Was an IDF Strike responsible for the Al-Ahli Hospital Explosion?
➕
Plus
773
Ṁ750k
resolved Dec 22
Resolved as
1.0%

Original title: Did the IDF just blow up a hospital on Gaza?

Original description: If the answer is unclear, I will resolve it to my best guess of the median informed observers' subjective probability based on all reliable sources.

Referring to the following attack:

https://www.reuters.com/world/middle-east/diplomats-renew-calls-gaza-aid-iran-warns-israel-2023-10-16/


Edit (Oct 17, 5:32 PM): A more extensive description of the resolution criteria is forthcoming now that this market is getting more activity; the actual outcome will of course stay the same, and I will be resolving the market in the spirit of the question.


Oct 18, 12PM: Some notes on specific questions in the comment section:

  • This is essentially a market about who launched the attack. If Israel intercepted a missile over the hospital and it landed tragically, that won't count as Israel blowing up the hospital for the purposes of this market.

  • "Reliable sources" includes both reliable primary and secondary sources. I will be interpreting all of the evidence I can to make a decision, not just taking the word of particular media outlets. I will also be evaluating (with a grain of salt) any released evidence or assessments from the IDF, USA, and Gaza, taking into account that these are biased parties and they may be lying or otherwise not giving us the full story.

  • Market resolution goes based on what I expect the median informed, intelligent person reading the same evidence would come away with. If I personally have some galaxy brain interpretation of the evidence that I don't think Average Joe Rationalist or disinterested manifold users would endorse, and that interpretation hasn't appeared anywhere in the media and is just my invention, I'll leave that at the door and stick to the apparent facts of the case.

  • I am punting final judgement for now, but I will probably not be resolving this question NO merely because it turns out that the attack hit the hospital's parking lot instead of the hospital itself. That seems to me like the sort of nitpicky resolution judgement that hurts prediction markets' ability to inform others, given the part of the description where I elaborate that I'm referring to the specific attack reported in the media at the time, and given the fact that I and the betters and the people who saw this market on Twitter read the market probablity as asking who was responsible for the reported attack.

Get
Ṁ1,000
and
S3.00
Sort by:

Will give a larger explanation if people want it.

Key points:

  • The physical evidence rules out an airstrike. The cause of the explosion was very likely a rocket launched within the strip.

  • The target (anglican hospital parking lot) was of no military significance. No one has pointed out a target next to the hospital that Israel might have liked to destroy. And it's in Israel's interest to limit civilian casualties.

  • Hamas controls the area surrounding the hospital and has failed to produce any evidence that would help their case, or let external observers in. They also lied in the initial stages of the attack about numbers of dead. They clearly investigated and could not find any details that would help them.

  • Israel has released phone calls that would reveal an unusual level of dishonesty to forge, even for wartime democracies. It's one thing to put witnesses you think might be shakey in front of Congress. It's another thing to lie directly, in a way you can't deny later if caught. It's another thing to completely manufacture recordings of your opposition talking about something they didn't.

  • Most things are not false flags.

So we're still waiting for the actual explanation of how 1% was reached, right?

@jim I pinned the creator explanation.

"Our latest findings on the al-Ahli hospital blast: using 3D trajectory analysis, we dispute the Israeli military’s claim that the hospital was struck by a misfiring Palestinian rocket from a salvo of 17, with most of the damage caused by unspent rocket propellant."

https://twitter.com/ForensicArchi/status/1758563590406086960

@AbdullaFaraz These same propagandists who refer to the IDF as the IOF have been discredited before.

predictedYES

I basically quit manifold markets after this resolution.

predictedNO

@AbdullaFaraz good riddance

predictedYES

There was only one rocket barrage, a failed rocket from which could have caused the explosion at the hospital [17 rockets, starting at 43.4 and ending 30.1 seconds before the hospital explosion, launched from ~ ( 31.490707, 34.414611)].
(There was another, concurrent, rocket barrage [3 rockets] from a different launch site ~ ( 31.542869, 34.453290 ), that is to the north of the hospital, which is not relevant).

