GIA, AGS inconsistent? What Gives?

[Fife]

Here, GIA and AGS both graded the same diamond.

AGS Color: H
GIA Color: J

AGS Clarity: VS2
GIA Clarity: VS1

AGS Report 3/14/12
GIA Report 1/3/12

Is AGS typically less reliable when it comes to color?

 

[Gypsy]

Can be a little less reliable when it comes to color.

 

[TRIALnERROR]

Here, GIA and AGS both graded the same diamond.

AGS Color: H
GIA Color: J

AGS Clarity: VS2
GIA Clarity: VS1

AGS Report 3/14/12
GIA Report 1/3/12

Is AGS typically less reliable when it comes to color?

GIA and AGS are not consistent. AGS has been known to be soft on color by as much as two grades when compared to GIA. AGS typically grades harder on clarity. I've found that vendors will shop grading labs in order to maximize profits, sending it to GIA first and if not satisfied with the results, then sending it to AGS. Here is a recent example of a vendor shopping labs in order to maximize profits, AGS color being soft by two grades, and GIA clarity soft by one grade: ([URL='https://www.pricescope.com/community/threads/found-the-setting-now-i-need-help-finding-the-diamond.177857/?hilit=shopping']https://www.pricescope.com/community/threads/found-the-setting-now-i-need-help-finding-the-diamond.177857/?hilit=shopping[/URL]).

 

[Christina...]

Can be a little less reliable when it comes to color.

There has been lots of talk lately about the color grade differences between GIA and AGS, but until recently it doesn't appear that it's been brought up much. Do you think that there has been a change in color grading with either GIA or AGS? It seems that all the threads accuse AGS of being soft on color grading, but could it be that GIA has become more strict?

 

[diamondseeker2006]

I have only seen AGS one grade up from GIA. I do not understand a two grade difference. That has a tremendous influence on price and should concern everyone.

 

[TRIALnERROR]

Does anyone else have a problem with vendors shopping labs for the sole purpose of maximizing profits? This is the second recent example of AGS being off by two color grades which, like DS stated, has a significant impact on price. I'm sure the reason we don't see more cases is because the GIA report simply "gets lost" and none is the wiser. Personally, I will only purchase a GIA stone going forward for this very reason. If you take the time to read through the thread I posted earlier, you may find it very enlightening. This topic probably deserves its own dedicated thread...

 

[2023]

Here, GIA and AGS both graded the same diamond.

AGS Color: H
GIA Color: J

AGS Clarity: VS2
GIA Clarity: VS1

AGS Report 3/14/12
GIA Report 1/3/12

Is AGS typically less reliable when it comes to color?

GIA and AGS are not consistent. AGS has been known to be soft on color by as much as two grades when compared to GIA. AGS typically grades harder on clarity. I've found that vendors will shop grading labs in order to maximize profits, sending it to GIA first and if not satisfied with the results, then sending it to AGS. Here is a recent example of a vendor shopping labs in order to maximize profits, AGS color being soft by two grades, and GIA clarity soft by one grade: ([URL='https://www.pricescope.com/community/threads/found-the-setting-now-i-need-help-finding-the-diamond.177857/?hilit=shopping']https://www.pricescope.com/community/threads/found-the-setting-now-i-need-help-finding-the-diamond.177857/?hilit=shopping[/URL]).

This link by trialnerror...interesting! definitely could be another thread, i never read it before because of the title.

 

[Fife]

I've emailed AGS about this with the links. Who bets they'll respond?

 

[jkozlow3]

This is disturbing, especially since AGS is supposed to use a Sarin Colorimeter to obtain a preliminary color grade before the final grade is assigned by the GG. Details here: https://www.pricescope.com/journal/diamond-laboratory-process-basic-functions-and-procedures

 

[denverappraiser]

I think it’s a mistake to assume that the ‘error’ was made by AGSL, not GIA (or both) but there’s no doubt in my mind that the reason the dealer sent it in to AGS was exactly because they thought it was a higher color grade and they were unhappy with the J from GIA. This is a common practice and 2 grades is a pretty unusual variation but it’s not out of the range of possibility. 1 grade each for example. Frankly I’m surprised you had the opportunity to see both reports. Standard practice would be to quietly shred the GIA and sell it using the AGS as a sales tool. May I ask how it was presented in the sales pitch?

 

[Fife]

It was presented using the stats from the AGS report.

 

[Laila619]

Does anyone else have a problem with vendors shopping labs for the sole purpose of maximizing profits? This is the second recent example of AGS being off by two color grades which, like DS stated, has a significant impact on price. I'm sure the reason we don't see more cases is because the GIA report simply "gets lost" and none is the wiser. Personally, I will only purchase a GIA stone going forward for this very reason. If you take the time to read through the thread I posted earlier, you may find it very enlightening. This topic probably deserves its own dedicated thread...

Me too. I was never convinced my AGS H diamond was really a true H color.

 

[TRIALnERROR]

I think it’s a mistake to assume that the ‘error’ was made by AGSL, not GIA (or both) but there’s no doubt in my mind that the reason the dealer sent it in to AGS was exactly because they thought it was a higher color grade and they were unhappy with the J from GIA. This is a common practice and 2 grades is a pretty unusual variation but it’s not out of the range of possibility. 1 grade each for example. Frankly I’m surprised you had the opportunity to see both reports. Standard practice would be to quietly shred the GIA and sell it using the AGS as a sales tool. May I ask how it was presented in the sales pitch?

