What should be the criteria for taking any decision? In life, we won’t have simplistic mathematical probability that the success rate is 2% or 95% and so on. We will have some knowns and some unknowns. We will have cost benefit analysis, pros and cons of all possible actions. How to take critical decisions in a war like situation or in general with so many unknowns?
probabilistic thinking, and a grave appreciation that uncertainty can never be eliminated.” A longer answer would take the many books in the “decision-making under uncertainty” section of the library.
That is NOT a glib response. It is a sensible and thoughtful response. Many times times we can quite easily make a rather obvious or simplistic decision either because the evidence is clear and/or the the consequences of poor outcomes are negligible. When they aren’t though, humility and caution must reign supreme.
Dan why is that “glib”? The only thing I would add to it is that decision makers must always remember that their decisions—including a decision to do nothing—have real consequences for real people up to, and including, death!
Sometimes I think the harder the truth, the less engagement, which you mentioned head-on. I resonate with your comment on the amount of analysis it would take to get a valid or useful conclusion.
Often, I have very little interest in current events because of this reason, I don't have enough information (most of which is intentionally hidden from the public) to make a meaningful judgment nor would my judgment actually mean anything, as I have very little agency in the outcomes of geopolitics as it currently stands. Sure, I'll speculate given the information I know, but it means absolutely nothing to me. There is no reward, usually no relevance, and no means of validation.
Still, there are those who yell and banter, cut their friends and family off for disagreeing with them, and have all these unreasonably strong opinions, maybe for the sake of maintaining psychological homeostasis (reducing dissonance, uncomfortable feelings of uncertainty) or as social heuristics for signaling group membership.
Epistemic tyranny drives me crazy, my biggest gripe with Trump. We need to normalize uncertainty.
I do think there's a disanalogy that makes judging decisions by outcomes more defensible in realistic cases than in your example. When it comes to betting your life savings on a roulette spin, we know exactly what odds you're facing. Even if you win, we can know you were extremely unlikely to; the outcome doesn't provide any important evidence about the process, because we know all we need to know about this very simple process.
When it comes to military decisions, there's lots more uncertainty, not just about the outcomes of potential actions, but about the chances. If I, with my extremely limited information, think some course of action is likely to lead to bad results, and then it succeeds, I could think they just got lucky. But I might also reasonably revise my judgment about how improbable success was. In a setting where I know to a mathematical certainty how probable success is (your casino case), this dynamic is absent.
Yes. It’s not hard to imagine practical circumstances in which the outcome tells us something valid about the quality of the decision. Hence my “generally” and “often” waffling….
Couldn’t agree more. Effects - the outcomes of decisions - have three broad causes: systemic forces, your agency/actions, and randomness. You control only one. Unfortunately, we overestimate the importance of that one - especially when we like the outcome. And this leads to problems when we don’t. All we can do is learn to judge our decisions by the quality of the process we use, given the time pressure we face and the information we have.
Bracketing the other content, my favourite take on the "if the person at the top of the pyramid indicates what information they want to receive, this changes the quality of the information, etc." stuff is the old Brad Delong "the Cossacks work for the Czar" stuff.
I think that the very early, classified assessments of the impact of the US’s strike on Iran’s nuclear assets were just that - very early and classified. They should not have been made public whether they confirmed or contradicted the President’s claims.
There’s a whole discussion to be had about “intention”, whether conscious, semi-conscious, or UNCONSCIOUS (!). Announced purpose is at the heart of political practice: persuade the right (or wrong) people & you’re in business.
