Grading systems and 'gaming' them: An eternal story?

[John P]

PS is considerably different because the community's 'recommendations' are influenced by the impact on the customer, especially when it comes to the fifth C - cost. If everyone had unlimited funds, it would be great to sit here and extoll the virtues of Kobe beef, but when not all budgets will support Kobe beef, something's gotta give, right? Beyond that, there are some people who just don't GET the masses fawning over Kobe beef because they truly do prefer a nice cheeseburger.

Precisely. And some Kobe beef lovers can't understand cheeseburger POV. Thankfully, for those who enjoy both, we have Kobe beef burgers (available with foie gras topping here). And, to the point, maybe I can get ten people excited by that - but if they have never tasted kobe beef and foie gras in-person we must realize our descriptions can reach a certain level, but after that it's up to their taste buds.

Since there is no way to 'legislate' or standardize preferences that will universally apply to all, it's hard to consider PS a 'grading system'. There may be times when people do indeed say "good enough", but that's often based on sensitivity to the poster's OWN stated values and priorities, and it's really not the place of the community to tell him why he should prefer something else.

To me, PS is more of a virtual neighborhood where people can solicit opinions. It makes sense that if I were to poll the 8 neighbors on my street about who their preferred landscaping providers are, their recommendations would be influenced by their buying values (cost, time to complete the job, variety of plants, professionally dressed employees, etc.) and who they happened to use. If I ask for their opinions, I do recognize their responses may weight things that aren't important to me and I'd adjust the weight of those recommendations to me accordingly. I also realize their experiences are limited to those they've used and may not include other great providers.

In many cases here it's also influenced-by what is repeated here time and again. Take the "PS-attitudes" regarding fluorescence, perception of lab-standards, "H&A" and "Ideal" themes, recommended angle ranges, etc. While I don't think this is negative (since most widely-adopted positions are reasonably defensible) it does reflect a pattern of repetition until dissenting views are lost or even trampled. In the neighborhood sense, it would be like a new neighbor moving in, saying "hey I have a landscaper named Bob with lovely white lilacs" and having the "established" folks ignore Bob completely, recommend the neighborhood-approved landscapers and say there's no reason to go with white lilacs since off-white lilacs are cheaper and no one can tell the difference anyway (apologies to flower lovers if I butchered all that)... Hey, those neighbors may be giving sound advice but it does not foster growth, it just repeats a (limited) system that has proven safe and economical.

Likewise, there are several folks on PS who prefer better clarities, better colors, and better makes, but that may not be important in the values of the poster asking.

PS mission statement wasn't to cultivate cut-nuts, though that certainly was a predictable side effect of the learning curve. It's was to create a place where people could get meaningful data to inform their buying choices, not to tell that the only right buying choice is the one Skippy would choose (apologies to Skippy for being the example study this time). In fact, most times, PSers explain *why* they each prefer something, and that will help an inquirer determine if that's a valid reason for him, too.

Well-said and agreed, with the stipulations above.

 

[Rockdiamond]

David- thanks for the reply.
For sure this highlights some of what I perceive as the reason the science behind classifying diamonds using the tools is still controversial.

"Normal Light"- a workable analogy is the way diamonds are graded for color.
We need to have standardized lighting to be able to determine the ever so slight differences between D and E.
When we look at the stones in different lighting, it may be difficult- if not impossible to see that slight difference- yet in the case of color, the difference in color is fairly undeniable.

Cut, however, is simply different. In the case of color, we're looking through the pavilion ( for colorless anyway) which eliminates the light play seen through the table.
Judging cut is done through the table- which introduces a ton of variables.
What one person sees as "brighter", may not be what another person perceives as brighter- even in standardized lighting.
We can attribute this to contrast, patterning, and other factors that go into what makes a diamond bright, shinny, scintillation, what have you.

In both color, and cut, introducing "real life" lighting creates a whole new set of variables- but again, in the case of cut, the variables are that much greater.
John- to my eye, whatever the weaknesses of the "BrillianceScope" also affect the other methods such as HCA.
ASET and IS are different in that they use physical properties and photography.
That's where I see the biggest advances coming- or at least the biggest advances that will be more universally accepted.
There's many in the trade , for example, that disagree ( vehemently) with some of the research produced by AGSL.
I don't see the changing by 2020.

For me the bottom line is that "Excellent" cut- such as Infinity's goods- can be demonstrated using the technologies in question- however other types of excellent cut styles may be arbitrarily downgraded using the same technologies.

I'm still waiting for a good example of, and explanation of a GIA EX Cut grade steep deep- and it's deficiencies. Pick a sample stone, and just use the measurements, so we are not picking on anyone's inventory.

