Odds of recently discovered asteroid hitting Earth drop nearly to zero

https://www.washingtonpost.com/science/2025/02/20/asteroid-2024-yr4-odds-hitting-earth-near-zero/

If it were a continuous random walk where they would know for certain one way or the other by then, and everything is calibrated correctly, then the odds would be 60% (= 2.4% / 4.0%)

(edit: See below discussion) But the random walk isn't continuous, almost all of them only go down in probability of impact.

@DanielTilkin I believe they usually go pretty monotonically up and then suddenly down - up because for a while the uncertainty range of the trajectory narrows whilst earth remains within it, then down more rapidly as uncertainty narrows further and excludes earth.

So I think what we expect is for the probability to keep going up, and either to continue to ~100%, or to suddenly drop to zero at some point.

@chrisjbillington Yeah, looks like either can happen:

From https://cneos.jpl.nasa.gov/sentry/faq.html,
Most often, the threat associated with a specific object will decrease as additional observations become available, and so objects will be posted to, and later removed from, our Impact Risk Page.
...
On the other hand, in the unlikely case where a particular potential impact event persists until the orbit is relatively well constrained, the impact probability and associated risk will tend to increase as observations are added.[...] The impact probability will tend to grow as the orbit is refined and alternative and safer trajectories are eliminated. Eventually, the impact probability will drop (usually quite abruptly) to zero or, if the asteroid is really on a collision trajectory, it will continue to grow until it reaches 100%.

https://cneos.jpl.nasa.gov/sentry/details.html#?des=2024%20YR4

Currently this page shows it at 2.4%, while not impossible it seems intuitively unlikely that the odds would almost double unless it's a near miss