Here’s another object-level tactic I haven’t seen mentioned yet. (Assume LV will not just kill Harry for speaking of non-magical powers. I have a way of increasing the likelihood of this assumption being true)
Harry could explain the Power of Expected Utility Calculations and subtly attempt a Pascal’s Mugging on LV, convincing him that LV can’t possibly assign a probability of less than one in twenty that killing Harry will indeed avert the prophecy, or for that matter cause it, and that the rational action to take is to not kill Harry. He can present it as a “power” to stop the timer and buy a life, regardless of if LV accepts the conclusion, since it is a valuable tool for the future and was probably not in the books Harry gave him to read.
Harry can also explain the Power of Bayesian Probability Updates, both to buy a life and to provide a framework within which to argue that the probability that LV killing Harry backfires is much higher than he previously expected. If the Mugging alone doesn’t work, then Harry can combine this with EV calculations to construct a valid argument that LV shouldn’t kill him.
I’m starting to develop a way to chain this with some other arguments and strategies into a cohesive strategy and I’m starting to feel pretty good about it. Thoughts?
Yeah, this is basically the route I’d do. Except I added one more ingredient. Here, I’ll just quote my review.
Ramble to Voldemort about how you have a better knowledge of decision-making systems and scientific research, and about how if you do destroy the world, it won’t be because you have some world-destroying-property, it will be a result of bad actions—actions that Voldemort, who is starting to learn muggle science, might take just as easily. Indeed this seems likely, since you make such similar decisions, and he may find the prophecy talking about him instead.
If he really wants to stop the prophecy, the way is not to very thoroughly kill one child, the way is to understand the causal path that leads to bad things happening, no matter who does it, and divert it. For this, it would be beneficial if he had you, Harry, alive and well. In fact, you have several ideas already, which for obvious bargaining reasons you will not mention. Since this is fairly honest, you can even say something in parseltongue about how this increases his chances of survival, playing on Tom Riddle’s overriding fear of death.
Then, when he’s distracted considering this, kick him in the grill.
I expanded my previous post in a full solution (very long) with a pretty thorough line of reasoning. In the end, I convinced myself that Voldemort is not acting in his self-interest by killing Harry and that he’s dangerously overconfident in his understanding of the Prophecy and his ability to avert it. Here are the relevant excerpts from my solution:
Tell Voldie that original prophecy (“born as the seventh month dies...”) has not yet been fulfilled, and argue that this calls for rethinking killing Harry because attempting to kill him might cause this prophecy to come true in a bad way and might provide mechanism for “remnant” of Harry to survive and fulfill second prophecy. [Evidence: Snape said “No, I would know if it were fulfilled.” Apparently, neither Dumbles or Minnie found anything strange with the idea that “If it were complete, it would make sense to me,” so that’s likely a known/plausible attribute of prophecies.] LV was very sure that he understood it enough to take risky action that time, but was wrong. “Seems you learned wrong lesson from previous mistake with prophecy, teacher.” Argue that it was lack of understanding of previous prophecy that led to loss of LV’s first body, not the act of trying to turn it to his advantage.
LV’s actions now are driven solely by prophecy. Just as before, it will make him fall victim to it. How can you not realize you’re actions are being completely determined by what you heard?
Tell LV that he strongly believes that things will turn out badly for LV and the Prophecy will not be averted if he tries to kill him (Harry never meant to end his immortality, just temporarily disembody him [so the curse is still in play and LV cannot kill Harry, though my solution has Harry not share this information])
Harry has heard a third prophecy that seems to indicate that Harry will survive beyond this day. It’s better not to tempt fate. (...and 3 their devices...)
If still haven’t gotten him to give in and spare you: “Okay, seriously? You’re Trying to tell me that you still assign a greater than 50% probability to the idea that killing me here tonight is in your best interests?”