 

[Karl_K]

It's a given that in jewelry showrooms one can expect lighting to be arranged to maximize appearance. This is logical, just as in car-dealers’ showrooms, fashion shows, etc. It is critical to overcome the single-environment for comparisons. In my own travels I've consulted with showrooms to create a "diamond-walk," where clients can take a tray or temporary settings with candidates and walk through a span of different illumination scenarios: Multiple direct spotlighting (normal in most showrooms), an indirectly lighted area (darkened or beneath a counter), diffused lighting (like a classroom or office) and natural lighting (near a window or -better- outdoors). These are some of the staples. In some cases the jewelers have needed to create space and adjust lighting to do this, which has taken some convincing at times. Pleasantly, the payoff has been gratitude from clients for creating a unique process seen as fair and educational. I urge all sellers to try it. I also urge them to encourage return-trips for such comparisons, during both the sales process and the return period.

All that is nice and better than the typical set up but the results will still have a high chance of changing if the comparison was made over time in the persons actual environment.
At best in my opinion the results are on par with selecting one that has high potential using the PS system.
Both are valid and good ways to select diamonds but I dont see a huge advantage either way.
When it comes to what is most common in stores it is no longer par but the advantage shifts to the PS online shopper who spends the time to fully vet the diamond when they get it by looking at it in for them everyday situations and lighting.

 

[Venice]

I'm still waiting for a good example of, and explanation of a GIA EX Cut grade steep deep- and it's deficiencies. Pick a sample stone, and just use the measurements, so we are not picking on anyone's inventory.[/quote]


I am curious to see this, too. From what I read here the "steep-deep" stones are made out to look quite awful in person. What GIA grade should these cuts have? Very Good? Good? Fair?

 

[WinkHPD]

Earlier in this thread I saw a point expressed that differentiation of the top 1% of diamonds was not important. (Sorry, no time to go find it and get exact quote.) I spoke with Peter Yantzer and asked him about the percentage of diamonds that potentially qualified as AGS0 light performance. This is only about light performance, as you can have AGS0 light performance and still fall short on other grading factors and end up with a lower cut grade than AGS0.

He tells me that the theoretical performance of the AGS0 light performance grade currently sits at approximately the top 7% of the scale. That is a lot of room for differentiation. Yet these diamonds often appear extremely similar in the static technologies like IS and ASET.

I have observed the above in practice and I am of the opinion that people need to see a diamond to know for sure if it is their perfect flavor, but people like Karl are unhappy with me because I offer no "scientific proof." Well, I can tell you that before 1930 nobody could possibly offer scientific proof that Pluto existed because the technology did not exist to see it. But that did not mean that it wasn't there…

More important, this isn't just about the limits of technology, it's also about perception and taste. We may NEVER be able to prove "perfect flavor" any more than we can prove vanilla is a better flavor than chocolate, or vice-versa. Even when we can qualify a diamond in that top 7% (with obvious room for differences beyond the grade and still-images) we will never be able to prove the percentage of people who like A more than B, as there will always be taste issues to be factored in.

This to me is the real issue and I think that it warrants more conversation than, “Hey, you diamonds that are more expensive - get out of the pool!”

Wink

 

[oldminer]

Wink, very interesting that AGS0 stones are at the top 7% of the pool. Why can't or why won't the AGSL give the trade their top 1% super-duper grade to work with, too? Why leave everyone guessing or uninformed? Likely as not, this super grade would be very near the center point of each zero parameter range they have set as the widest limits, but that is not anything we know for sure. It just sounds logical, but pure logic may not be the way this grading structure has been constructed. Since optics and physics are purely by defined rules of nature, it would be sensible that the very best is at the center of the current top grade range. I can't claim to know.

Part of the reason we see gaming is the wide latitude that obviously exists at 7% around a top, 1%, center point of cut grading. I bet there would be very little or much less visual difference among stones cut to the top 1% level. Could a human eye see the difference between a top 1% and a top 2%? I don't know, but probably no consumer could see a difference. Then again, very, very few consumer can see the difference between a VVS2, a VVS1 or an IF stone even with a binocular microscope let alone with their naked eyes.

If the labs had chosen to create a system for grading that could not be easily "gamed" they ought to have defined the extremely top pinnacle grade besides the widest possible range for a secondary top grade. Then folks like Paul could work toward meeting that very top pinnacle with assurance of recognition for their superior efforts and people who would prefer to game a system would work the fringes in the top 7% arena and still likely have good success. The way it exists now is that the field is wide open and those who are doing the finest work compete with little real benefit or recognition against mass producers of the lowest possible and widest interpretation of top cut grading.

When I made the AGA Cut Class system I came at the problem by defining the pinnacle first, 1A, and then broadened the grades as one came down the cut quality slopes. I saw the issue of gaming the system in the beginning, back in the mid-1980's, and knew that by making an extremely tight top grade that gaming would be impossible to play at that level. The next level, 1B, was a lot less complicated to meet although still categorized as "Ideal". This gaming is nothing new. Technology and marketing has changed the game and its rules, but the game players will always work any system to make it pay off for them. Creating a system that recognizes and rewards the very best players makes good sense to me.

