Theoretical Tiers for PFL to Bizygomatic % Ratio

[i heard the bell]

I haven't been able to find a definitive answer to this ratio's tiers, so I decided to theorize one. You may notice that there's a steeper penalty for PFL's that are shorter rather than longer, because longer eyes generally create a more striking eye area.

Here is my (theoretical) table

Tier	Deviation from Average (Below)	Deviation from Average (Above)
T1	0% Deviation (19-22%)	0% Deviation (19-22%)
T2	0-0.5% Deviation (18.5-19%)	0-2% Deviation (22-24%)
T3	0.5-1% Deviation (18-18.5%)	2-3% Deviation (24-25%)
T4	1-1.5% Deviation (17.5-18%)	3-3.5% Deviation (25-25.5%)
T5	1.5-2% Deviation (17-17.5%)	3.5-3.75% Deviation (25.5-25.75%)
T6	2-2.5% Deviation (16.5-17%)	3.75-4% Deviation (25.75-26%)
T7	>2.5% Deviation (<16.5%)	>4% Deviation (>26%)

I would love some input on this.

 

[Purecel]

First of all, which column is correlated to what? 2nd of all, wdym by tier? And what the the first percent have to do with the 2nd ones in brackets?

 

[Quetila]

First of all, which column is correlated to what? 2nd of all, wdym by tier? And what the the first percent have to do with the 2nd ones in brackets?