as coin # 67...especially if there are only 50 guys collecting it
monsterman
out of rockets ...out of bullets...switching to harsh language
I follow your logic now. Thank you for the explanation.
What you are saying is that it is important to understand where a coins stacks up to its competition and how this can skew the populations of top-pop coins (or those very near to top-pop).
I do find some fault in your logic when talking about coins that are not among the highest in any given grade, for instance my MS 66 RD which spurned this disucssion, but I understand and agree with your logic for the top coins of any issue.
At the end of the day, the old adage "buy the coin not the holder" is still very much the overarching principle all of us should follow. My analysis was not intended to refute this, but rather to quantify the average "looseness" of NGC versus PCGS...Mike
Personal Ebay Auctions
My Website
---------------------------------------------
ALWAYS LOOKING TO PURCHASE
TOP 100 MORGANS / HOT 50
TOP 50 PEACE VAMS
In honor of the memory of Cpl. Michael E. Thompson
<< <i>I think you would be better off generating a p-value between the populations than simply comparing the weighted average. Even if you don't want to do the statistics for a p-value, you can obtain the standard deviation of each population and then see if they overlap. It is my opinion, based simply on the very small difference in weighted average over the sample size, that the standard deviations will overlap and produce a difference that is not statistically significant. Could you do this for us to prove your point? I am likely not as well versed with statistics as you are since I am not an engineer, I am a scientist but did quite a bit of statistical analysis for my PhD thesis. Thanks. >>
The p-value is 0, they are statistically different.
In honor of the memory of Cpl. Michael E. Thompson