 

[Serg]

Earlier in this thread I saw a point expressed that differentiation of the top 1% of diamonds was not important. (Sorry, no time to go find it and get exact quote.) I spoke with Peter Yantzer and asked him about the percentage of diamonds that potentially qualified as AGS0 light performance. This is only about light performance, as you can have AGS0 light performance and still fall short on other grading factors and end up with a lower cut grade than AGS0.

He tells me that the theoretical performance of the AGS0 light performance grade currently sits at approximately the top 7% of the scale. That is a lot of room for differentiation. Yet these diamonds often appear extremely similar in the static technologies like IS and ASET.

I have observed the above in practice and I am of the opinion that people need to see a diamond to know for sure if it is their perfect flavor, but people like Karl are unhappy with me because I offer no "scientific proof." Well, I can tell you that before 1930 nobody could possibly offer scientific proof that Pluto existed because the technology did not exist to see it. But that did not mean that it wasn't there…

More important, this isn't just about the limits of technology, it's also about perception and taste. We may NEVER be able to prove "perfect flavor" any more than we can prove vanilla is a better flavor than chocolate, or vice-versa. Even when we can qualify a diamond in that top 7% (with obvious room for differences beyond the grade and still-images) we will never be able to prove the percentage of people who like A more than B, as there will always be taste issues to be factored in.

This to me is the real issue and I think that it warrants more conversation than, “Hey, you diamonds that are more expensive - get out of the pool!”

Wink

Scientific hypothesis of existence Pluto was 30+ years before astronomers see Pluto.
Approximate coordinates of Pluto had been calculated much early than 1930. It was very important and necessary information to find Pluto

This history is very good example what science and scientific proof is.( most interesting what old calculation The Pluto coordinates was wrong, because had been used wrong information about Neptune mass. similar history was with Tolkowsky round cut . )
Firstly you need develop hypothesis ( predict some phenomena), then proof it( confirm hypothesis)
To develop good hypothesis you need develop theoretical model of phenomena .
If you have not such model, you have not real chance to develop good hypothesis and then proof it.



"In the 1840s, using Newtonian mechanics, Urbain Le Verrier predicted the position of the then-undiscovered planet Neptune after analysing perturbations in the orbit of Uranus.[20] Subsequent observations of Neptune in the late 19th century caused astronomers to speculate that Uranus' orbit was being disturbed by another planet besides Neptune. In 1906, Percival Lowell, a wealthy Bostonian who had founded the Lowell Observatory in Flagstaff, Arizona in 1894, started an extensive project in search of a possible ninth planet, which he termed "Planet X".[21] By 1909, Lowell and William H. Pickering had suggested several possible celestial coordinates for such a planet.[22] Lowell and his observatory conducted his search until his death in 1916, but to no avail. Unknown to Lowell, on March 19, 1915, his observatory had captured two faint images of Pluto, but did not recognise them for what they were.[22][23]

Due to a ten-year legal battle with Constance Lowell, Percival's widow, who attempted to wrest the observatory's million-dollar portion of his legacy for herself, the search for Planet X did not resume until 1929,[24] when its director, Vesto Melvin Slipher, summarily handed the job of locating Planet X to Clyde Tombaugh, a 23-year-old Kansas man who had just arrived at the Lowell Observatory after Slipher had been impressed by a sample of his astronomical drawings.[24]

Tombaugh's task was to systematically image the night sky in pairs of photographs taken two weeks apart, then examine each pair and determine whether any objects had shifted position. Using a machine called a blink comparator, he rapidly shifted back and forth between views of each of the plates, to create the illusion of movement of any objects that had changed position or appearance between photographs. On February 18, 1930, after nearly a year of searching, Tombaugh discovered a possible moving object on photographic plates taken on January 23 and January 29 of that year. A lesser-quality photograph taken on January 21 helped confirm the movement.[25] After the observatory obtained further confirmatory photographs, news of the discovery was telegraphed to the Harvard College Observatory on March 13, 1930.[22] http://en.wikipedia.org/wiki/Pluto "

 

[John P]

I'm still waiting for a good example of, and explanation of a GIA EX Cut grade steep deep- and it's deficiencies. Pick a sample stone, and just use the measurements, so we are not picking on anyone's inventory.

I am curious to see this, too. From what I read here the "steep-deep" stones are made out to look quite awful in person. What GIA grade should these cuts have? Very Good? Good? Fair?

"Awful" is harsh. But then again, I am a Kobe beef guy who respects the people who are happy with Sizzler. And while GIA EX does not have such a dramatic range (Kobe to Sizzler, with apologies to Sizzler) there are levels of differentiation which (to me) have less appeal than others. Primarily this has to do with a reduction in brightness, better seen in some lighting conditions than others.

