To be honest perhaps more chance than skill.
Its a real pain to change it because it means disassembling and adjusting the spacer length, to change the spring preload. So the desire to get it right the first time was very strong. And once its assembled, checked, and you find its not quite what you desired, the thought of pulling it apart to change the spacer length is a pain in the proverbial. Honestly, if in my case it did not come close to where I wanted it, I am not sure I would have turned around and made adjustments any time soon.
If one had progressive springs, then I think it would be next to impossible to calculate, because of the varying rate of compression. (please checkout Greencan's method below this post, as thats a very clever way to do so) In the case of fixed rate springs its a lot easier. In my case the Sonic Springs are rated at 1.2 Kg per mm to compress. Having two forks then 2.4 Kg per mm of compression of the front end.
The info I had read on Sonic Springs sort of indicated that perhaps starting with 19mm spring preload or 19mm of compression would provide a safe and good starting point for adjustment.
I tried to do up an excel spreadsheet that would allow me to look at varying loads and see what would sort of sag would occur. The thing that stumped me most was I had no idea what the weight sharing was between front and back wheels. The spreadsheet was based on 50/50 front and back but I am certain that this number is wrong. (you will see why I say this later)
The way I worked it out was,
Take total bike weight, add rider and luggage weight for the average riding condition. (I used 394KG) The assumption was half of the total weight was now on the front wheel. (197Kg) That means I would have 98.5 Kg on each fork spring. Hold that number for future reference.
I then calculated that if I wanted 32mm of Sag then then I would have to preload the spring leaving just 38.4 Kg as sag. Why 38.4? Because 32mm sag, times 1.2Kg/mm equels 38.4Kg.
Its at this point I used the 98.5Kg of fork weight and subtracted the 38.4 Kg of sag to tell me how much to compress the spring. 60.1Kg or 50.1mm (60.1Kg divided by the 1.2Kg/mm = 50.1mm of compression)
50.1mm is a long way from the 19mm starting point others had mentioned. I tried a spacer to obtain 50.1mm of compression but it was too tough to install and I knew something was wrong on both counts. My 50/50 basis must be way off the mark.
This is where a bit of luck came into play. I was guessing that 19mm was to soft and I would have to much sag. 50mm was going to be to hard with to little sag. What could I start with and allow me to cut it down if needed? I made a guess and compromised at 25mm, because I knew I could compress this and install, especially on the RH fork cartridge which is harder to work with than the LH fork.
So I cut the spacers giving each fork spring 25mm of compression. I fully expected I would have to later disassemble to fine tune the length of the spacers to obtain the 32mm sag I deemed as being optimal.
Honestly I was blown away that the 25mm spring preload I used gave me exactly the 32mm I wanted. Miles away from the 50.1mm I had calculated. I concluded that my guess of the 50/50 front and back wheel load sharing is what really throws my calculations out the window, either that or I am completely an idiot and have absolutely no idea at all about whats going on, and just fluked it, in more ways than one.
Trying to answer your question made me go back to my spreadsheet and re-gig it trying to find out at what load between front and back would match the results I experienced. I am guessing that 65% of the weight at rest (incl rider) is on the rear wheel and conversely 35% is on the front wheel.
I am sort of hoping someone might be able to confirm this if they have access to two independent scales one for each wheel. It might prove that all of the above is just gibberish.
I hope this makes a small amount of sense and answers it for you. Happy to answer any other questions.