I just found an fascinatingly coincidental calculation while planning an extension to my network. So I already have my base constellation set up around Kerbin: 3 sats equally spread around an equatorial orbit with an orbital radius of ~1400km (800km altitude) and each carrying a mid-range omni and a relay dish. Next step in my exploration program is to send an unkerballed rover to the farside of the Mun and for that I obviously need to be able to relay the signal. I like to keep things lite so I decided to aim for a single relay satellite at the Mun end rather than the full three. The first problem was how to make sure the relay stays above the farside so that I can get full-time coverage of my already-sure-to-be-tediously-longwinded rover mission. If this was the real world then I might consider shoving it into the L2 point and hoping it stays there long enough to complete my mission. But KSP uses a single-body physics approximation so there aren't any Lagrangian points. So instead I settled for putting my relay just beyond the Mun's SOI, so that it's orbiting Kerbin slightly faster than the Mun. This means that eventually it will overshoot the Mun and become useless but I did the calculations and made sure it would be in position long enough to land my rover and have a good look around. (Interestingly this puts it very close to where the L2 point would be, because L2 is generally found close to the radius of the smaller body's Hill sphere - presumably the real-world basis for calculating KSP's SOIs) The next problem was to make sure that if I put a relay satellite here it wasn't going to be inside the shadow of my Kerbin constellation. So I drew some diagrams and did the calculations based on the point where my 3 KerbSats would be casting their longest shadow. I used their actual orbits in the calculation and then got lazy (while erring gently on the side of caution) with plotting the signal's path past the Mun (I pretended the Mun was a cube). But the figure I arrived at for the longest signal shadow cast by the Mun was: 2,431,028.1 m... And the SOI of the Mun (according to KSP wiki)? 2,429,559.1 m This seems to be more of a happy coincidence than anything derivable (although if someone wants to prove me wrong, I would love to see it) but it seemed like too beautiful a coincidence not to share