#1 #7 and #9 from the study below are such examples. I don't know that someone would call them "awful" in live viewing but they do have less overall brightness (to me) compared to other examples. From memory, #7 in that study was considered least-bright. I remember that because seven is my favorite number and it was somewhat "Murphy's Law" for me that the least-favored (by our group of observers) wound up drawing that position.

https://www.pricescope.com/journal/laboratory_cut_grades_what_report_doesn’t_show

 

[Serg]

I'm still waiting for a good example of, and explanation of a GIA EX Cut grade steep deep- and it's deficiencies. Pick a sample stone, and just use the measurements, so we are not picking on anyone's inventory.

I am curious to see this, too. From what I read here the "steep-deep" stones are made out to look quite awful in person. What GIA grade should these cuts have? Very Good? Good? Fair?

"Awful" is harsh. But then again, I am a Kobe beef guy who respects the people who are happy with Sizzler. And while GIA EX does not have such a dramatic range (Kobe to Sizzler, with apologies to Sizzler) there are levels of differentiation which (to me) have less appeal than others. Primarily this has to do with a reduction in brightness, better seen in some lighting conditions than others.

#1 #7 and #9 from the study below are such examples. I don't know that someone would call them "awful" in live viewing but they do have less overall brightness (to me) compared to other examples. From memory, #7 in that study was considered least-bright. I remember that because seven is my favorite number and it was somewhat "Murphy's Law" for me that the least-favored (by our group of observers) wound up drawing that position.

https://www.pricescope.com/journal/laboratory_cut_grades_what_report_doesn’t_show

  John,

re:"#1 #7 and #9 from the study below are such examples. I don't know that someone would call them "awful" in live viewing but they do have less overall brightness (to me) compared to other examples. From memory, #7 in that study was considered least-bright."

How is about Fire for samples 1,7,9? Did you compare Fire?


Even on "MONO " photos with light environment for grading Brightness, the stone 7 have more big and colorful Fire flashes than most others diamonds from this test .

"Mono" Photos do not give right comparison even for Brightness. ( you need compare these diamonds in Stereo, then you have rights to say each diamond is more bright for Human)

 

[John P]

How is about Fire for samples 1,7,9? Did you compare Fire?

Even on "MONO " photos with light environment for grading Brightness, the stone 7 have more big and colorful Fire flashes than most others diamonds from this test .[/b]

"Mono" Photos do not give right comparison even for Brightness. ( you need compare these diamonds in Stereo, then you have rights to say each diamond is more bright for Human)

The comparisons were made live Sergey. Not sure about fire. It was two years ago but I do have notes somewhere (if I can locate them). I remember #7 simply because of the mnemonic device mentioned.

 

[Serg]

How is about Fire for samples 1,7,9? Did you compare Fire?

Even on "MONO " photos with light environment for grading Brightness, the stone 7 have more big and colorful Fire flashes than most others diamonds from this test .[/b]

"Mono" Photos do not give right comparison even for Brightness. ( you need compare these diamonds in Stereo, then you have rights to say each diamond is more bright for Human)

The comparisons were made live Sergey. Not sure about fire. It was two years ago but I do have notes somewhere (if I can locate them). I remember #7 simply because of the mnemonic device mentioned.

John,

Cut Light Performance is not just Brightness, but many ASG, IS, ASET Biases do not see important difference between Light Performance and Brightness. Many pro consumers become blind when they see "death ring", leakage on diamond Photo.

 

[WinkHPD]

<Snip>

This history is very good example what science and scientific proof is.( most interesting what old calculation The Pluto coordinates was wrong, because had been used wrong information about Neptune mass. similar history was with Tolkowsky round cut . )
Firstly you need develop hypothesis ( predict some phenomena), then proof it( confirm hypothesis)
To develop good hypothesis you need develop theoretical model of phenomena .
If you have not such model, you have not real chance to develop good hypothesis and then proof it. <Snip>

Hey! Pluto was more on-topic than I ever intended! Thank you for the history, Serg.

I am not a scientist but the proof in my world is in the history of consumer reactions. One day maybe we will have the tool or computer program to explain them. We do not at present.

For now I hope you'll forgive me though, if I don't go analyzing perturbations in "Uranus!!!"

Wink

 

[Venice]

I'm still waiting for a good example of, and explanation of a GIA EX Cut grade steep deep- and it's deficiencies. Pick a sample stone, and just use the measurements, so we are not picking on anyone's inventory.

I am curious to see this, too. From what I read here the "steep-deep" stones are made out to look quite awful in person. What GIA grade should these cuts have? Very Good? Good? Fair?