No rockets from the 17 rocket barrage had any discernible slowdowns or shutoffs at lower attitude indicating engine failure nor any of them had a steep enough trajectory to impact at the hospital.

Footage of this rocket barrage is available for 3 POVs - Bat Yam, Netivot and HaAsara.

There is a lack of investigation into multiple aircraft, that were doing bombing runs at that time.

2 of those had trajectories and timing relevant to the hospital explosion, both were approximately on approach towards the hospital from the sea:

The first started doing U-turn ~2-3km ground distance from the hospital ~11-16 seconds before the T-23.4 explosion we see on the Netivot footage (that bomb would have been flying approximately over the hospital).

The second started doing U-turn ~1-1.5km ground distance from the hospital ~10-14 seconds before the T-1.7 explosion and the T-0 hospital explosion. (T-1.7 would have been flying approximately over the hospital).

There have been a bunch of other strikes before and after the hospital explosion at locations close to T-23.4 explosion, that would suggest that air-launched bombs were flying over the Al-Ahli hospital.

predictedYES

The base odds ratios of significant explosions in Gaza being caused by IDF artillery (as opposed to IDF airstrikes or rocket misfires or other), divided by the ratio of the fraction of times Gazan authorities have presented to international observers the collected debris from Israeli artillery strikes to the fraction of times they have done so for other explosions, further multiplied by the odds ratio of IDF attacks hitting military targets as opposed to civilian ones, further multiplied by the odds ratio of an IDF statement that contains verifiable claims being true.

That is what the median informed observer would believe, even if “airstrikes are entirely ruled out” was a credible claim, rather than specifically “JDAM” has been ruled out.

My below median research for an informed observer:

Israeli artillery strikes and airstrikes in the last year are so numerous that they aren’t enumerated. Hamas rocket failures are enumerated. So the relevant feature for the base rate is the ratio of artillery strikes to airstrikes. Artillery is significantly cheaper than aircraft operations, therefore I start with a “it is probable that it was an artillery strike” just from “an explosion happened in Gaza”.

I don’t see any cases where Gazan authorities have collected debris from any site of explosion. That makes the observed P(a)=p(a|b)=p(a|~b), not evidence.

IDF attacks in general target apparently civilian targets more often than apparently military targets. The IDF regularly justifies this by claiming that fighters are representing their military operations as civilian operations. I find that particular claim credible, since fighters with apparently military operations would be quickly killed by the IDF. With no way to identify non-civilian operations, I cannot update on the apparent civilian state of the hospital refugee camp as regards any incentive of the IDF to avoid actual civilian targets. Observing IDF reports, I see that they estimate that one-third of the people killed by the IDF are Hamas fighters; taking that report as gospel, my priors on the maximum fraction of people in Gaza who actually fighters is 10%. Assuming that somehow none of the people in the refugee camp were Hamas fighters, the odds ratio would be 1:3 against the IDF having performed the operation. The IDF has definitively shelled refugees, so no update of the form “they wouldn’t do that” applies.

That leaves the update based on the “communications intercept” released by the IDF being accurate and applicable. If it was a true communications intercept, the operational security lost by revealing that the IDF could and did intercept that particular communication has strategic significance to the war as a whole. I do not believe that the Israel intelligence community would release that much of a proof of their capabilities merely to debunk a false claim that they did one particular thing. Since publicly available analyses of the recording didn’t overwhelmingly validate it, the priors against military intelligence choosing to reveal their capabilities dominate, and the release of a probably fabricated recording are largely independent of the source of the explosion.

Being generous at each stage, 3:2 explosions in Gaza are from Israeli artillery, Gazan authorities don’t make debris from any explosions available for international examination, times 1:3 that if all the refugees were civilians they would be less likely to be killed by IDF actions. The Israeli intelligence apparatus is unlikely to choose to disclose their capabilities quickly or for little benefit, and I don’t see any reason why they would be more likely to falsify a recording quickly for a failed rocket than for an artillery strike that caused more civilian deaths than planned.

