But the problem, if anything, was that too much dense information was conveyed at all. Based on the analysis in the post, of the engineers had replaced that slide with one that said "Don't go forward with reentry", that might have saved lives better than any change in medium. To be clear, I'm in favor of abolishing PowerPoint for any non-ephemeral use, but the problem here was focus and framing of the info.

▲ xp84 8 days ago | parent | next [-]

I agree completely. My deck would have been:

Slide 1: 48-point font

  Don't go forward with reentry

Slide 2: 24-point font

  * Our foam collision dataset from experimentation only included pieces below X cu in.

  * Evidence points to this piece being at least Y cu in - 200 times more massive

  * Catastrophic damage to the wing cannot be ruled out

This would have been a great PowerPoint, and I'm not convinced handing them only an academic paper with dozens of pages of facts and figures would have had the effect that my above deck would have had.

▲

notahacker 7 days ago | parent | next [-]

Yeah, a slide like that would have been fine; fundamentally the slide is exactly like an academic paper in that some of the most interesting implications are just banal statements of the numbers in the data and qualifiers. In fact, the slide is much better than the academic paper... it doesn't contain much in the way of irrelevant data and qualifiers.

In practice Tufte and bloggers and commenters are retconning messages engineers not possessing foreknowledge of what was going to happen didn't wish to convey. The slide isn't supposed to say "no reentry" not because engineers don't know how to say no using PowerPoint, but because what the engineers are actually saying by selecting those points for consideration is "damage is theoretically possible but not in our simulations which test data suggests are actually on the conservative side; the test data is only at a very small scale though". If they'd dumbed it down, the slide would have said "it could go wrong but the limited data we've got suggests it won't"

	▲	xp84 6 days ago \| parent [-]
		I'm basically with you and let's be clear, I know nothing about the domain. But the word "limited" is doing a lot of work in that last sentence. Maybe a diagram showing relative size of the biggest foam chunk they'd tested vs. the size of the actual one would have been useful.

▲

andrewflnr 8 days ago | parent | prev [-]

Agree. Though to be honest I still think a paper with an executive summary that said "Don't go forward etc" would have probably been even better. Then the powerpoint slides can be illustrations of how far outside the testing data this is, simulations of possible damage, and other, you know, useful stuff.

▲ cxr 7 days ago | parent | prev | next [-]

> Based on the analysis in the post

The analysis in the post is dogshit and misrepresents the review board's actual conclusions.

> But the problem, if anything, was that too much dense information was conveyed at all

That's totally opposite to what the members of the review board identified as the problem.

▲ stinkbeetle 7 days ago | parent | prev | next [-]

The person I was replying to said they had no idea how it had anything to do with PowerPoint. The article quotes the report as identifying over-use of it as one of the problems that contributed to the communication break down.

I'm not making the argument and I'm not interested in engaging with this quibbling, I'm just explaining how the article said the expert who conducted the investigation found a problem with their use of PowerPoint. If you have a problem with that conclusion, then take it up with the investigation report, not me. I would be fascinated to see you provide a rebuttal of it.

▲ throwaway173738 7 days ago | parent | prev [-]

The engineer in question was probably trained by the organization to leave decision-making to management. In organizations like that you refrain from statements that sound like decisions.

Saying “more testing must be done before deciding to re-enter” would be equally valid.