Hi Mieke and @reeflings, I just made a VERY ROUGH simulation of the price fork: https://c301.nl.tabdigital.eu/f/178895.
Technical point on just how rough it is
I have not bothered to impose the condition that the sum of the price of all units is equal to the total construction cost.
SUM(apartment price x apartment surface) = medium price x total surface
In real life, this is important to have, because it makes sure that, once everyone has paid for their unit, the costs of the project are entirely covered.
Since there are more lower floor units than upper floor ones, there will be a discrepancy. it is likely that the coefficients in my simulation (-10%, -3.33%, +3.33%, +10%) might need to be tweaked upwards to make up for the discrepancy. For example -8.5%, -2%, +3.33%, +10%. You are welcome to add this condition if yoy feel like it.,
Each unit is assigned a “discount coefficient” from -10% to +10% depending on the floor it is in, with two exceptions. The logic is explained in the spreadsheet.
You can also play with it, for example with a narrower fork of -5% to +5%. To do so, enter the value you want in cell J1.
The real thing will be more sophisticated, but in the end the floor reflects sunlight, quiet and privacy, so it is a reasonably good proxy.
It seems reasonable that at least O8, O9, O10 and I5 would be in the top category, yes. Jury is out for A7 and A8.