"Awful" is harsh. But then again, I am a Kobe beef guy who respects the people who are happy with Sizzler. And while GIA EX does not have such a dramatic range (Kobe to Sizzler, with apologies to Sizzler) there are levels of differentiation which (to me) have less appeal than others. Primarily this has to do with a reduction in brightness, better seen in some lighting conditions than others.

#1 #7 and #9 from the study below are such examples. I don't know that someone would call them "awful" in live viewing but they do have less overall brightness (to me) compared to other examples. From memory, #7 in that study was considered least-bright. I remember that because seven is my favorite number and it was somewhat "Murphy's Law" for me that the least-favored (by our group of observers) wound up drawing that position.

https://www.pricescope.com/journal/laboratory_cut_grades_what_report_doesn’t_show

Thank you, Mr. Pollard, for taking the time to respond to my question and for sending the link. I have seen the link before, but it was nice to examine it again.

Perhaps the word "awful" was a bit harsh, but I get the notion we should avoid them like the plague. Again, maybe a bit harsh. Would you say that they would fall into the Very Good category?

Thanks again. I appreciate and value your opinion.

 

[Karl_K]

I have observed the above in practice and I am of the opinion that people need to see a diamond to know for sure if it is their perfect flavor, but people like Karl are unhappy with me because I offer no "scientific proof."

Wink

You rang?
Your often repeated and consistent theme is that in your office people preferred a certain brand of diamond.
What does that prove? That in your office some people liked one particular diamond type.
That does not prove that everyone would make the same selection.
That does not prove that if they lived with the diamonds for 2-3 days they would make the same selection.
Read what I wrote in this thread and have been saying for years. People need to live with a diamond for a few days in their world to see if it is the right one for them.
How it was selected, online by data or b&m looking at it does not change that at all.
Further when you try and apply your theme to any diamond not in the comparison it falls apart because that diamond was not compared.

 

[Serg]

<Snip>

This history is very good example what science and scientific proof is.( most interesting what old calculation The Pluto coordinates was wrong, because had been used wrong information about Neptune mass. similar history was with Tolkowsky round cut . )
Firstly you need develop hypothesis ( predict some phenomena), then proof it( confirm hypothesis)
To develop good hypothesis you need develop theoretical model of phenomena .
If you have not such model, you have not real chance to develop good hypothesis and then proof it. <Snip>

Hey! Pluto was more on-topic than I ever intended! Thank you for the history, Serg.

I am not a scientist but the proof in my world is in the history of consumer reactions. One day maybe we will have the tool or computer program to explain them. We do not at present.

For now I hope you'll forgive me though, if I don't go analyzing perturbations in "Uranus!!!"

Wink

Wink,
re:For now I hope you'll forgive me though, if I don't go analyzing perturbations in "Uranus!!!"

For diamonds You can use more simple method than elliptical orbit calculations :

"Tombaugh's task was to systematically image the night sky in pairs of photographs taken two weeks apart, then examine each pair and determine whether any objects had shifted position. Using a machine called a blink comparator, he rapidly shifted back and forth between views of each of the plates, to create the illusion of movement of any objects that had changed position or appearance between photographs. "

Something similar could be very helpful for diamond comparison . Science has very limited number of methods , these methods are universal for any type scientific work

 

[diagem]

David- how would you say "performance" is objectively measured?

I agree with Karl on this point- there is no "normal" lighting. This is part of my argument of why any light performance standards can't possibly cover all the possibilities.
But I do ask in earnest- what means are you suggesting to measure and grade light performance?

I am using the words "Normal Lighting" in the scientific way, not in the regular terminology of daily speech. Normal lighting in science means a specifically chosen, arbitrary lighting scheme which the scientist has decided to use in the measurement of some optical or physical feature of an object. This lighting should have relevance to some common, regular model of lighting when possible, and must be highly repeatable in all aspects. Lighting created with the use of calibrated LED's of known color temperature, distribution and voltage will help make a chosen lighting scheme repeatable and consistent. If such a scheme provides the right environment for meaningful collection of data and the scientist uses it on a regular basis, then that lighting scheme is said to be "Normal Lighting". I hope my statements now make a bit more sense. There are limitless lighting schemes in regular daily life.

In my 13+ years working along with the team at ImaGem, Inc, I have been exposed to digital devices which analyze the amount and character of light from known sources in their "normal lighting" scenario. These devices are based primarily on analysis of very high definition images. When one knows the exact amount of light being projected on a diamond, one can measure with great accuracy and repeatability certain characteristics of the light coming back to the observer's eye which in digital image making is the sensor in the digital camera's sensor. How much light comes back, how much potential scintillation there is, how well defined a pattern and contrast is being shown in the image. By sub-pixel analysis, these "facts" can be "measured". When these numeric, repeatable results are combined with studies of human preference, we have data which strong correlates to how pretty a diamond will look to the majority of observers. Even if the beauty analysis is imperfect in that not every observer is the same, the data is rather perfect and clearly defines performance. A higher amount of performance does not always equate to "better", but for those interested in interpretation of what numerical results mean, getting solid data on a repeatable and consistent basis is very meaningful. If a diamond were to be made with too much light return and not sufficient contrast, it would lack an element required for beauty even though the measure of light return was very high. A diamond with high sparkle and overall lower light return might well appeal to some folks, but in terms of measured performance it would be somewhat lower in light return. It is a balancing act to create a beautiful diamond and not a contest of who can make every measured aspect higher and higher at the same time. Such a diamond won't look so good to most of us.