Thoroughly enjoyed the piece until you went (tangentially in your words) to comment further about the future of US intelligence. Agree that Trump's outburst not surprisingly with his NPD and past performances were way out of line. But to the point of information available to him before the US intervention, the basis of his decision, seems the Mossad has had more accurate information quantitatively and qualitatively than the very IAEA that you quote today. You could just as easily (? and accurately?) used/referenced the several posts by David Albright (previously of the IAEA) both before and after the strike. He didn't make iron-clad guarantees but offered a lot of insight about the situation before and after the strike. Anyway, I have no idea what the impact of Trump's inappropriate outbursts will have on our intelligence gathering globally going forward, but at least as relates to intel on Iran's nuclear ambitions is concerned the Mossad likely knows (and won't likely say) where the "maybe missing" enriched uranium is at this very second and also won't be hampered by comments by our intemperant president. Nice article. Looking forward to more.
What about situations in which we say, “right decision, wrong reasons”? The decision-making process can be separated, analytically, from the decision as well from the results, right?
There's a key difference with your analogy. With roulette, the odds are known by everyone in advance. But in geopolitics, some people have more information than others. The administration had reports from US and Israeli military intelligence, the result of years of research and spying, so they had a much better estimate of the odds than the public.
If the mission succeeded, we should revise our estimates of the probability of success upwards. If they failed, we should revise it downwards.
If our belief is something like "we should attack if the probability of success is at least 50%", then we can indeed change our opinion of the quality of the administration's decision based on the mission's eventual success or failure.
The roulette example only illustrates that decision quality and outcome can be entirely separate. I did not suggest it was analogous with Trump’s decision.
And in your comments, you’ve switched the focus from Trump’s decision to the judgement of the IC. Trump decided, presumably, after being given odds by the IC, and the former, not the latter, is my concern here. As I said, there are circumstances where outcomes contain information that can and should be included when judging prior judgements.
What should be the criteria for taking any decision? In life, we won’t have simplistic mathematical probability that the success rate is 2% or 95% and so on. We will have some knowns and some unknowns. We will have cost benefit analysis, pros and cons of all possible actions. How to take critical decisions in a war like situation or in general with so many unknowns?
My glib answer is “with humility,
probabilistic thinking, and a grave appreciation that uncertainty can never be eliminated.” A longer answer would take the many books in the “decision-making under uncertainty” section of the library.
That is NOT a glib response. It is a sensible and thoughtful response. Many times times we can quite easily make a rather obvious or simplistic decision either because the evidence is clear and/or the the consequences of poor outcomes are negligible. When they aren’t though, humility and caution must reign supreme.
Dan why is that “glib”? The only thing I would add to it is that decision makers must always remember that their decisions—including a decision to do nothing—have real consequences for real people up to, and including, death!
Sometimes I think the harder the truth, the less engagement, which you mentioned head-on. I resonate with your comment on the amount of analysis it would take to get a valid or useful conclusion.
Often, I have very little interest in current events because of this reason, I don't have enough information (most of which is intentionally hidden from the public) to make a meaningful judgment nor would my judgment actually mean anything, as I have very little agency in the outcomes of geopolitics as it currently stands. Sure, I'll speculate given the information I know, but it means absolutely nothing to me. There is no reward, usually no relevance, and no means of validation.
Still, there are those who yell and banter, cut their friends and family off for disagreeing with them, and have all these unreasonably strong opinions, maybe for the sake of maintaining psychological homeostasis (reducing dissonance, uncomfortable feelings of uncertainty) or as social heuristics for signaling group membership.
Epistemic tyranny drives me crazy, my biggest gripe with Trump. We need to normalize uncertainty.
Husetics? Heuristics I think?
I do think there's a disanalogy that makes judging decisions by outcomes more defensible in realistic cases than in your example. When it comes to betting your life savings on a roulette spin, we know exactly what odds you're facing. Even if you win, we can know you were extremely unlikely to; the outcome doesn't provide any important evidence about the process, because we know all we need to know about this very simple process.
When it comes to military decisions, there's lots more uncertainty, not just about the outcomes of potential actions, but about the chances. If I, with my extremely limited information, think some course of action is likely to lead to bad results, and then it succeeds, I could think they just got lucky. But I might also reasonably revise my judgment about how improbable success was. In a setting where I know to a mathematical certainty how probable success is (your casino case), this dynamic is absent.
