Author Topic: Varieties of silver Historical Figures, 1984 to 1993  (Read 126040 times)

0 Members and 2 Guests are viewing this topic.

Offline badon

  • Trade Count: (0)
  • Hero Member
  • *****
  • Posts: 4487
  • Karma: -81
Re: Varieties of silver Historical Figures, 1984 to 1993
« Reply #195 on: December 10, 2011, 09:02:12 PM »
By the way, I think I can make it so you won't have to upload and/or mark an image for the type. The randomness feature will work to put those images from the specimen on the type page, automatically. I like things that are automatic. That's ultimate user-friendliness :)

This worked perfectly. Thanks for the brilliant feedback DiggingNorway. It works so well I was able to remove the "Add a main image" link from all pages that no longer need them. That's much more user friendly. Now, the only time anyone needs to mark an image as "main" is if there isn't already one available from lower levels. Since there's nothing lower than specimens or sometimes sightings, the images will automatically be displayed from wherever they were uploaded. It's a beautiful thing! Now, reloading the page shows you different coins each time.

I think it makes the pages much more interesting - except for the reverses, because they're all the same!

Offline SANDAC

  • Supporter
  • Global Moderator
  • Trade Count: (+8)
  • Hero Member
  • *****
  • Posts: 2265
  • Karma: 118
Re: Varieties of silver Historical Figures, 1984 to 1993
« Reply #196 on: December 10, 2011, 09:28:15 PM »
Could we make the random obverse/reverse smaller or otherwise designate them as temporary until a more permanent pictures are available?  I'm concerned that new user would see those spots as already taken and therefore uncomfortable with replacing them with more suitable pictures.  For an example, the best picture for CCT13 should be a picture of the 1984 4-coin set, but without a prior knowledge of how the random picture come about, a new user may be reluctant to replace the random pictures with a more suitable permanent one.

Offline SANDAC

  • Supporter
  • Global Moderator
  • Trade Count: (+8)
  • Hero Member
  • *****
  • Posts: 2265
  • Karma: 118
Re: Varieties of silver Historical Figures, 1984 to 1993
« Reply #197 on: December 10, 2011, 09:40:35 PM »
Krause put the mintage of Chen Wen (and the other 3 1987 HF) at 4000! Mr Ge puts it around 9000. Chen Wen's rather rare appearances point more to the 4000 figure.

The mintage of HF remains a mystery to me.  Some years, such as 1991/1992/1993 are difficult to find, yet their mintage is much bigger than 1989 which is plentiful on eBay.  Certain coin within a year is more readily spotted than other.  Princess Chen Wen is hard to find, but other 1987 coins are easily located.  My original rationalization was that mintage within a year is not identical, such that many of the coins are locked up in a set which circulates slower than the surplus loose coins.  Mr Ge's latest figures shot a big hole in that theory, because some of the most readily spotted coins actually have lower mintage than others in the set.  There are definitely disconnect between the published actual mintage and what are actually circulating in the market place.  Maybe selective hoarding is the answer (which is my current theory concerning the 1992 HF set), but why?

Offline badon

  • Trade Count: (0)
  • Hero Member
  • *****
  • Posts: 4487
  • Karma: -81
Re: Varieties of silver Historical Figures, 1984 to 1993
« Reply #198 on: December 10, 2011, 09:48:49 PM »
Could we make the random obverse/reverse smaller or otherwise designate them as temporary until a more permanent pictures are available?  I'm concerned that new user would see those spots as already taken and therefore uncomfortable with replacing them with more suitable pictures.  For an example, the best picture for CCT13 should be a picture of the 1984 4-coin set, but without a prior knowledge of how the random picture come about, a new user may be reluctant to replace the random pictures with a more suitable permanent one.

I agree that for a 4 coin set type, the main images should be images of the 4 coin set. But, what about a very general type like CCT10: Chinese historical figures? For those pages, I think random is best usually. What do you think?

For the pages where a permanent picture is better, I think it should be someone already familiar with the type that selects the permanent picture. I'll make the form instructions easier to understand, and it should be OK from there.

