Luna 25 lander mission, Russian lander following on from the Soviet-era lunar program Options

[Phil Stooke]

The impact site has been calculated now: 57.910° S, 61.450° E. (From the NK forum)

Phil

[rlorenz]

And presumably the "expected success probability" for something like ESA Mars landings needs to be taken with a healthy grain of salt, since there have been no successes...

You miss the whole point of the Cromwell-Laplace estimate (discussed in the supplement to the paper -
the success probability estimator of most use may be informally termed the
Cromwell‐Laplace estimate, equal to (k+1)/(n+2) where k is the number of successes out of n trials. It
essentially encodes the idea that one can never be 100% certain (from Oliver Cromwell’s appeal “I
beseech you, in the bowels of Christ, think it possible that you may be mistaken” to the Church of
Scotland in 1650). Equivalently, it introduces the possibility that one’s luck could run out and the next
trial may fail, even in an otherwise unblemished record so far.)

In effect k+1/n+2 dilutes the track record by the prospect that the next attempt could go either way. It embodies the prior that
no system is 100% reliable or 100% unreliable, and starts with a 50:50 guess if there is no track record, then asymptotically
tends as data accumulate to the frequentist probability k/n. So 'the grain of salt' is baked into the method via Bayes rule.

[rlorenz]

And the Soviet moon landings had a lot more than 1 failure, and a lot more than 5 attempts (if that is what "trials" means).

You're quite right. The supplement notes
Data in this section are drawn from a variety of sources, principally Wilson (1987), Ball et al. (2006) and
Harland and Lorenz (2006) together with contemporary internet sources (Wikipedia entries for Soyuz,
Falcon 9, Luna Program, Cromwell's Rule, downloaded November 2018). Various judgements are made
regarding the accounting of specific examples which could lead to small adjustments in estimated
probabilities. The principal message, that probabilities of success are not (and are sometimes far from)
unity, is robust to the subjective component of these assessments
.....
.....
The Soviet lunar exploration program featured 2 successful semi‐hard landings (Luna‐9 and ‐13), two
successful landers with Lunokhod rovers (Luna‐17 and ‐21) and one rover that failed to depart Earth.
This set of 4 successful landers yields PL~0.83, although adding sample return landers to the set brings
this probability rather lower. Soviet Mars landing attempts Mars‐2, ‐3, ‐6 and ‐7 yielded only 20
seconds of transmission from the surface from Mars‐3 – depending whether that is considered a success
or not, we have 0.17<PL<0.33, although factors contributing to this particularly dismaying performance
were known at the time (e.g. Perminov, 1999).


I forget why I excluded the sample return landers. Note that missions lost on launch would have been rolled up
into the launch success number (i.e. PL is conditional on getting delivered to a landing trajectory, although I
forget why I included Lunokhod-0....) Anyway, I maintain the remark above, that the overall conclusion (that
loss probability on landing attempts is ten to tens of per cent, and not just a few per cent as the Challenger-like
risk assessments always seem to say) is robust...

[fredk]

In effect k+1/n+2 dilutes the track record by the prospect that the next attempt could go either way. It embodies the prior that no system is 100% reliable or 100% unreliable, and starts with a 50:50 guess if there is no track record, then asymptotically tends as data accumulate to the frequentist probability k/n. So 'the grain of salt' is baked into the method via Bayes rule.

I agree that the Cromwell‐Laplace estimate (aka the "rule of succession") gives the best estimate (expectation of the posterior) of the probability, P, of success for independent trials when we assume a uniform prior on P.

My point instead was simply that the variance of the posterior for P gets large for small n. Eg, for ESA Mars landings (k = 0, n = 2) we have mean P = 0.25 but standard deviation of 0.19 (for the beta distribution posterior). In other words, we really can't say much about P in that case except that it's vaguely somewhere between maybe 0.06 and 0.44. That's what I meant by taking P = 0.25 with a grain of salt.

Of course the variance gets smaller as n gets larger, so when there are many launches the value of the expectation P is more meaningful.

[rlorenz]

I agree that the Cromwell‐Laplace estimate (aka the "rule of succession") gives the best estimate (expectation of the posterior) of the probability, P, of success for independent trials when we assume a uniform prior on P.

My point instead was simply that the variance of the posterior for P gets large for small n. Eg, for ESA Mars landings (k = 0, n = 2) we have mean P = 0.25 but standard deviation of 0.19 (for the beta distribution posterior). In other words, we really can't say much about P in that case except that it's vaguely somewhere between maybe 0.06 and 0.44. That's what I meant by taking P = 0.25 with a grain of salt.

Of course the variance gets smaller as n gets larger, so when there are many launches the value of the expectation P is more meaningful.

Sure. To put it another way that avoids people having to mess with Beta distributions, you can get an idea of the size of the required salt grain (i.e. some measure of the uncertainty) from the difference between (k/n) and (k+1/n+2), here 0/2 vs 1/4, or 0.25

[Thorsten Denk]

...you can get an idea of the size of the required salt grain...

Haha, I like this way to express uncertainty.
Thorsten

[kenny]

The French News Agency AFP reports the following, repeated by Moscow Times:

Kremlin vows to pursue Moon race after Luna-25 crash
by AFP Staff Writers
Moscow (AFP) Aug 29, 2023

The Kremlin said Tuesday that Russia would not give up its ambitions to land a craft on the Moon after its first lunar mission in nearly 50 years failed this month.
The Luna-25 module crashed on the Moon's surface after an incident during pre-landing manoeuvres. An Indian mission days later successfully landed near the Moon's south pole.

"We know that the way to the stars is through thorns. The main thing is to continue the Russian programme," Kremlin spokesman Dmitry Peskov told reporters.

"The plans are quite ambitious and they will be realised," he said, adding that the failed mission was not a reason to "tear your hair out".

Moscow Times - Luna program to continue

[Phil Stooke]

https://www.nasa.gov/feature/goddard/2023/lro-luna-25-impact

Luna 25 impact crater - very probably. The story is cautious but I think the faint downrange spray of bright ejecta is pretty conclusive. Dark markings near the impact are also seen at other anthropogenic impacts. Difference image attached.

Phil

[kenny]

NASA Goddard has released an alternating GIF that shows the candidate site for Luna 25, before and after the presumed impact.


Luna 25 impact site before and after