Set Registry Rewards the Better Set???

tradedollarnut · June 12, 2002 2:01AM

Spooly got me thinking by claiming that the PCGS Set Registry is superior because it rewards completion. So I did a little calculating. Which is the better set from the below? Which is more complete? Which is rated higher?

Morgan Dollar Set One

An AU58 1884-S (CU price guide $325) and an AU58 1886-O (CU price guide $105) - total of two circulated fairly common coins.

Morgan Dollar Set Two

14 glittering MS67 examples - 1879-S,80-S, 81-S, 82-S, 83-O, 84-O, 85, 85-O, 86, 87, 98-O, 1900-O, 02-O and 04-O

Now in my mind, there is no doubt that even ignoring the high value of the common date MS67 coins that EVERY collector would agree that Set Two is the better set, the more complete set, the set you'd want to have. BUT, Set One is the higher rated set by the PCGS Set Registry.

So much for rewarding completeness!!

CarlWohlforth · June 12, 2002 2:13AM

Well I don't know Morgans. But I do know Dragon does and he says the PCGS Morgan weights are all wrong.

Is there another set where this happens?

DAM · June 12, 2002 2:19AM

tradedollarnut,

I've seen this discussion for months, and it seems there is never a final solution to the question. Since I wasn't a part of those threads I'd like to put in my DAM Two Cents worth.

First: I believe it's important to allow collectors to display their "works in progress". I've got a couple myself, a 3CS set currently 17% complete and a couple of Type Sets in various stages of completion. Using my 3CS set as an example, if I were the only one listed I shouldn't be considered as having the finest known set. But that's what's happening. I wonder... are the sets PCGS is going to recongize at the end of June going to be "complete" sets. Or will they include "works in progress"?

Second: Using your example, if those were the only coins listed that would be an argument for not displaying % complete numbers, and ranking sets, until the set reaches a specific level of completion.

DAM

tradedollarnut · June 12, 2002 2:25AM

Merc Dime Set One

An AU58 1919-D (CU price guide $82) and an AU58 1924-S (CU price guide $102) - total of two circulated fairly common coins.

Merc Dime Set Two

12 glittering MS67 examples - 1916, 35,36,37,39-D,41-D,41-S,42-D,43-D,44-D,44-S,45-D

Gerry · June 12, 2002 2:42AM

TDN:

Is the underlying point to your anomolous result, in brief, that you think set ratings should reflect value (in some sense) and that the only way to achieve that is to have weights that vary with the grade of an issue, and not merely have the same weight for all grades in an issue?

If so, I remember you making that point on the message board before - maybe even a couple of times. If that is your point, I agree with you; however, David Hall and PCGS are not prepared to build a system on that basis.

CarlWohlforth · June 12, 2002 2:53AM

Are those sets out there? While the first two Mercs have more weight than the weights of the second set's coins totaled I think the grades are important too. If the second set was all AU I think the first set would be in front of it. As you've stated it I don't know but suspect the second set would rank 1'st. But someone could figure it out by doing the math. Perhaps someone has a spreadsheet set up or something?

Wouldn't the second set cost more? It would then rank higher over at NGC.

My main thing is why are we trying to compare a rare date to a common date? Neither of those sets (hypothetical?) are complete. Only when both are complete can you compare them coin for coin.

Personally I never liked any weights at all. I know, I know, I was in a very small minority. But to me why would weights matter if both sets were complete? After all if the guy with the super high grade rare date coin has a AU 45-D and really wanted to be in first he could always cough up $125 or whatever to get one in 67 FB. So I figure just use the simplest way possible for partial sets, a simple average.

tradedollarnut · June 12, 2002 3:33AM

The result is not anomolous. Every weighted series can have a result such as this. You can double, or even triple, the size of the comparative sets and still have the same result.

I have no point other than to question Spooly's assertion that PCGS is "completion friendly". It's obvious that a much more complete and beautiful set, in every weighted series, can be outpointed by a smaller circulated, conditionally rare set. So please don't make blanket statements about the advantages of any set registry over another.