The users can still easily upload an image for the page, but not making it too easy to replace the existing images means that the user will need to become more familiar with the CC before they start replacing images. By being more familiar, it means better images will be selected, and we won't spend all day every day changing things back to the way they were every time someone wants to experiment. Does that make sense? What do you think? Should be we try it this way for a while, and maybe add refinements as we go?

I like that there's 2 less links that the user needs to figure out what they do. And, the random images all come from the main images in subtypes, so they're probably always going to be good representatives of the supertype. As for making them smaller, I could do that, I don't think it's necessary since the label underneath says "Random Obverse" and "Random Reverse" whenever the images are random.

Offline badon

  • Trade Count: (0)
  • Hero Member
  • *****
  • Posts: 4487
  • Karma: -81
Re: Varieties of silver Historical Figures, 1984 to 1993
« Reply #199 on: December 10, 2011, 09:52:52 PM »
The mintage of HF remains a mystery to me.  Some years, such as 1991/1992/1993 are difficult to find, yet their mintage is much bigger than 1989 which is plentiful on eBay.  Certain coin within a year is more readily spotted than other.  Princess Chen Wen is hard to find, but other 1987 coins are easily located.  My original rationalization was that mintage within a year is not identical, such that many of the coins are locked up in a set which circulates slower than the surplus loose coins.  Mr Ge's latest figures shot a big hole in that theory, because some of the most readily spotted coins actually have lower mintage than others in the set.  There are definitely disconnect between the published actual mintage and what are actually circulating in the market place.  Maybe selective hoarding is the answer (which is my current theory concerning the 1992 HF set), but why?

I agree, this is a huge mystery. I don't think Mr. Ge has "shot a hole" in anything yet, since we know nothing about his research. Hopefully our population statistics will make the AVAILABILITY of the coins more clear. Restrikes might explain some of it, but we'll have to do more research.

I think part of the confusion is because the historical figures series market is much more immature than other series, so the market is not a good representative of true rarity. I know a lot of the later dates came to the market in big hoards, while the earlier dates like 1989 are circulating furiously because they're supposed to be the key set and everyone is buying and selling them.

Offline SANDAC

  • Supporter
  • Global Moderator
  • Trade Count: (+8)
  • Hero Member
  • *****
  • Posts: 2265
  • Karma: 118
Re: Varieties of silver Historical Figures, 1984 to 1993
« Reply #200 on: December 10, 2011, 10:03:19 PM »
I agree that for a 4 coin set type, the main images should be images of the 4 coin set. But, what about a very general type like CCT10: Chinese historical figures? For those pages, I think random is best usually. What do you think?
Where there are many similar subtypes underneath a type (such as a variety of the type), randomly displaying the subtype is actually desirable.  It is a thing of beauty to behold.  Where subtypes are substantially different than the type, like CCT13 and all the different types below it.  Randomness seems less suitable.  It begs to be replaced with a permanent picture.

Offline badon

  • Trade Count: (0)
  • Hero Member
  • *****
  • Posts: 4487
  • Karma: -81
Re: Varieties of silver Historical Figures, 1984 to 1993
« Reply #201 on: December 10, 2011, 10:25:11 PM »
Where there are many similar subtypes underneath a type (such as a variety of the type), randomly displaying the subtype is actually desirable.  It is a thing of beauty to behold.  Where subtypes are substantially different than the type, like CCT13 and all the different types below it.  Randomness seems less suitable.  It begs to be replaced with a permanent picture.

I'm glad you agree - it is very attractive to see :) You can get introduced to fascinating coins you have never seen before, just by doing normal things on the site. I'll give it some more thought to see if I can come up with a way to make the "begging" easier by showing that link that used to be there.

The problem is (and the CC's strength) is that types aren't different from each other. You'll notice that they all have the same basic data to start out with, and the only thing that makes them different is what people do to them. In order to add a link to certain "kinds" of types, I would have to figure out a way to make them distinct. That's one more thing to keep track of if we change something. Keep in mind that there's already over 600 types on the site! If you have to check a checkbox to mark a type as the kind that needs main images, why not just add the main images?

I can't do that automatically, because it will take a human to decide which of the 600 types should "beg" for main images. That's a fair amount of work that would be better spent either doing something else, or just adding the images whenever you're ready.

