I read on https://gizmodo.com/spacex-starlink-satellites-dodge-137-objects-daily-1850616506 by Passant Rabie:

  • A new report revealed that Starlink satellites had to make 25,000 collision avoidance maneuvers in the six-month period between December 1, 2022 to May 31, 2023, Space.com first reported.

  • SpaceX had previously stated that it would move its Starlink satellites if there is a greater than 1 in 100,000 chance of a collision with a piece of orbital debris or another satellite.

Does that mean that with the current fleet of Starlink satellites and the current threshold to move a satellite based on the chance of a collision, SpaceX expects 1 collision every 2 years? That seems a bit high to me, not to mention that the fleet is expected to increase tenfold (~4k->~40k). Or are some of those numbers off or outdated, or is my understanding incorrect?

  • 2
    $\begingroup$ I'm pretty sure they mean "1 in 100'000 chance of a collision with a piece of orbital debris if nothing is attempted to avoid it". Since they do take evasive action, one would expect the actual collision rate to be several orders of magnitude lower. $\endgroup$ Commented Jul 12, 2023 at 12:23
  • 1
    $\begingroup$ Operators move satellites if other satellites come close (and their definition of "close" is a lot further away than your or my definition of "close"). $\endgroup$
    – RonJohn
    Commented Jul 13, 2023 at 3:41

1 Answer 1


I think you've made a bit of a logical/statistical misstep.

Primarily, the "1 in 100,000" figure is a bit misleading because it's not a statistic. It's an operational risk level threshold that's been set by SpaceX which triggers a preventative measure. It does not mean that SpaceX expects a collision every 100,000 maneuvers.

The 25,000 count of avoidance maneuvers simply indicates the number of times in that time period where the risk exceeded this factor, and then was corrected by the maneuver, because after this process, the risk once again drops to basically zero.

Perhaps the best way to explain this is with a counter example. You'd be correct if the lines went something like:

A new report revealed that Starlink satellites risked collision and did absolutely nothing about it 25,000 times in the six-month period between December 1, 2022 to May 31, 2023


SpaceX had previously stated that they define a satellite as "at risk of collision" if there is a greater than 1 in 100,000 chance of a collision with a piece of orbital debris or another satellite

Or, another example:

A pot of water is being boiled. If left unattended, there is a 99% chance of the pot boiling over. If it is attended, the pot will not boil over because the attendant is attentive and will remove it from the stove or turn it off. In this situation, the stove is constantly being attended by a vigilant cook.

Obviously, the chance of the pot boiling over in this scenario is not 99% because that doesn't include the fact that the cook is standing next to it. Sure, there still is a small chance (the attentive cook could suddenly suffer a stroke), but nowhere near as high.

  • 11
    $\begingroup$ It might be better to not format hypothetical statements as quotes. $\endgroup$
    – asdfex
    Commented Jul 11, 2023 at 18:02
  • 4
    $\begingroup$ Gizmodo July 2023: Starlink satellites executed 25,000 avoidance maneuvers over a recent six-month period—an orbital situation that's set to become even more challenging. makes me wonder if there can be a "dynamical Kessler" where (in the future) a flawed or imperfect algorithm can start moving so many satellites so frequently that it reaches a point of instability and starts crashing them into each other. Sounds like a plot for a bad SciFi movie at least. $\endgroup$
    – uhoh
    Commented Jul 11, 2023 at 22:17
  • 2
    $\begingroup$ I think if there were both that many satellites in orbit and we could still actively control them then most likely we would start de-orbiting them en masse. $\endgroup$
    – Turksarama
    Commented Jul 13, 2023 at 5:53
  • $\begingroup$ @asdfex Any suggestions? I tried a bunch of different formatting methods, but in the end I found the quote format to be the best from a readability and aesthetics viewpoint. Yes, maybe it's a tiny bit misleading but without them, I feel like the answer loses structure otherwise. $\endgroup$
    – Dragongeek
    Commented Jul 13, 2023 at 17:54
  • $\begingroup$ Maybe some simple preformatted text block - that makes it visually stand out without implying a quote. $\endgroup$
    – asdfex
    Commented Jul 13, 2023 at 19:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.