Yes. It’s not hard to imagine practical circumstances in which the outcome tells us something valid about the quality of the decision. Hence my “generally” and “often” waffling….
“it’s universally respected head”
—> “its universally…”
Couldn’t agree more. Effects - the outcomes of decisions - have three broad causes: systemic forces, your agency/actions, and randomness. You control only one. Unfortunately, we overestimate the importance of that one - especially when we like the outcome. And this leads to problems when we don’t. All we can do is learn to judge our decisions by the quality of the process we use, given the time pressure we face and the information we have.
I pine for politics that embraces humility and complexity. It was never completely here. But it is now almost completely absent.
And the true idiot is the winner who turns around and bets it all on a single number again.
We all do something stupid at least once in our lifetime (and live to tell the tale), but only idiots do the same stupid thing again.
Sports is full of evaluations succumbing to outcome bias. But here’s a great example from a veteran pitcher of not letting a younger pitcher fall into the bias trap. See first five paragraphs in https://www.sportsnet.ca/mlb/article/lauer-fitting-in-to-blue-jays-veteran-rotation-as-they-take-series-over-red-sox/Lauer fitting into Blue Jays' veteran rotation as they take series over Red Sox - Sportsnet.ca
Bracketing the other content, my favourite take on the "if the person at the top of the pyramid indicates what information they want to receive, this changes the quality of the information, etc." stuff is the old Brad Delong "the Cossacks work for the Czar" stuff.
https://delong.typepad.com/sdj/2008/01/do-the-cossacks.html
I think that the very early, classified assessments of the impact of the US’s strike on Iran’s nuclear assets were just that - very early and classified. They should not have been made public whether they confirmed or contradicted the President’s claims.
There’s a whole discussion to be had about “intention”, whether conscious, semi-conscious, or UNCONSCIOUS (!). Announced purpose is at the heart of political practice: persuade the right (or wrong) people & you’re in business.
Thoroughly enjoyed the piece until you went (tangentially in your words) to comment further about the future of US intelligence. Agree that Trump's outburst not surprisingly with his NPD and past performances were way out of line. But to the point of information available to him before the US intervention, the basis of his decision, seems the Mossad has had more accurate information quantitatively and qualitatively than the very IAEA that you quote today. You could just as easily (? and accurately?) used/referenced the several posts by David Albright (previously of the IAEA) both before and after the strike. He didn't make iron-clad guarantees but offered a lot of insight about the situation before and after the strike. Anyway, I have no idea what the impact of Trump's inappropriate outbursts will have on our intelligence gathering globally going forward, but at least as relates to intel on Iran's nuclear ambitions is concerned the Mossad likely knows (and won't likely say) where the "maybe missing" enriched uranium is at this very second and also won't be hampered by comments by our intemperant president. Nice article. Looking forward to more.
What about situations in which we say, “right decision, wrong reasons”? The decision-making process can be separated, analytically, from the decision as well from the results, right?
Quote
The ends justify the means
Unquote
Niccolò Machiavelli
There's a key difference with your analogy. With roulette, the odds are known by everyone in advance. But in geopolitics, some people have more information than others. The administration had reports from US and Israeli military intelligence, the result of years of research and spying, so they had a much better estimate of the odds than the public.
If the mission succeeded, we should revise our estimates of the probability of success upwards. If they failed, we should revise it downwards.
If our belief is something like "we should attack if the probability of success is at least 50%", then we can indeed change our opinion of the quality of the administration's decision based on the mission's eventual success or failure.
The roulette example only illustrates that decision quality and outcome can be entirely separate. I did not suggest it was analogous with Trump’s decision.
And in your comments, you’ve switched the focus from Trump’s decision to the judgement of the IC. Trump decided, presumably, after being given odds by the IC, and the former, not the latter, is my concern here. As I said, there are circumstances where outcomes contain information that can and should be included when judging prior judgements.