The more I think about it, the more I have a hard time figuring out how it can or should be done, without requiring lots of unproductive work - which also needs to be undone once the images are finally added (so the type stops "begging" for main images). That last part could be automated, but that type of data entry is the kind that I've spent a lot of effort trying to avoid in the database design.

Remember the switch in terminology for obverse and reverse images? That was a productive part of the database design, but still took a bit of effort to do. When there's millions of images, it could take an army of users months or years to change that many images. So, adding an unimportant, temporary bit of data to each page, possibly millions of them, just so it will beg is something I really want to avoid if possible.

However, there is another approach that could work. I can generate lists of types that don't have main images yet. That way, people can add images to the ones that need them, and ignore the ones in the list that don't need them. It would work just like the lists of images marked as obverse and reverse that was on the main page before. Once the job was done, I can easily just removed the code that produced those lists (we can always add it again later if someone uploads a lot of images incorrectly).

I've spent some time working on the software for generating a list of types without main images, but I had to refocus my attention on more important things. When I have some free time, I can keep working on it, if you think that is a good idea. What do you think?

By the way, my top priority is to help people use the sight, and learn what issues they're running into, and what features they think are important. After everyone goes to bed in the West, that's when I work on the new stuff :)

The last "new stuff" I did was making it possible to show data from all subtypes, sightings, images, etc. It took some reworking of the database design, but I was able to automate most if it. The results are truly incredible. It is now easily possible to surf on top of enormous quantities of data easily, in a pleasant and informative way, without sinking and drowning in the data.

Hmm, that was a nice analogy that's actually pretty close to how the works. I should put that in the help page.


Offline badon

  • Trade Count: (0)
  • Hero Member
  • *****
  • Posts: 4487
  • Karma: -81
Re: Varieties of silver Historical Figures, 1984 to 1993
« Reply #202 on: December 11, 2011, 12:56:35 AM »
After some thought and testing, it appears we need the "beg" links on types where no main images exist, and where there are no main images in subtypes that can be used for randomization. Everything seems to be working now, but I'm going to clean up the code a bit to make it easier to modify in the future. This simple feature was a bit more complex than I thought it would be :)

Offline SANDAC

  • Supporter
  • Global Moderator
  • Trade Count: (+8)
  • Hero Member
  • *****
  • Posts: 2265
  • Karma: 118
Re: Varieties of silver Historical Figures, 1984 to 1993
« Reply #203 on: December 11, 2011, 02:00:52 AM »
After some poker around, I'm becoming a even bigger fan of the random-images-for-the-main-image approach.  As long as one populate the image at the end of the chain, it will automatically fill in all the intermediate hierarchy with relevant images.  The range of the images automatically increases as the hierachy is filled to reflect the larger population.  Genius.  It should be a recommended style to not populate the intermediate main images, just let the random image generator does its job.  At or near the highest level of hierarchy, one may want to showcase a selected group of notable coins within that hierarchy.  Is there a way to mark certain coins for displaying by the random generator?

Offline SANDAC

  • Supporter
  • Global Moderator
  • Trade Count: (+8)
  • Hero Member
  • *****
  • Posts: 2265
  • Karma: 118
Re: Varieties of silver Historical Figures, 1984 to 1993
« Reply #204 on: December 11, 2011, 02:10:04 AM »
You can send in more than 10 coins at a time just by crossing out the serial number on the top of the subsequent forms (don't cross out the first one), and then numbering your forms 1 of 2, 2 of 2, etc.

Also, if you're sending in more than one of the same coin, you can use just one line. Just put the quantity in the notes for that line. Of course, you won't be able to keep track of your NCS before and after photos as well that way, because you won't know what numbers the coins receive.
I bet not many people know that trick.  If you look up NGC certification number in the 3xxxxxx series (the number generally assigned to NCS conserved coins), it generally ends at -010.  So most people just fill out one page of the form and sent in 10 coins.