Considering GIA developed the D-Z color scale, isn't it safe to say that AGS made the errors in these examples? Additionally, GIA relies on multiple graders independent opinions to determine the color. Whereas AGS's standard is simply "at least two graders."


How GIA grades color: "Since light source and background can have a significant impact on the appearance of color, the diamond's color is graded in a standardized viewing environment. Color graders submit their independent opinions into the system. During this phase, graders are not privy to color opinions entered previously. The color grade is determined when there are sufficient agreeing opinions."

How AGS grades color: "Color normally refers to where a diamond's Color falls within the range from colorless to light yellow or light brown. At the AGS Laboratories, the Color of a diamond is determined by at least two experienced color diamond graders using a set of master diamonds."

 

[denverappraiser]

Hi Trial,

Who developed the system or when has nothing whatever to do with it. There's not a single grader in the lab who was even born when the scale was developed. Both use the same color scale and identical methodology, even though they don’t describe it the same way on their respective advertisements. At both labs, the approach is that at least 2 out of 3 graders must agree before a particular grade is assigned. If the first two agree, the third is unnecessary so two is ‘sufficient agreeing opinions’ at both. Both labs do their color grading in a neutral environment that are carefully maintained for exactly this purpose.

That said, GIA does have a curious position. For most people, any particular grade, let’s say I color, is a stone that could be reasonably expected to be called I by GIA-GTL if submitted to them. This is not exactly the same thing as a stone that GIA DID call I on some date in the past, but it’s similar and it’s certainly a good start. On the other hand,it doesn't make them right. Will it still be an I tomorrow? Maybe, probably even, but maybe not. Grading is done by humans and humans have variations. Actually, so do machines but that’s another topic. Again, this applies at both labs. Grades can and do get protested and the graders can and do change their call. A J can become an I on second inspection. An H can become an I for the same reason, although customers rarely seem to protest that direction.

 

[TRIALnERROR]

Hi Trial,

Who developed the system or when has nothing whatever to do with it. There's not a single grader in the lab who was even born when the scale was developed. Both use the same color scale and identical methodology, even though they don’t describe it the same way on their respective advertisements. At both labs, the approach is that at least 2 out of 3 graders must agree before a particular grade is assigned. If the first two agree, the third is unnecessary so two is ‘sufficient agreeing opinions’ at both. Both labs do their color grading in a neutral environment that are carefully maintained for exactly this purpose.

That said, GIA does have a curious position. For most people, any particular grade, let’s say I color, is a stone that could be reasonably expected to be called I by GIA-GTL if submitted to them. This is not exactly the same thing as a stone that GIA DID call I on some date in the past, but it’s similar and it’s certainly a good start. On the other hand,it doesn't make them right. Will it still be an I tomorrow? Maybe, probably even, but maybe not. Grading is done by humans and humans have variations. Actually, so do machines but that’s another topic. Again, this applies at both labs. Grades can and do get protested and the graders can and do change their call. A J can become an I on second inspection. An H can become an I for the same reason, although customers rarely seem to protest that direction.

Denver,

Who developed the grading scale and master set should account for something. After all, it's not like AGS developed their own scale, they adopted GIA's. Of course human error can occur in either lab, that goes without saying. However, it seems much less likely at GIA: According to GIA "...depending on size, as many as 40 pairs of eyes may analyze each stone." That statement doesn't seem to coincide with "if two agree, the third is unnecessary."

 

[denverappraiser]

The scale was introduced by Robert Shipley, the founder of both GIA and AGS. GIA was a school for training jewelers about gemstones and AGS was a professional society for jewelers. Neither one started out as a lab.

‘As many as 40’ contains the possibility of ‘as few as 2’, as does ‘sufficient agreeing opinions’. I disagree with your premise that GIA lab has more people that look at a particular stone during their color grading process but I would also argue that this isn’t the heart of the current topic either. 21/40 experts agreeing on something does not generally convince me more than 2/3 nor would I expect it to produce significantly more consistent results, especially since there’s no way for an end user to tell what the ‘real’ score was (40/40 would be considerably more convincing than 21/40 for example). Mind you, I’m not picking on GIA. It’s a fine lab and I think the 2 out of 3 approach is entirely reasonable other than for serious outliers.

 

[Fife]

Grades can and do get protested and the graders can and do change their call. A J can become an I on second inspection. An H can become an I for the same reason, although customers rarely seem to protest that direction.

I understand one mark higher or lower - but going from H to J? That's pretty shoddy.

 

[denverappraiser]

Grades can and do get protested and the graders can and do change their call. A J can become an I on second inspection. An H can become an I for the same reason, although customers rarely seem to protest that direction.

I understand one mark higher or lower - but going from H to J? That's pretty shoddy.

Yes it is.

The missing information is WHY there's such a big difference. I've never seen the stone, so I can't make any statement about which, if either, I would agree with but a 2 grade range is a lot. It translates to thousands of dollars. I'm just pointing out that it could be either or a combination of both labs.

Two color grades isn't a tiny detail. As mentioned, I think the dealer/cutter obviously didn't agree with the J as evidenced by the fact that they immediately sent it to AGS upon return from GIA. I've no clue if they were right but they believed it enough to PAY a second lab so there's at least a certain amount of confidence there. It's still a grade out, and that's still a pile of coin, but I do think it's likely that GIA was a participant in this. That's not the same as calling AGSL innocent but we really don't know. It would be interesting to resubmit it to GIA for a 'color recheck' to see what they say a second round but, of course, whoever owns it isn't going to take the time and money to deal with a lab a THIRD time.

Fife, did you buy it?