Gerry · June 12, 2002 4:06AM

TDN:

I thought your point was a different one, but now that I see what you are getting at, let me try this:

The weights of the coins you put in set 1 are 9 and 8, respectively. This adds to 17. The weights of the 14 coins in your set 2 are all 1, so the total is 14. If you measure a completion as the percentage of the weights in the set, which is how it really should be measured in the PCGS rating system, then set 1 is more "complete" than set 2, and that is why it is the higher rated in the PCGS registry.

Spooly's point is valid in a different way than you interpreted it. PCGS does put a premium on completion in its registry ratings system, if you measure completion on the basis of the total of weights in a set. You would have to have a very contrived example to find a case where one set had the higher percentage of the weights, yet was rated lower.

tradedollarnut · June 12, 2002 4:22AM

The bottom line is that both Registries can have one or two coin "wonder sets" that perhaps don't deserve their ranking - NGC on the high end and PCGS on the low end. In both situations, a set with many more coins can be outpointed by a set with fewer coins. Neither Registry is any better or worse than the other in this instance.

psxchelly · June 12, 2002 5:29AM

This is a valid topic for all sets....i asked a question a few days ago regarding "tied" sets... specifically the Modern Type Proof set 1950-pres

here is an example just off the top of my head..( these sets may not exist.)

Set A is ranked 4th, Weighted GPA is 69.90, 100% complete, 93% DCAM, Set Rating is 69.90
Set B will tie for 4th, Weighted GPA is 69.90, 100% complete, 97% DCAM, Set Rating is 69.90

Now, my point being, is that I do not think its right that "the set registered first (in this case set a, which i was told is PCGS's policy-i could be wrong) for goodness sakes is the "better" set. In my opinion, since with most modern coins, the "cam" and "dcam" designation is important. Set B should be ranked higher because of the 97% dcam. I may be totally off base with this, but as a modern coin collector, DCAM and CAM designations to my collection are important.

LincolnCentMan · June 12, 2002 5:59AM

PCGS list them in order of when they were registerd in case of a tie. This order has nothing to do with the ranking, just the order they're listed. With the format of the table they use to list the sets, it is not practical to list tied set's side by side... and even if you did, there would be one to the left of the other.... presumably the first spot.

If you go to the Lincoln cent biz strikes 1909-1958, there are several sets that are tided at close to the bottom. There are three sets ranked at #75. There's no set listed at 76 or 77. The order at which they registered their sets affects only the order they're displayed, not their ranking.

...on a seprate note, I can see you're point about the CAM vrs DCAM. That would be a matter of programing the script to set the DCAM as the secondary ranking factor... in the event of a tie in the numeric standing.

David

bigt · June 12, 2002 6:09AM

Morgan weights are a complete JOKE

Spooly · June 12, 2002 9:08AM

First of all I never said "PCGS Set Registry is superior because it rewards completion"

Secondly you can run all the numbers you want, pick and choose numbers all day long to come to whatever conclusion you want. That applies to everything in the world of numbers. My opinions are based on the Registry methods and structures.

My points is this: PCGS system rewards completion.

If you take a One million dollar coin a put in the NGC Registry. (Just one coin)
Then place a complete set (Say worth 650,000)

The complete set will rank number 2! Period!

NCG system takes the Highest "value" of all the sets............ makes it the number one set!

Then the number 2, 3, 4 sets are Ranked by a percent of "value" to the number one set!

Number 1 set is always 100% ( doesn't matter how many coins are in it )( As long as it's has the highest "worth")

Number 2 set is....... 80% of the value of the Number One set!

Number 3 set is .......70% of the value of the Number One set!

ect..... ect

How where does the NGC system reward completion??? Where? It doesn't! (That is ok with me, I have stated before I like the idea of having two different systems)

The PCGS system levels out the heavy weighting issue as the set advances to completion.

It's all about the curve.... PCGS has one towards set completion. NGC doesn't! Period!!

----------------------------------------------------------------------------------------------------------
Post from Arch (Web Architect at NGC) (He is the answer man at NGC about the Registry)

"Completeness (currently) doesn't matter.

Each coin is worth X points. Each set is worth the sum of the X's.

