First thanks for this great ressource! I believe I may have found a bug, or am I just holding it wrong?
I recently was looking at a diamond, SKU 11039206. I found it on the efficient frontier in the “Diamond Prices” tab. I did make sure to select the Ideal proportions filter. The diamond appears on the curve as a brown square (which, IIUC means “ideal proportions”).
I then went to James Allen and asked for the certificate of that diamond. It mentions: Table: 57.5%, Crown angle: 36.4°, Pavillon: 40.1°.
According to the Cut estimator, this is “AGS Excellent/GIA Very Good”, which is quite far from “Ideal”. Is it a bug, or did I just misunderstand what “Ideal Proportion” means?
Thanks in advance for your help!
For James Allen, I use computer vision to evaluate cut instead of going by the numbers. In this case, the evaluation of the optical symmetry disagrees with the numbers. I trust the vision more than the numbers when it’s close (e.g. 34/40.6, 36.5/40.6), but this might be too shallow of a pavilion angle to go purely based off of vision. No machine learning classifier is perfect, so I’d attribute it to that.
It’s also somewhat moot. I think the options below would be called as “ideal” by my algorithm and had better proportions if you request the grading report.
Shape,Carat,Color,Clarity,Cut,Price (USD),Affiliate Link