Experts needed on Diamond Proportions.

[Texas Leaguer]

Garry,
I have a question about this comment above:
"HCA is near enough but it is not good enough, and if I were to add the minor facets, which is covered in my patent, then you have the GIA system."
Why have you chosen not to add the minor facets? (Not that it would be a simple undertaking but you have available the advanced expertise.) It seems that it would make the tool much more robust. As you mention, it would position it as an alternative to GIA facetware. And because it would not reward steep/deeps and it would give credit to other makes that get dinged unfairly in the GIA system, it seems like it would make the tool even more useful as an "advanced filter" on the GIA system.
Good question Bryan, by the time I got to that point, and started a patent war with GIA that cost serious 6 figures, I had realised through my re-education from Sergey that it was a dead end pursuit. It would have taken 100's of hours (before DiamCalc Pro) to do the minor facet computations (which were relatively easy for the table crown and pavilion stuff) and why? By then I knew that unless you could use this approach fr fancy shapes it was a mikey mouse approach. But yes, I could do it inside a week now with DCpro, but it still is a dead end.

A couple of follow-up questions out of curiosity- from the HCA results page is this mention of updates:
"HCA scores were adjusted Dec. 15, 2001 and Feb. 6, 2003" Leonid and I reworked some data glitches.
This would suggest that the tables have not been tweaked since 2003. Is that correct? It seemed to me that I saw a poll you did a year or so ago about fire vs brightness and you mentioned that you were considering re-weighting them in the model. I have added symmetry bonus and penalty to the Android App which is available from play store. The Apple app - latest from very slow developer 10 minutes ago "The app works in my development environment and on my phone, but when I put into production/the App Store it breaks. This is the last bug I am facing before it is complete.. the email from Apple is me trying to test it in the app store."

Also on that page is this statement:
"Even though HCA grades cut more effectively than systems like the AGS, it does not yet factor in symmetry and minor facets."
Was this statement made prior to the release of the AGS LP grading system? Or do you still think HCA grades cut more effectively than the AGS system?

Yeah!!! very very old. where is it - send it to me and Andrey and we will fix it!!!!
Remember when AGS used a really dumb approach with shallow stones in one corner, deep in the other, and nice thru the diagonal middle. First time I met Pete I told him in my brutal aussie manner why they were wrong and how wrong they were. We sat down with DC and he got it. Check it with him

Here's the page to be updated:
http://www.pricescope.com/tools/hca (hit the go button to get to the results page below)


Re: adding minor facets to HCA - if you can do it in a week I personally think it would be VERY worthwhile. Of course it is extremely easy for me to spend your 40 hours! (and maybe you meant a week for your team).

 

[John P]

Re: adding minor facets to HCA - if you can do it in a week I personally think it would be VERY worthwhile. Of course it is extremely easy for me to spend your 40 hours! (and maybe you meant a week for your team).

I think so too. While improving the tool you might also make it more apparent that the given numeric result is less important than the broad classification, meaning "HCA EX" candidates all deserve equal consideration. Right now the first impression is that 0.8 is better than 1.2 is better than 1.8 (etc.) Your usage notes disclaim this Garry, but not everyone gets it.

 

[Karl_K]

Re: adding minor facets to HCA - if you can do it in a week I personally think it would be VERY worthwhile. Of course it is extremely easy for me to spend your 40 hours! (and maybe you meant a week for your team).

I think so too. While improving the tool you might also make it more apparent that the given numeric result is less important than the broad classification, meaning "HCA EX" candidates all deserve equal consideration. Right now the first impression is that 0.8 is better than 1.2 is better than 1.8 (etc.) Your usage notes disclaim this Garry, but not everyone gets it.

With GIAs horrible rounding of the lowers it puts on the grading report there will be a lot of edge cases where it could go either way based on the gia numbers. How would you recommend handling them?
I see it on here all the time that the only possible comment is that might be a nice diamond within the gia rounding and that's all that can be said without images.

 

[John P]

With GIAs horrible rounding of the lowers it puts on the grading report there will be a lot of edge cases where it could go either way based on the gia numbers. How would you recommend handling them?
I see it on here all the time that the only possible comment is that might be a nice diamond within the gia rounding and that's all that can be said without images.

I feel exactly the same way. Especially with 75% given LH, and the difference between 73-77 is critical to know (80% is a close second).

That said, I don't think the HCA should stagnate based on what's out of its control. Cup-half-full: Cut-focused PS sellers take their own local scans (I know, I know that was redundant as soon as I said "cut focused" ). Also, however, a lot of suppliers - especially Indians - are doing 3D scans and have more specific information internally.

 

[Karl_K]

With GIAs horrible rounding of the lowers it puts on the grading report there will be a lot of edge cases where it could go either way based on the gia numbers. How would you recommend handling them?
I see it on here all the time that the only possible comment is that might be a nice diamond within the gia rounding and that's all that can be said without images.

I feel exactly the same way. Especially with 75% given LH, and the difference between 73-77 is critical to know (80% is a close second).

That said, I don't think the HCA should stagnate based on what's out of its control. Cup-half-full: Cut-focused PS sellers take their own local scans (I know, I know that was redundant as soon as I said "cut focused" ). Also, however, a lot of suppliers - especially Indians - are doing 3D scans and have more specific information internally.

I look at the hca as answering if the crown and pavilion angles work well together to sort diamonds for further review.
I am not sure adding lgf really changes that given the data most commonly available.
On the other hand it would eliminate a group of false positives but how large a group is it?
Is there enough information out there in an easy enough form to answer that?

But in the end I don't think it would change how I look at the hca or use it.