David, are you talking about still face up LP measurements or are light/Diamond/observer moving while being measured as well?

 

[WinkHPD]

I have observed the above in practice and I am of the opinion that people need to see a diamond to know for sure if it is their perfect flavor, but people like Karl are unhappy with me because I offer no "scientific proof."

Wink

You rang?
Your often repeated and consistent theme is that in your office people preferred a certain brand of diamond.
What does that prove? That in your office some people liked one particular diamond type.
That does not prove that everyone would make the same selection.
That does not prove that if they lived with the diamonds for 2-3 days they would make the same selection.
Read what I wrote in this thread and have been saying for years. People need to live with a diamond for a few days in their world to see if it is the right one for them.
How it was selected, online by data or b&m looking at it does not change that at all.
Further when you try and apply your theme to any diamond not in the comparison it falls apart because that diamond was not compared.

Karl,

I am disappointed that you only quoted the part of my post which was about you. It is far less important than Peter Yantzer's point: While these diamonds often appear extremely similar in the static technologies there is a lot of differentiation, even at the top of the scale.

Regarding “my office.” You are trying to make this just about me but it is not. These choices happen in the offices of appraisers and other jewelers around the country all the time. And those people all have the chance to live with the diamonds for far more than 2-3 days.

Karl, it seems like you are trying to find differences in our beliefs where they do not exist. We have never said differently than each other with regard to these things. I am an online seller and think the online process is sound, but I also think Paul, whos in the business of CREATING diamonds, has a point. So do respected science professionals. We can learn something here.

And with respect, your repeated theme seems to be that you can choose better for people by-internet than people can do in real life?

I am sorry my friend but that is far off the path of reality. Far, far off. We can identify great flavors here, but can we tell them which flavor they will like?

Wink

 

[oldminer]

With ImaGem technology a diamond is rotating perpendicular to the lighting. During the rotation, the stage is stopped several times and a detailed image is taken for analysis. Multiple image analysis are used in giving the numerical end results. It is testing done in the face up position which is what most believe to be the far most important position of a diamond in use and for consistently malking repeatable measurements. This is what AGS and GIA have also encouraged in their systems of grading cut. We all know that diamonds are worn in many different light environments and often viewed at less than optimum angles, but to grade diamonds or to create repeatable measurements one has to choose a model which can work and give reuslts which will have meaning. These results cannot cover every varied position or lighting angle and they should not. Making a measuring system one must keep in mind that the system must be understandable and repeatable. Creating more variables will make most systems just about useless. To an extent already, we have parametric and ray tracing systems in place which use more variables than can be truly relied upon. Garbage in get garbage out Rounding and averaging do not make for optimal grading results. Such is the laboratory version of "gaming" to a certain extent.

 

[diagem]

With ImaGem technology a diamond is rotating perpendicular to the lighting. During the rotation, the stage is stopped several times and a detailed image is taken for analysis. Multiple image analysis are used in giving the numerical end results. It is testing done in the face up position which is what most believe to be the far most important position of a diamond in use and for consistently malking repeatable measurements. This is what AGS and GIA have also encouraged in their systems of grading cut. We all know that diamonds are worn in many different light environments and often viewed at less than optimum angles, but to grade diamonds or to create repeatable measurements one has to choose a model which can work and give reuslts which will have meaning. These results cannot cover every varied position or lighting angle and they should not. Making a measuring system one must keep in mind that the system must be understandable and repeatable. Creating more variables will make most systems just about useless. To an extent already, we have parametric and ray tracing systems in place which use more variables than can be truly relied upon. Garbage in get garbage out Rounding and averaging do not make for optimal grading results. Such is the laboratory version of "gaming" to a certain extent.

I believe advanced laboratories are at an intersection in regards to cut.
The more advanced the research the need of movement assessment becomes an important missing link in the cut grading formula.
You are correct that face up position is important, but seldom is a Diamond viewed in it's optimum position IRL.
As I have been studying light behavior in the last couple of years, I came to a conclusion that fancy cut Diamonds have the ability to look less than optimum in the face up position while enhancing their LP in various other positions specifically in motion.

The future of Fancy Cut Diamonds looks promising.

 

[Rockdiamond]

I'm still waiting for a good example of, and explanation of a GIA EX Cut grade steep deep- and it's deficiencies. Pick a sample stone, and just use the measurements, so we are not picking on anyone's inventory.