A filler coin cannot hurt, it can only help - in terms of points. TDN's observation, though is worth noting. If you buy the filler and then upgrade, you're ultimately stuck with a filler that you have to try and resell (perhaps successfully, perhaps not). So the only place a filler can hurt you IS in the pocket book."

Arch

(Note X Points are really X dollars)

-------------------------------------------------------------------------------------------------------

As Gerry pointed out........................ You are not hearing or understanding my points .... I am sorry if you can't understand the over all principles of the two systems.

I am surpised you would try to ambush me with contrived examples........... I am talking "big picture" methods and structure.

Good Day!

(edited to fix typo)

merz2 · June 12, 2002 9:34AM

TDN
I have to agree with Spooly on this issue.There are definitely pros and cons to both NGC and PCGS Registries.One (NGC) gives you value based,while the (PCGS) gives you a better idea of how your set stands up to other sets.IMHO

DIMEMAN · June 12, 2002 10:22AM

You all have good points. I personnaly like the PCGS method better than NGC. The cost should not determine rank over completeness. One really expensive coin should not be able to BUY you first place over someone who has a nice complete set.

The Morgan sets mentioned have to be way down on the list and it will all come out in the wash when the sets are complete.

Also a 84-S is not common. But should not count more than 7 common coins in 67. That's crazy.

Just my thoughts.

Jon

oreville · June 12, 2002 10:37AM

Now I remember why I don't like registry set collecting. It is all about numbers and lets face it folks , I do numbers all day for a living, see too many numbers already when collecting raw or slabbed coins (numerical grades), watch and play sports which has numerical scores for determining how well you or your team won (or lost), then have to look at my children's report cards and lecture them on the low numerical scores and then when I see registry set weighted averages being batted around, well that drives me bonkers!

Arrrrrgggggghhhh.

grip · June 12, 2002 10:53AM

Spooly,
Well presented and i agree with you.I like both systems you have a choice.I also like PCGS it gives everyone a chance.Al

Fairlaneman · June 12, 2002 11:04AM

OK...OK !!!... I will Post to this ONE MORE TIME !!!

Both Registries have Flaws and Everyone knows that. People that take the Registries in a Dead Serious aspect certainly have Brain Damage or Some Other Mental Problem.

I believe both NGC and PCGS started the Registries for Our Enjoyment and Not as a Vehicle to Debate or Fight over.

If You like NGC Go There. If You like PCGS Stay Here. If You like Both Just Accept the Way Each is Set Up.

Ken

LincolnCentMan · June 12, 2002 11:04AM

There will always be contriversy reguarding the weighting system. And it doesnt matter how you change the wieghts, you'll always be able to "tweak" the numbers to make a point.

The problem with weighting is that there are some coins that are very difficult in grades above XF. Others are tough in grades above 64. You'll always be able to pick out a tough date, pick the grade right below the big price jump, and set it against a bunch of high dollar common dates and make TDN's point.

I will say I was total opposed to the wieghtings to begin with. However, everyone knows what the weights are. Everyone knows where the "rules" lay. If your ranking is the most important thing to you in you're collecting, you should bye those keys and semi-keys in the grade right below the big price jump. If you'd rather have something that's pretty, then collect the common dates in high grades.

The wieghts do not discriminate against collectors. Collectors discriminate againt weights.

In the end, no matter how you look at it. The PCGS set registry does encorage competition. You need only look at what top grade coins are realizing to see this.

David

Spooly · June 12, 2002 11:18AM

Fairlaneman that is my postion too! If collectors ask questions about the methods and structures of the the two systems I tell them. That way the collector can choose which one works best for them. Debate is healthy.... as long as it's not taken to a personal level.

Fairlaneman · June 12, 2002 11:25AM

tradedollarnut · June 12, 2002 11:27AM

I have nothing against either registry, nor Spooly. I just took his post on the NGC board differently - it appeared he was stating that NGC's Registry was inferior because it did not reward completion like PCGS's. Well, it's apparent that PCGS has problems rewarding completion, just like NGC - it just depends on where you look!

