Do you think each date and mint in a set should have different weights in each grade?

lclugza

Do you think each date and mint in a set should have different weights in each grade, based on that date and mint's rarity in that grade? The 1901 silver dollar is given a higher weight in the current set than the 1885CC, however, the 1885CC is a much rarer coin overall. Would it be better to have a set where the 1885CC is given a higher weight in circulated grades, and the 1901 a higher weight in "Mint State" grades?

Flash

Yes I do. I would be quite a monumental task for somebody at PCGS to tackle, but at least they'd have something to spend all that money on that I've sent them!




Matt

tompkins

On the surface, this seems like it might help, but, if you think about it, I would have to say most emphatically "NO!" for a number of reasons. first (and most obvious) is the huge amount of work in defining the weights, and the ensuing arguments and discussions. But, equally important is some of the mathematives. let me posit an example (and I don't think this is too far fetched, albeit a very short "series").

Suppose there were a set of silver Ike MS coins (71-S - 76-S) And I'm using these coins to show what I think is a realistic example, albeit a very abbreviated set). Suppose for the 2nd highest grade (I'll ignore the MS69's for this illustration) we use the existing PCGS weights of 3, 1, 1, 1, and 2; and for the highest grade we arbitrarily say everything is a weight of 5; and everything below the second highest grade has a weight of 1. So what you have is the following weights:
71-S: 1 (MS65), 3 (MS66), 5 (MS67)
72-S, 73-S and 74-S: 1 (MS65 through MS67); 5 (MS68)
76-S: 1 (MS65, MS66); 2 (MS67); 3(MS68)

Now, suppose your set consists of an MS65 (71-S and 4 MS68's---the weighted grade works out to 67.86
You upgrade your 71 to a 66: because of the higher weight, the grade GOES DOWN to 67.74
You upgrade again to a 67: the grade rises to 67.8, but it is still lower than when you started with the MS65.

Sure, you can tweak the weights to solve this situation, but no matter what you do situations like this will arise somewhere.

So it sounds nice, but I just don't believe it would work because of problems like this, unless the distribution of the grades is constant for all coins in a series (which it clearly is not).

Pete

lclugza

I think you'd have to throw out the familiar Sheldon numbering system completely for it to work, and you'd might have a system where a MS65 1893S dollar was worth 100 points vs. a MS common date 5, and the 1901 worth 2 points in EF versus 40 in MS63, etc. This might work, but it would also be complicated and confusing. Perhaps a better system is to rank each date and mint on overall rarity (instead of just MS rarity, as it seems to be now).