The anti-satellite side has a budget hundreds to thousands of times cheaper, based on average ball bearings to satellite density in orbit, with a decent spread and multiple orbit intersections per day.
Approximately as much as a finished satellite or block or gold of similar mass. The bulk of the cost is in launch and aiming.
There are 40,000 SUVs randomly located on the planet.
You fire randomly in the horizontal plane.
What's the probability of you hitting an SUV, and not yourself in the back 84 minutes later?
The earth is big. The probability isn't zero, but for any given orbits period it is low. And both the ball bearings and the satelites will re-enter after a while.
Why randomly, if you're aiming for objects in an known orbit?
> with a decent spread and multiple orbit intersections per day.
If you miss one — and you probably will without active guidance, imagine aiming said gun at an SUV seven time zones away — everything's in a different place by your next close approach. Your ball bearings and the target both, thanks to chaotic pertubations, boosts, drag, etc.
Probably not for Starlink. You’ve got mass-manufactured satellites in a constellation launched on a reüsable, profitable platform on one hand. And on the other hand you have experimental expendable ASAT weapons.