Some interesting observations - an unwritten rule?

[Tradedollarnut]

Went thru the NGC Registry with a fine tooth comb looking for what criteria they used to determine the best sets. For the classic sets, the numbers that stood out were Total Set Points and Points per Coin. For the top 25 or so highest Total Set Points sets [leaders in their category only], I calculated Points per Coin. All 5 award winners came from the top 14 in this group - a pretty high correlation. I then did a little more digging and calculated the percentage of NGC coins for these sets. Below I list the top 15 Points per Coin sets [from the top 25 Total Set Point sets only]:

 

Set #1)has 8594 average Points per Coin and is 11% NGC

#2)8279 ;100% NGC ***

#3)8277 ; 19% NGC

#4)7129 ;100% NGC ***

#5)5877 ; 35% NGC

#6)5680 ; 29% NGC

#7)5418 ; 21% NGC

#8)5220 ;94% NGC ***

#9)4858 ; Closed set ***

#10)4524 ; 21% NGC

#11)4403 ; 0% NGC

#12)3383 ; 48% NGC

#13)3312 ; 30% NGC

#14)3070 ; 56% NGC ***

#15)2966 ; 9% NGC

 

Bingo! Direct hit - go right down the list and throw out any set not 50% or more NGC and the winners stand out immediately [winners shown as ***]. One winner has a closed set, but I'd be willing to bet a PF70DCAM 1963 cent [ ] that his set is greater than 50% NGC.

 

So, apparently those who wanted PCGS only [or majority] sets ineligible for NGC Top Set awards were definitely thinking the same as NGC!

[jtryka]

TDN, my head hurts so could you summarize that into one sentence for me? Are you saying that the winning sets were all majority NGC sets? Is there a majority PCGS set that should have won an award but didn't?

[Tradedollarnut]

I believe that the above criteria is exactly how the sets were chosen - there's [edited to: high odds against] a 100% correlation by chance. So it appears that all the sets without 3 stars above were bypassed for a top set award because of their percentage of NGC coins. If so, that's fine. After all, it's NGC's Registry - but how about letting us all in on the secret.

 

Bob Moreno [#14 above] has a very nice Buffalo Nickel set. But is it really in the same class as Cardinal's Bust Dollar set [#3 above]? No. But Cardinal only has 19% of his coins in an NGC holder while Moreno has 56% of his in the white plastic.

 

I thought that the coins meant more than the plastic here, but it's pretty apparent that without 50% or more of your set in the right holder.......

 

Disappointing.

[Wihlborg]

I thought that the coins meant more than the plastic here, but it's pretty apparent that without 50% or more of your set in the right holder.......

 

Disappointing.

 

 

TDN,

 

I agree that the coins should be the most important thing, but even If you are correct that one of the criteria NGC used to determine the top sets was that the sets needed to contain 50% or more of NGC graded coins, lets not forget that at least NGC allows PCGS coins to compete in their registry. At PCGS you are not allowed to add any NGC graded coins to compete in their registry, let alone 49%.

 

IMO it is PCGS's registry that is disappointing.

 

John

 

[jtryka]

TDN,

 

Correlation does not equal causation. There may have been other criteria used to select the winners as well, but right now it is all speculation. The NGC site is rediculously slow for me even though I am on a T1 line otherwise I would check for myself, but could you tell us if this trend holds for the other winners? Particularly the modern sets, as almost certainly there should be a strong PCGS presence in those sets.

If the trend holds then I would agree with your assertion, and NGC.

[Tradedollarnut]

I ain't treadin' in modern territory!

 

I always considered the NGC Registry a slightly better registry, a more encompassing registry. The scoring was more accurate, and the ability to include more coins was a huge advantage in presenting the best sets in the world.

 

PCGS Set Registry awards are repeatedly bashed because the field is so narrow - sets composed of 100% PCGS coins. The NGC award should have meant something more - instead it appears it's limited to sets 50% or more NGC. Quite frankly, that's less sets to pull from than those that are 100% PCGS.

 

If it's a rule, fine. Publish it so we know.

 

 

[jtryka]

Well, I guess I don't care too much since I have virtually no chance for an award, but I would think that these are so subjective that it would be almost impossible to have a complete list of rules. After all, how do you pick out the prettiest girl at the dance?

[jtryka]

Ok, modern set winners:

 

C SOTO JEFFERSON COLLECTION: 40% NGC, 60% PCGS

Eric Pacheco Cameo Cent Collection: 100% NGC

Everest Collection of Proof Franklin Halves 79% NGC, 21% PCGS

The Houston Collection Of Washington Quarters: 80% NGC, 20% PCGS

The Renard Collection: 98% NGC, 2% PGCS

 

So it seems like it might be somewhat suspect, but clearly not a defined rule but perhaps a bias.