3:2 times 1:3 odds ratios gives a final of 3:6, a 33% chance that Israel caused that particular explosion. I’m missing about a 1:50 odds ratio to get down to 1%.

@DanPowell You were wrong lol, get over it.

predictedYES

@Shai Do you have hidden 50x evidence or the longer explanation promised?

@DanPowell

Your analysis is just wrong.
>I don’t see any cases where Gazan authorities have collected debris from any site of explosion. That makes the observed P(a)=p(a|b)=p(a|~b), not evidence.

Were there any other cases where dozens of international news and non-profit organizations requested such evidence? This is not the typical case by a wide margin and so the equation is wrong.

How do you explain no journalist in the area snapping a picture and selling it for what I can only imagine to be huge sums of money?

Why did hamas claim that the rocket evaporated?


Other parts of your analysis have similar glaring flaws but I'd like you to respond to this part first.

predictedYES

@Shai When did Hamas claim the rocket evaporated? What did Hamas claim the rocket evaporated? Who did Hamas claim the rocket evaporated? How did Hamas claim the rocket evaporated?

In what other cases has debris from an explosion in Gaza been presented to international observers?

On closer analysis, I did make a rather serious error. I accidentally thought that MH excluding airstrikes meant that I should take the base rate of an explosion being a non-airstrike IDF action. Reviewing the math, excluding the possibility of an airstrike means that I should have considered the base rate of a non-airstrike explosion being an IDF action, which is a base rate on the order of 10^2 :10^4 (dozens of non-IDF non-airstrikes to thousands of IDF non-airstrikes) . So I’m looking for a total of four decades of evidence to get all the way down to a 99:1.

@DanPowell From Human Rights Watch, the word used was "vaporized":

A Hamas official said the remnants would “soon be shown to the world.” More than a month after the events, this has not happened. Ghazi Hamad, a senior Hamas leader and deputy minister in the Hamas-led Gaza governing authority, told the media on October 22 that “the missile has dissolved like salt in the water.… It’s vaporized. Nothing is left.” Human Rights Watch noted that substantial portions of munitions typically survive a detonation, even if parts of munitions are designed to break apart and may be made unrecognizable by thermal damage.

https://www.hrw.org/news/2023/11/26/gaza-findings-october-17-al-ahli-hospital-explosion

predictedYES

@Shai And how likely is it that Hamas made debris available from a particular Israeli artillery shell? That’s a necessary component of evaluating strength of evidence.

@DanPowell Given that they said they would and that doing so would have benefitted them tremendously I would say that the chance of them presenting the debris (conditional on that debris existing) is 90%.

It's also notable that dozens of news organizations have hundreds of journalists working for them in gaza and any one of those could have made that evidence available independently of Hamas.

predictedYES

@Shai Okay, so you claim at most a decade of evidential value on that point, if you generally assume that Hamas is unwilling to affirmatively misrepresent evidence.

If you maintain that Hamas is willing to affirmatively misrepresent evidence, presenting journalists with fragments from a different Israeli artillery shell that struck nearby that night or of one of the artillery shells that struck the hospital in the preceding few days is a possibility, lowering the P(Hamas doesn’t present IDF artillery shell fragments for inspection|the explosion wasn’t caused by an IDF artillery strike) to some value, presumably greater than the P(no fragments presented to media|IDF artillery strike) of 10% that you eventually asserted. I don’t agree with that 10% estimate, but the evidence that would convince me otherwise is either hundreds of sets of artillery fragments having been presented to media or a comprehensive list of particularly notable IDF strikes and at least some of them have had bomb fragments produced on request.

@DanPowell There were not hundreds of other cases where the cause of a blast was disputed and made international headlines for days. Your faux-Bayesian analysis is a sad attempt to muddy the waters. IDF openly says when it bombs Mosques, Hospitals or schools, so evidence is neither requested from nor shown by Hamas. This is not what happened here.