Offline badon

  • Trade Count: (0)
  • Hero Member
  • *****
  • Posts: 4487
  • Karma: -81
Re: Varieties of silver Historical Figures, 1984 to 1993
« Reply #205 on: December 11, 2011, 02:17:22 AM »
After some poker around, I'm becoming a even bigger fan of the random-images-for-the-main-image approach.  As long as one populate the image at the end of the chain, it will automatically fill in all the intermediate hierarchy with relevant images.  The range of the images automatically increases as the hierachy is filled to reflect the larger population.  Genius.  It should be a recommended style to not populate the intermediate main images, just let the random image generator does its job.  At or near the highest level of hierarchy, one may want to showcase a selected group of notable coins within that hierarchy.  Is there a way to mark certain coins for displaying by the random generator?

Right now, the random images are taken from the lower level images that are marked as main obverse and main reverse. It does not use any of the slab images, since the coins are too hard to see in those photos when they're so small. So, all you have to do is set a main obverse and main reverse image on the lower level types, and everything else is automatic!

Eventually, there may be millions of images in the Coin Compendium. Even with just a few thousand, it is becoming important to only select the best images. That's why only one main image is allowed for obverse and reverse of each type. If you mark two images as main obverse or main reverse, both images will show up, and it will look strange.

I'm sure eventually I will need to make a tool that finds all types with more than one main obverse or main reverse image, so we can easily find them and remove one of them. There have been a lot of concerns about validity of the data, but despite this database's potentially huge size, it remains pretty easy to find the problems so someone can fix them. Well, it's easy to find pretty much anything and everything :)

I'm glad you liked the random image trick. Thanks to DiggingNorway for discussing that with us, and bringing it to our attention. It will be exciting to see your coins become "famous" when they show up at the top level for everyone to see :)

Speaking of that, eventually the front page will be much nicer, and all those lists will be moved to some sort of "tools" page. When it is nicer, we can have "coin of the day" or something like that, to promote coins that have a lot of interesting information recorded about them.

Offline badon

  • Trade Count: (0)
  • Hero Member
  • *****
  • Posts: 4487
  • Karma: -81
Re: Varieties of silver Historical Figures, 1984 to 1993
« Reply #206 on: December 11, 2011, 02:19:29 AM »
I bet not many people know that trick.  If you look up NGC certification number in the 3xxxxxx series (the number generally assigned to NCS conserved coins), it generally ends at -010.  So most people just fill out one page of the form and sent in 10 coins.

Haha, if NGC got smart, they would use website forms with 1000 lines, so they will make more money :) But, they like antique carbon paper from the 1820's:

http://www.nytimes.com/1998/11/12/technology/carbon-paper-still-messy-still-in-use.html?pagewanted=all&src=pm

Offline badon

  • Trade Count: (0)
  • Hero Member
  • *****
  • Posts: 4487
  • Karma: -81
Re: Varieties of silver Historical Figures, 1984 to 1993
« Reply #207 on: December 11, 2011, 02:43:37 AM »
I just discovered that sighting data is not being fed into specimen data, so when I fix that, we'll have some more images. The sighting images are often pretty ugly though, since they come from screenshots or catalog scans. I may have to filter out screenshots from the random image display, except when no other images are available.

Offline SANDAC

  • Supporter
  • Global Moderator
  • Trade Count: (+8)
  • Hero Member
  • *****
  • Posts: 2265
  • Karma: 118
Re: Varieties of silver Historical Figures, 1984 to 1993
« Reply #208 on: December 11, 2011, 02:29:36 PM »
I was tracking down a lead on NGC certification site and ran across this interesting one.

http://www.ngccoin.com/certlookup/CertResults.aspx?CertNumber=3306026-019

There are two things interesting about it.  1.  This is another example of the "super puffy sleeve" Li Chun.  2.  NGC slabbed all 1987 4-coin set in a "super slab", so if anyone has such super slab, please take a picture of it and put it as the main picture of CCT16.

Offline badon

  • Trade Count: (0)
  • Hero Member
  • *****
  • Posts: 4487
  • Karma: -81
Re: Varieties of silver Historical Figures, 1984 to 1993
« Reply #209 on: December 11, 2011, 03:10:59 PM »
Interesting research. That "super slab" has all 4 of these certification numbers:

3306026-017
3306026-018
3306026-019
3306026-020

So, each coin would be entered into the CC as individual specimens, with a slab photo that shows all 4 coins. Once entered, you can click on the certification number navigator (under the CC number) to find more.