[TomB]

TDN, this is a fascinating thread and it inspired me to do my own homework on the results posted.

 

I followed your parameters, as I understood the way you described them, and then added some of my own. The rules I ended up following, prior to any analysis or data acquisition, are listed-

 

1)Include only classic sets

2)Look at those sets that have the highest point totals for their class

3)Exclude sets with less than 100,000 total points

4)Do not analyze sets that are clear extensions of other sets, such as Early Dollars with Varieties vs Early Dollars or Roosevelt Dimes 1946-2003 vs Roosevelt Dimes 1946-1964

5)Disqualify sets that are closed

6)Do not consider sets that have, in my opinion and without any quantification, too many holes present

 

This led to 35 classic sets that were first in their class with at least 100,000 points. Of these, four sets were removed because they were closed. From the remaining 31 sets there were three removed because they were extensions of other sets. The last 28 sets had another three removed because there seemed to be too many holes present. This leaves 25 sets to consider. My top 15 sets are-

 

41.67% 10777.3

11.11% 8594.1

100.00% 8279.0 winner

100.00% 7129.2 winner

35.00% 5876.6

100.00% 5691.8

29.55% 5680.9

38.10% 5560.3

18.42% 5418.7

60.00% 4629.0

20.55% 4524.3

0.00% 4454.6

75.00% 4196.6

57.14% 3987.0

3.70% 3858.0

 

The columns are %NGC and points per coin. The winning sets are noted. One of the winning sets was a closed set and so was disqualified by me on that basis. The other two winners are in positions 18 and 20, on my list, as far as points per coin are concerned. We are in disagreement as far as some of the numbers here are concerned but that is likely due to miscounting of coins on my part as I did this quickly from the computer screen. Also, I might have counted coins that were only for show purpose and were not used in the calculation of point totals. This is a trivial matter.

 

The way these numbers came out I find it hard to ignore the Early Dollar, Trade Dollar, $3 Gold, Seated Dollar and Barber Half Dollar sets. I mention these five sets because they scored so high relative to points per coin and also because they struck me as being particularly tough to finish in any high quality, meaningful way.

 

So, a long way to get around to saying that I agree with your conclusion that NGC essentially required a 50% NGC set in order to win in this competition. I don't believe this is a bad thing, but it might have been nice to let people know it beforehand.

 

One last thing, you can have a 100% correlation by chance, it would just be harder and harder to come by with the more points you need to correlate.

 

[Tradedollarnut]

I guess it actually makes sense they'd go that route to encourage collectors to buy NGC coins, but it was a surprise to me since they didn't announce it. The thing that made me look is when they were pointing out the NGC * coins in each set - made me wonder if only sets that had a * coin in them were eligible. Turned out to be something else......

 

 

[Fairlaneman]

Dirty Rotten Bias NGC people anyway. Geez...I'm really PO'ed about this development.

 

 

and to think I promoted their Registry on the other forum in the beginning just to get some people to enter the darn thing. Now I'm double PO'ed.

 

 

 

 

 

 

[jtryka]

I am really surprised at the conclusions that we are jumping to, especially folks like TomB and TDN who I know think quite logically. I have already shown that there was one modern set award winner that was 60% PCGS, and that one example disproves the idea that any award winner must have more than 50% NCG coins. All we have shown here is that there is a definite NGC bias among award winners, but I still don't think we've adequately proved what the actual "rule" is. Perhaps someone from NGC can explain this to us, since right now all we are doing is engaging in speculation (not that it isn't fun, just not expecially productive).

[TomB]

I'm not "jumping" at any conclusions, I simply went through all of the data supplied for the classic sets and then analyzed it on a custom-built excel spreadsheet. The results are what I reported and a logical conclusion based on those results is consistent with an intent on the part of NGC to have a majority of NGC coins in their winning sets within the classic series. Please note, depending on what you want to show, you might certainly be able to make other logical arguements, though these would likely be more qualitative than quantitative and that is why I initially stayed away from them.

 

From memory, I can only recall that two of the winning sets were in Buffalo nickels and in classic commems. The Buffalo nickel set looked like it was of superb overall quality and is from the quintesential American series. The classic commem set I have actually seen displayed at the ANA and it is awesome and imposing. I have no problems with either set being a winner even if they didn't score as highly as some other sets using the completely arbitrary parameters that were listed above. Looking at some of the sets that did not win, however, was also mighty impressive. The Seated dollar and MS Barber half sets appeared to me to be the most impressive of the group and I couldn't help but think that I would likely have chosen them to be included. Perhaps they were the last finalists that were cut? I don't know. Since I don't remember what other sets were winners I really don't know if NGC was perhaps trying to take a mixture of metals, denominations and eras for the winning sets. This could very well be the case. They may also have been trying to keep the winning sets within certain monetary outlay groups. I don't know. If you look through the analysis closely you will see that there are a number of high percentage NGC sets that did not win and that these sets are higher in the arbitrary points per coin category than those that did win. Obviously, NGC had a reason for rewarding the collections that they rewarded, but given the numbers in front of me, I think that at least a small amount of their reasoning dealt with the certification status of some of the coins. In truth, I don't think that this is a bad idea at all. After all, how might it look if one or more of the winning sets had virtually no NGC coins in them? Certain parties could attempt to take a gain from that.