Do you think the missile actually vaporized? If not, why do you think the Hamas official say it did?

predictedYES

@Shai I’ve already acknowledged your assertion of one decade of evidence, if you don’t have the type of evidence that I would expect and don’t have another way of establishing that Hamas could spare the manpower to perform a forensic investigation but doesn’t want to because it knows what that investigation would say, it would require speaking directly to that.

I think that the quote you are currently leaning on was most likely a figure of speech, translation error, or quote taken more out of context than the NYT asserts. What do the independent sources who have followed up with that Hamas official without portfolio say?

If you want to add “Israel takes credit when it commits a war crime” as a source of evidence then you can, but keep in mind that someone took credit for Israel having to attack a Hamas facility in the refugee camp and was later repudiated as an Israeli official, despite being just as much a “high ranking Israeli official” prior to the repudiation as the Hamas official with no portfolio who was the single-source quote that only spoke to the NYT.

And if you’re not trying to waste my time, please give your estimate of the strength of evidence when you present evidence.

@DanPowell

Just to be clear, you still believe Hamas failing to provide anything to support their case is not evidence at all? Despite saying that they would and then lying about the rocket vaporizing?

predictedYES

@Shai P(A|B) =P(A) x P(B|A)/P(B). I think that P(B|A) and P(B) are within a factor of 1.5 of each other, where A is “an explosion was caused by an Israeli artillery strike” and B is “Hamas officials recover explosive fragments and make images of them available to reporters”.

What values do you ascribe to P(B|A) and P(B), if you don’t assign them 0.9 and 0.9P(A) respectively, as you have suggested you do.

[P(B) can be decomposed to P(B|A)P(A)+ P(B|~A)P(~A), if you have trouble estimating the overall odds that Hamas would recover fragments from an explosion an estimate of P(B|~A) can be used instead if you don’t think that if Hamas producing artillery fragments it would be ≈infinitely strong evidence that it was an artillery strike.

@DanPowell So you no longer believe P(A|B) =P(A), that's great.

predictedYES

@Shai Stop wasting my time and put out some numbers or shut your lie-hole.

@DanPowell When did I lie? Please don't be mean :(

predictedYES

@Shai Numbers or GTFO.

And you lied when you either said that “within 1.5x” is the same as “not equal to” or when you tried to gaslight me about my own position. I never agreed with your claim that Hamas is 90% likely to provide IDF fragments “from a particular IDF strike” if it’s an IDF strike but will refuse to provide IDF fragments “from a particular explosive site” if it’s not an IDF strike.

Frankly, I think it’s just as likely that Hamas would produce IDF artillery fragments in response to media inquires whether or not it’s an IDF strike, and I’m surprised that you would consider the production of such easily falsified evidence to be so conclusive!

@DanPowell It wouldn't be *that* easy to convincingly fake evidence but I agree with you it that P(B|~A) is not zero.

But what we got isn't simply P(~B). We got Hamas doing a press release minutes after the explosion among the bodies. A suspicious silence after the IDF presented their evidence. And then occasionally when asked why they aren't providing evidence they would either lie that they will or make up some nonsense about evaporating rockets.

I also don't know why you kept ignoring my point about journalists in Gaza providing the evidence independently of Hamas. It makes all this math kinda worthless.

predictedYES

@Shai Not numbers, you used to be wasting my time. Anyone who wants to try to defend MH can bring numbers.

@DanPowell Here's a number for you: there's a 1% an IDF strike was responsible for the Al-Ahli hospital explosion

predictedYES

@Shai your assertion that you’re pulling three decade of evidence out of your thin air convinces literally nobody with any epistemic hygiene. While I admit that I wasn’t perfectly honest, and put an intentional error in my own initial analysis to see if anyone actually applied the math at all rather than just asserting their own conclusion established with reckless disregard for actual evidence, I didn’t make it hard to find and I carefully applied all of the evidence.

@moderators: pick a fucking side.

Comment hidden
© Manifold Markets, Inc.Terms + Mana-only TermsPrivacyRules