I am curious to see this, too. From what I read here the "steep-deep" stones are made out to look quite awful in person. What GIA grade should these cuts have? Very Good? Good? Fair?

"Awful" is harsh. But then again, I am a Kobe beef guy who respects the people who are happy with Sizzler. And while GIA EX does not have such a dramatic range (Kobe to Sizzler, with apologies to Sizzler) there are levels of differentiation which (to me) have less appeal than others. Primarily this has to do with a reduction in brightness, better seen in some lighting conditions than others.

#1 #7 and #9 from the study below are such examples. I don't know that someone would call them "awful" in live viewing but they do have less overall brightness (to me) compared to other examples. From memory, #7 in that study was considered least-bright. I remember that because seven is my favorite number and it was somewhat "Murphy's Law" for me that the least-favored (by our group of observers) wound up drawing that position.

https://www.pricescope.com/journal/laboratory_cut_grades_what_report_doesn’t_show

Thanks for posting that excellent link John.
Would you say it's fair to state that some observers would pick #7 based on it's attributes? I believe that a fair number of shoppers would indeed pick that type of stone due to it's lack of patterning ( doesn't show H&A) and longer LGF's..

In terms of "steep deep"- I agree that it's a bit deep- and indeed, will spread every so slightly less than most 1.16ct EX or ) cut grade stones- but ever so slightly.
#9 Steep deep has an average ( good) spread for it's weight
#1 steep deep is also a bit too deep for my taste- and also has a very slight spread disadvantage.
But based on the photos and other info, it would seem to me that it's only the very slight spread loss that is a "clear" disadvantage to two of the steep deeps shown in that article.
Appearance wise, it really comes down to a taste issue.

 

[John P]

Thank you, Mr. Pollard, for taking the time to respond to my question and for sending the link. I have seen the link before, but it was nice to examine it again.

Perhaps the word "awful" was a bit harsh, but I get the notion we should avoid them like the plague. Again, maybe a bit harsh. Would you say that they would fall into the Very Good category?

Thanks again. I appreciate and value your opinion.

You're very welcome Venice.

GIA defines all of those diamonds as part of their "Excellent" span of appeal. I did find differences, personally, but any opinion I offer is just me (and GIA is much larger than any one person), based on my own preferences and standards.

With that stipulated, I regularly find GIA EX diamonds I'd not select for myself; most-often with pavilion and crown angle pairings higher than 41.2/35.5. This is not a universal rule but it's a tendency for me. There are also some GIA VG combinations (too shallow for EX) which I like better than some EX combos. Again, not universal.

In my supplier-role I have a fun time with my dealer-family. Some principals and store gemologists love to see if they can "trick" the super-duper-ideal guy by handing over a diamond (typically an outside combo) and quizzing me about it. My track record is good enough that they keep bringing the challenges, and I think not-knowing what I'm being quizzed on keeps my saw sharpened. To the point; once in a while I'll be surprised at how lively a diamond is relative to its paper stats, or how another specimen doesn't "pop" like I would have predicted based on average numbers.

Therefore, based on the average numbers of #7 above, all I can say is that it (and a few others) did not "pop" for me in that particular field of candidates...while other diamonds in the study did.

 

[John P]

Thanks for posting that excellent link John.
Would you say it's fair to state that some observers would pick #7 based on it's attributes? I believe that a fair number of shoppers would indeed pick that type of stone due to it's lack of patterning ( doesn't show H&A) and longer LGF's.

You're welcome David.

It's fair, just as it's fair to say some observers will pick a marquise over any round. Some will pick an emerald cut. Some will pick a pearl over a diamond. So yes, I think #7 would find (and probably has) someone to love and take-home and cuddle-with.

I find that the majority of people tend to prefer the angular combos the GIA, AGSL and other systems all award their highest grade. Therefore #7 - which was not rated highest in all systems - could often lose (for the majority of people) when compared side-by-side with candidates which DID score highest in all the systems.

On-topic, all of this depends on micro-taste differences between different observers. There are definitely tendencies - but not everyone is confined to the same taste or preferences.

Q: If you took all of the diamonds in this study and gave them to 1000 people to live-with for a week who would choose what?

* Do you think all 1000 people would end-up with the same diamond? I don't.
* If we cut the field in half do you think all 1000 would end-up with the same diamond?...
* If we cut the field to two, would all 1000 pick the same diamond?...

 

[whatmeworry]

How would you cut grade a diamond with nano etching on the pavilion to increase fire? GIA doesn't give it a cut grade because it's a non traditional approach.
Seems that the direct assessment tools Brilliantscope and Imagem would be better suited here.

 

[kenny]

How would you cut grade a diamond with nano etching on the pavilion to increase fire? GIA doesn't give it a cut grade because it's a non traditional approach.
Seems that the direct assessment tools Brilliantscope and Imagem would be better suited here.