[jtryka]

Tom,

 

Correct me if I am wrong, but it seems like the point of this thread is that NGC had an "unwritten rule" that winning sets had to be majority NGC coins. Well, as you have shown, that situation holds for classic sets, but as my data showed, it does not hold true for moderns. What about the best presented sets? What is the "rule" there? My only point is that I do not see a rule, written or unwritted, or at least not one being followed consistently, so the best I can conclude it that there certainly seems to be an NGC bias, which you would expect given that it's their registry. Again, I would love some comments from the NGC folks themselves on this.

[TomB]

Jeff, I really don't care about what the point of the thread is as far as what other people determine, I care about what the point of my posts are.

 

My posts are limited to, as I have stated before, looking at what we consider classic sets and do not discount your observations as to modern or special presentation sets. In truth, I have looked at neither and never said that I had. Keeping that caveat in mind, I still think it logical to believe that NGC might have had percentage makeup of a set as at least one parameter for their decisions. Again, I don't believe this to be inherently wrong, it's just an interpretation of the available (read: examined) data.

[Tradedollarnut]

I am really surprised at the conclusions that we are jumping to, especially folks like TomB and TDN who I know think quite logically.

 

If it walks like a duck and quacks like a duck....

 

 

Politics of the coin grading services are strange and mysterious things......

[jtryka]

I give up, yeah, NGC had this unwritten rule that only sets with more than 50% NGC coins could win an award. Man that's just horrible.

[Tradedollarnut]

Don't know if you can honestly say it's horrible, so I assume you were being sarcastic. As a rule, it might make sense for NGC. But it should have been relayed openly to the participants ahead of time. And, no disrespect intended, but the pool of eligible sets is pretty narrow...

[EVillageProwler]

Assuming all this is true, if I were NGC I would probably have done something similar and I would not have advertised it in advance. In fact, if I were NGC, I would definately NOT willingly comment publicly about this.

 

It is my opinion that it is impossible for NGC to comment on this issue without stirring up more trouble. Unlike David Hall who likes to rumble with the masses, NGC seem far more careful (and reasonable) in how they deal with the masses.

 

EVP

 

[jtryka]

Don't know if you can honestly say it's horrible, so I assume you were being sarcastic. As a rule, it might make sense for NGC. But it should have been relayed openly to the participants ahead of time. And, no disrespect intended, but the pool of eligible sets is pretty narrow...

 

Yes, my entire post was sarcastic. My point is that there is at least one award winning set that did NOT have a majority of NGC coins. In fact, that nickel set had A MAJORITY OF PCGS COINS. So, I cannot logically conclude that there is an unwritten rule requiring a majority of NGC coins. The problem here is that they only picked 15 winners out of a significantly larger number of catagories, rather than just giving an award to the #1 set in every category. Once you introduce subjectivity into the process then anything goes. If we wanted to, we could problably come up with a number of "unwritten rules" if we only examine the 5 classic winners. One unwritten rule might be that awards for classic series exclude all current designs!

 

Anyway, I am still tired of arguing about this. I don't have a winning set, I don't have a competitive set, so why the heck should I care that a few elite sets filled with PCGS coins might have been snubbed? I would ask you the same question as well, why do you care?

[EVillageProwler]

I would ask you the same question as well, why do you care?

 

Perhaps because he has a great set that did not win? (Or, two?) Nor did I. (sniff, sniff. j/k)

 

I do think, however, that the whole process was waaayyy too subjective for anyone to meaningfully conclude much from a numerical analysis. Perhaps that was NGC's point? Or, am I being too paranoid?

 

EVP

 

[Tradedollarnut]

Doesn't much matter attal...

 

Just as well - Laura doesn't want me playing the registry game anyway.

[gmarguli]

Don't forget to factor in the population weights (# of coins certified by each service) as well as the percentage of coins certified by these two services not currently listed in the registry.

 

I think this is a job for Earl Warren!

[Peace_Fan]

One other data point, FWIW, is that 3 of the 5 sets awarded best presentation contained more than 50% PCGS coins (I'm too lazy to figure out the exact numbers)....