The myth that a "super coin" will buy you number one on NGC is false - NGC restricts all classic rarities from competing and also limits the value any one coin can add to a set. It appears to me far more likely that the problem would occur on a PCGS registry.

I enjoy both registries and participate in both - despite the apparent drawbacks and limitations of each.

dragon · June 12, 2002 1:39PM

Wow!! I'm glad to see someone else other than me point out just how grossly inaccurate and ill conceived the Morgan weightings really are. I started a thread regarding this a while back, and have emailed D. Hall about it citing specific examples of inferior sets ranked far higher than others. His reply was that PCGS is aware that the weightings are not "perfect" but they are fairly accurate for the sets that are 100% complete.......well considering that less than 7% of all the registered Morgan CS sets are 100% complete, that's not very impressive and the other 93% are way out of whack.

Dragon

lava · June 12, 2002 2:39PM

My thoguhts:
1. Tradedollarnut gives an excellent example of a flaw using morgans -- set two full of ms67s is far better than two rare date coins in not-so-rare au condition. One solution -- PCGS should offer bonus points for premium grades, regardless of date. A morgan ms67 anything is deserving of a bonus point, just as, IMHO, any ms66pl or ms66dmpl. After that, this concept of bonus points must become a little more selective. For some years maybe a ms65, or ms64pl or ms64dmpl deserves a bonus.
2. PCGS' system of weighting is a nice idea that needs some tinkering. In the morgan set, no bonus in the circulated set for a pl or dmpl. I know that has many of us scratching our heads. I note also that NGC is not perfect either. NGC's new pl set does not give credit for dmpls, because apparently dmpls only get credited in the dmpl set.
3. Are all set meant to be complete? Could Bill Gates complete a morgan dmpl set? I think there are probably about ten coins he will never find, another several that supposedly exist and that are locked away in collections, and another handfull that for all but Bill Gates are probably outside the budget.
4. I personally do not care for the PCGS solution to the completion problem, which is to make excpetions. In the pl registry, only 50% of the coins need to be pl. Since 80% of the coins could be obtained in pl, why not make it 80%?
5. In my informal poll posted here re: 66pl v. 65dmpl, we were split, half opting for the 66 grade, half opting for the deeper mirror in 65. The weights are across the board higher for dmpls. There is no right or wrong answer I guess.
6. I like PCGS because I feel more like I am competing against myself. I also like it because it encourages to buy PCGS coins, and the market seems more willing to pay a premium for PCGS coins than other slabs. I think the weights need to be adjusted however. I also like seeing the pop numbers when viewing the coins.
7. I like NGC because it accepts PCGS and NGC slabs, so I don't feel compelled to play the crossover game (although my ANACS coins are of no help unless crossed). What I don't like about NGC is the fact that the measuring stick is completely arbitrary. One guy's collection might be stink, but it might be number one. Gates' collection might put the second collection, which by any other standard might be awesome, and make it look bad, like at the low end of the 0-20% line. I like seeing the population breakdown, but you have to click a button to see it. Hopefully NGC will get wisdom and soon give dmpls credit in their pl registry.

Spooly · June 12, 2002 2:57PM

I am not debating the weights of each coin, just the total structure of the Registry.

100 total Morgan sets on the PCGS Registry.

The top 18 sets are 98 to 100% complete.

As you go up the Ranking scale the weighting curve rewards completeness. That is the upward trend.

(Sure it bounces around a bit as you go DOWN the ranking because some sets just have high grade keys.... which rank higher than than some sets with more coins (that is the nature of weighting).....but that flattens out the Higher you go up the scale towards completeness.)

TDN used fictional sets to prove a point. You have a 118 coin set with each coin weighted differently! TND own examples make up a 1% to 5% sample SLANTED in his favor. Mathematically it's a silly statement.

tradedollarnut · June 12, 2002 3:10PM

Spooly - that doesn't make sense. You call very specific examples "slanted" and "mathmatically silly", yet your total observation is based on the fact that "as you go up the set ratings that the sets are more complete"??? What kind of mathmatical observation is that?? The system is made so that when a set owner adds a coin, his set rating goes up. So...the sets at the top are more complete.....astounding mathmatical observation!