I would grade that a gimmick, not that there's anything wrong with gimmicks.
If you like it, buy it.

 

[Garry H (Cut Nut)]

How would you cut grade a diamond with nano etching on the pavilion to increase fire? GIA doesn't give it a cut grade because it's a non traditional approach.
Seems that the direct assessment tools Brilliantscope and Imagem would be better suited here.

It is a good example WMW - why all grading systems for basic light return functions should be open ended.
The OctoNus system uses tolkowsky 1ct = 1.00
This means 'say' the pav etched stones might achieve 1.50 for fire.
Brilliancescope, AGS and all other measuring systems I know can not do that. I do not think Imagem can either as it has ben presented.

It is an good example of how creativity and innovation is killed by current grading systems.

 

[Karl_K]

And with respect, your repeated theme seems to be that you can choose better for people by-internet than people can do in real life?
Wink

Not exactly, compared to any place local to me dollar for dollar they will get a much better diamond online and would pay a lot more in a store or end up with not as nice a stone.
The online edge isn't as large as it used to be but it is still there.
With very few exceptions that would be true almost anywhere.

I know of one store here who keeps a couple ags0 stones on hand, uses them to sell a ton of so-so egl graded stones.
"sure this one is better but it is $4000 more and as you can see not that much better"
If someone happens to spend the extra money the dealer still wins and orders another one.
I have seen the same thing with gia EX, use them to sell EGL and GIA VG at high margins.
HOF is often used this way.
For the price of that EGL/gia vg graded stone the person could move up to at least a decent gia EX if not AGS0/h&a/PS consumer approved online.
Many consumers here and even Sir John can confirm that goes on more than anyone who learns at PS would like.

 

[oldminer]

Imagem has no metric for fire because it was one more variable that can be measured, but one which does not set most consumer preferences. Since all well cut diamonds of any shape display some degree of fire, it was not one of the grade setting variables which appeared necessary in human testing to determine which diamond looked better....to the vast majority. All of us understand that there are a few people who might select a diamond because it has unusally high fire caused by nano-etching....

Imagem can measure fire since it can analize and count all pixels showing color and delete all body color pixels. It could create a numerical count of pixels of color other than body color or black.... Would this grade the stone? No. Could it be a metric for fire? Yes.
Imagems assessment is that measuring fire muddies the waters on cut grading. Too much information leads people to not being able to compute a decision on how to choose a stone. A smart grading system goes for the most important variables, presents the information in a way humans can interpret the results, makes the results highly repeatable and correlating to human perception.

The current AGS system seems to do the job without numerical reference to measured metrics, but I feel that providing factual numerical references taken from actual measurements is highly meaningful to consumers. Direct measurement of actual diamond performance metrics is the eventual method which will prove most reliable.

 

[Rank Amateur]

Is the available rough cut differently now due to the pressure of the grading systems or is there just more emphasis on cut "gaming"?

Or are we still talking about such a small fraction of stones that it is insignificant?

 

[oldminer]

In order to create a grading system with solid factual and repeatable results for fancy shapes one will end up choosing the face up position for making the call. Unless some organization can force a particular angle of viewing other than perpendicular to the table, logic would say that any other angle is so limited that it would not be better than perpendicular. You certainly can't be promoting using a number of different angles and mixing the results. That would be meaningless. I agree some diamonds will perform better at some random angle, but you wouldn't want to judge every other stone of that shape at that angle. To me, that would be both unscientific and rather naive.

 

[Serg]

Imagem has no metric for fire because it was one more variable that can be measured, but one which does not set most consumer preferences. Since all well cut diamonds of any shape display some degree of fire, it was not one of the grade setting variables which appeared necessary in human testing to determine which diamond looked better....to the vast majority. All of us understand that there are a few people who might select a diamond because it has unusally high fire caused by nano-etching....

Imagem can measure fire since it can analize and count all pixels showing color and delete all body color pixels. It could create a numerical count of pixels of color other than body color or black.... Would this grade the stone? No. Could it be a metric for fire? Yes.
Imagems assessment is that measuring fire muddies the waters on cut grading. Too much information leads people to not being able to compute a decision on how to choose a stone. A smart grading system goes for the most important variables, presents the information in a way humans can interpret the results, makes the results highly repeatable and correlating to human perception.

The current AGS system seems to do the job without numerical reference to measured metrics, but I feel that providing factual numerical references taken from actual measurements is highly meaningful to consumers. Direct measurement of actual diamond performance metrics is the eventual method which will prove most reliable.

re: Could it be a metric for fire? Yes.
No. Number color pixels has bad correlation with human FIre perception .

re:Since all well cut diamonds of any shape display some degree of fire,

It is not correct statement. Fancy cut with same level Brightness could have quite different FIre