Eventually, LV will tell him the prophecy. If not, Harry can just go from his (correct) best guess based on what he heard (“Tear apart the very s-”) Stars? Sun? His first 2 thoughts when hearing it, narrowed down to Stars from Firenze’s comments, and deducing the “End the world” part from the fact that his mother said the world would end if Lilly were nice to her sister and the fact that Firenze called him “Son of Lilly”.
Explain that he’s pretty sure it says he will tear apart the very stars and that it will be the end of the world [...] Harry explains Star Lifting, Dyson Spheres, and the amount of energy required for interstellar travel or other futuristic technologies.
If that’s what the prophecy means, which is way more likely than anything else at this point, since what LV was really doing was not “snipping all threads of destiny” but constraining the solution-space. If he tries/succeeds in killing Harry, then the prophecy will still come true, but neither of them would have any control over how it happened, because all choice has been removed. And, knowing the nature of Prophecy, that’s how bad things happen. So wouldn’t it be better to work together toward making sure the world doesn’t end and that Wizardkind gets a way to escape the Planet Earth, just like Voldie wants? Even if somehow Voldie manages to kill Harry tonight and the prophecy just doesn’t happen which has never happened in the history of ever, what amount of diminishing pleasure from torturing idiots could possibly be worth more than all that?
And at that point, once they’ve solved permanent death for everyone, people will probably want to play War with You. They’ll be lining up by the thousands. Every clever person on the planet will want to try their skills against Lord Voldemort. I’ll play War against you for as long as you want! It’ll be fun! That’s the kind of fun you can only really have in a post-scarcity, post-death society.
This is new as far as I can tell. Please write up a review based around this, and based on a cursory read through of Reddit, it might be best not to do this in prose, it takes even longer to evaluate apparently and Eliezer’s plan has backfired
I submitted it. Here’s the link to my whole solution (It’s long, with backup plans and a few unique mechanics) if you’re interested. I’m pretty proud of it, given the time constraints.
Here’s another object-level tactic I haven’t seen mentioned yet. (Assume LV will not just kill Harry for speaking of non-magical powers. I have a way of increasing the likelihood of this assumption being true)
Harry could explain the Power of Expected Utility Calculations and subtly attempt a Pascal’s Mugging on LV, convincing him that LV can’t possibly assign a probability of less than one in twenty that killing Harry will indeed avert the prophecy, or for that matter cause it, and that the rational action to take is to not kill Harry. He can present it as a “power” to stop the timer and buy a life, regardless of if LV accepts the conclusion, since it is a valuable tool for the future and was probably not in the books Harry gave him to read.
Harry can also explain the Power of Bayesian Probability Updates, both to buy a life and to provide a framework within which to argue that the probability that LV killing Harry backfires is much higher than he previously expected. If the Mugging alone doesn’t work, then Harry can combine this with EV calculations to construct a valid argument that LV shouldn’t kill him.
I’m starting to develop a way to chain this with some other arguments and strategies into a cohesive strategy and I’m starting to feel pretty good about it. Thoughts?
Yeah, this is basically the route I’d do. Except I added one more ingredient. Here, I’ll just quote my review.
Nice. I like it.
I expanded my previous post in a full solution (very long) with a pretty thorough line of reasoning. In the end, I convinced myself that Voldemort is not acting in his self-interest by killing Harry and that he’s dangerously overconfident in his understanding of the Prophecy and his ability to avert it. Here are the relevant excerpts from my solution:
Also can not assign less than 1⁄20 probability that keeping him alive will destroy the world. So...?
This is new as far as I can tell. Please write up a review based around this, and based on a cursory read through of Reddit, it might be best not to do this in prose, it takes even longer to evaluate apparently and Eliezer’s plan has backfired
http://tvtropes.org/pmwiki/pmwiki.php/Main/GoneHorriblyRight (look at last entry in “Fan Works” tab)
I submitted it. Here’s the link to my whole solution (It’s long, with backup plans and a few unique mechanics) if you’re interested. I’m pretty proud of it, given the time constraints.