BTW - most examples chosen to prove a point are "slanted" - otherwise, what's the point in using them as an example? That doesn't make them "mathmatically silly".

If you want to observe that, in general, the sets at the top of the Registry, any registry, are the most complete, I would have to agree with you - that's generally the point of the Registry concept! When you have 100 sets in a denomination, this will almost certainly be the case.

Fairlaneman · June 12, 2002 3:26PM

TDN:

Why did You Change that Post too Spooly ??

Anyway TDN is Correct on His Statement about the Completeness factor. I really thought He was Out To Lunch on this but He Is Not. Go To the Basic Mercury Dime Sets and Look at Sets 2...3...4. Then Go down and Look at Sets Compare 1 and Compare 2. You will see that TDN is Correct about PCGS Not Rewarding Completeness. It Appears to be Done on Weighting Only.

So Sorry TDN.

Ken

tradedollarnut · June 12, 2002 3:41PM

I changed it because I'm trying to avoid being overly confrontational - that it not my intention at all. It just seems that others, like you, have the mistaken belief that there is some magical "completeness bonus" built into the PCGS Set Registry. It just isn't true. It's just a total of weights and grades.

Looks like I have my "real life" example now! Thanks, Fairlaneman.

Spooly · June 12, 2002 5:13PM

Check the NGC board..... Arch just proved my point!

ARCH-----
" There is no direct bonus for completeness. There is only weighting by slot. There are a couple of IMPLIED completeness bonuses - they're not REAL, they're just a side effect."

"SO, here is the true "completeness bonus" that PCGS gives... ready?

If you're getting spanked by someone with a GREAT coin in a heavy slot in a very hard to get grade... just go buy the coin in a low grade for cheap. You'll get almost as many points, and if your set is more complete than their is in the REST of the slot, you'll win."

Thanks for proving my point! I never said it was a direct bonus......... IT'S A CURVE AFFECT OF THE WEIGHTING STRUCTURE!

If you take a hand full of low weight coin like TDN says and compare it to a heavy weighted coin.... of coarse the scale it not right. But the weighting is figured on a TOTAL WEIGHTED POINTS...........So as you add more coins you get a leveling curve effect!

1. Adding a coin
2. Playing the curve

Dame how many times do I have to say it!

NGC system is base on value! So you could deleted 3 coins worth $1000 total....... then add 1 coin worth $1000 in value and the set stays the same(Value Rank).... but you have losted 2 coins.... "The set is Less complete in the number of coins" !

Fairlaneman · June 12, 2002 10:06PM

TDN and Spooly:

Heck this Dense Person that is Typing this NOW See's both Points each of You are Trying to make.

I think I will go out and Buy the last 5 Dimes I need for the set seems how the Weight is the Same for a AG as it is for a MS67. Naw can't do it as that would be Throwing Good Money towards Bad Money.

This has been a Nice Little Debate that has Spanned the USA from California to Flordia.

Ken

Gerry · June 12, 2002 11:58PM

Ken summarizes this whole debate in a neat, concise way. You might go buy those last few dimes in any low grade in which you can get them if your goal is to push your set up to a higher ranking. That would be a far, far better way to achieve that goal than buying one of those coins in an ultra-expensive high grade but leaving your set incomplete. This is just a reflection of the fact that the PCGS system rewards completeness. But, his other good point is "do you really want to do that?" I guess that's a matter of individual choice: to some completeness is very, very important - to others it is not. Good Job, Ken!

Spooly · June 13, 2002 12:44AM

The point of buying "cheap" keys only works on the low to middle of the completness scale. It will burn you as you move from the middle to the top! Once you get closer to 100% complete the "effects" of the curve are "very small" in the top sets.

grip · June 13, 2002 1:58AM

This is one of the better factual post's i have read.Thanks .Al

scher · June 13, 2002 4:55AM

interesting topic..collect the best examples you can find/afford, collect what you like and let the chips fall where they may..
all that ranking stuff sure gets gets ya goin though and thats what pcgs wants...good marketing at work to be sure.
bruce scher

Set Registry Rewards the Better Set???

Comments