It’s not bullshit, it is a way to replicate the fact that we will never know with absolute certainty what exactly happened.
Our legal system is based on the idea that we won’t know what actually happened; that’s why the word “reasonable” is included in the phrase “beyond a reasonable doubt.” If the business about the numbers being destroyed were actually bullshit, it would be because the legal system was predicated on “beyond ALL doubt,” which it is not.
Wait, what?
Malacandra posed a probability problem that, much like the Monty Hall problem, has an unexpected / counter-intuitive answer. Our resident statistics expert answers it incorrectly, and defends her incorrect answer for multiple pages, steadfastly sticking to “it doesn’t matter how many cars are what color, the witness’s reliability is the witness’s reliability,” which perfectly mirrors a person arguing “it doesn’t matter what door I picked or what door Monty opened, there are two doors left so the odds are 50/50.”
And after reading all that, you conclude that Malacandra is the one acting like a person who can’t understand the answer to the Monty Hall problem?
As a side note, I’ve seen the red/maroon car probability problem multiple times over the years, so it’s driving me crazy that I can’t find an example of it with google. Can anyone else?
Yes, but your scenario implies that investigation won’t get us closer to the truth, that all we can do is look at percentages and thereby make our judgement call.
That’s not how our legal system works. If you can’t come up with something better than statistics, you can’t convict someone on statistics.
That’s why you investigate all credible claims, even if they’re claims of a rare event. And that’s why your analogy falls apart: it ignores the ability to investigate to get closer to the truth.
Daniel
True, that is where my analogy falls apart. But you seem to be under the impression that anybody ever advocated that the DoJ statistics should be enough to dismiss a claim or provide proof for conviction.
Nobody ever advocated that. In fact, nobody ever advocated that such statistics should even be used at trial.
So I’m not sure your analogy holds up any better than mine.
If I’m wrong and you are so confident that I’m wrong, then surely you can explain to the audience how I’m wrong. Where does my logic fall to pieces, oh wise smart one?
Can’t explain it?
Thought not.
Friend Huerta suggested that “probabilistic models” should be used to determine which cases get investigated which do not.
LHoD is not arguing a strawman. The whole issue that got this thread started was the claim that we should use “fact patterns” as a way of triaging allegations and deciding where to shuffle prosecutorial resources. Oh yeah, and as a way of assigning credibility.
I also noticed you didn’t answer his question. Wouldn’t you want to verify a claim that a 4 had been drawn just as much as you’d want to verify a 17? Why would you blindly accept one and not the other?
I did a little digging and found the word problem authored by Amos Tversky which reads thusly
(bolding mine)
Very similar to Malancandra’s right?
Interesting things to note about Tversky’s word problem.
Malancadra’s hypothetical failed to identify who the suspects were and therefore the “at-risk” population. Without that information, you can’t really answer the question with a sound answer. The main reason why it is illogical to treat the total population as the pool of suspects is because then there would be no real reason why our population should be limited to just the city since borders can be crossed. Limiting the population to just the city would be purely arbitrary, but extending it to the whole world just looks insane. That alone tells me that there has to be a specific sample of cars that has been defined by the incident itself that we should be looking at, not the whole population of cars everywhere. The car in Peking, China has a far less likely chance of being the culprit as the car that was spotted driving by the victim 5 seconds after the accident.
Tversky’s word problem does a better job of addressing the problematic assumptions that Malacandra’s neglects to do.
Now Ellis Dee if you see any weaknesses in my arguments, feel free to tackle them. I’m always open to ignorance-eradication. I’ve laid out all my arguments because I’m confident that they based on reason and not a pile of stupid assumptions. If you disagree with my reasoning, challenge me on rational grounds please.
BTW, I enjoyed this probability problem, Malacandra. It was fun.
It wasn’t an illogical answer. You were obliged to make up an example in which maroons were 33% prevalent to get away from the conclusion that we could not in fact put much reliance on the witness’s testimony.
What utter, utter tosh. You claim to know statistics? You have a problem with the fact that the way the prevalence of maroon cars is reflected in a small sample is that it becomes highly unlikely there are any maroon cars in the sample at all, and you think this in some way violates mathematics?
pointing and laughing back. There, that advanced the argument no end.
Again, changing the hypothetical because the one I gave you didn’t produce the answers you wanted.
I, on the other hand, am by now prepared to accept pretty much any display of denseness on your part.
But you cannot have it both ways. Discussions of randomness are inseparable from discussions of mathematical probability, Doctor. (Note how I use your title sarcastically only when I find the display of ignorance unusually startling from someone of your professional standing. Mind you, from you yourself I’m prepared for pretty much anything.) By Zoe’s model, all strings of numbers must be considered equally unique, and all discussion of their unlikeliness, remarkableness, un-randomness or whatever must be considered equally fatuous.
Now, note! I am not opposed to throwing away Zoe’s model; it is not I who have been advocating it. It’s your call.
Hoo-rah. Finally ywtf digs up something that’s mathematically similar to my problem. And, because it’s not posted by me, she can actually agree with the conclusion.
Same math (slightly different parameters). Different flavour text. All of a sudden you find the conclusion acceptable.
Some truly amazing babbling going on here: first of all to explain why cars from out of town should be considered, and then to explain why all the cars in town should not be considered. I’m back to the man-and-woman-on-desert-island story again here.
you with the face is now cool with the notion that “85% of the taxis on the road” is sufficiently restrictive, even though there is still not a whit of evidence to show how many taxis were in the vicinity of the accident location that night, or what the breakdown of colours was among those. No, it’s sufficient to know what the colour balance was of all the taxis operating that night.
“Tversky enters the field with the considerable advantage of not being the person who’s dared to contradict me.”
Mal is carried out on a stretcher in immediate need of oxygen and some serious drugs
Good. I have some more discussion to come on the reporting of unlikely events. Stay tuned.
If nobody’s advocating either of those things, then what relevance do DoJ statistics have?
Daniel
Don’t be so diffident. There’s no “may be” about it.
Dolt, there is much evidence on these very boards that I am nothing of the sort.
You dragged in the Monty Hall in order to line me up with those who Just Don’t Get It, and I’m obliged to point out that you can’t use that brush to tar me with.
No, you haven’t. Saying “You’re wrong!” isn’t pointing out anything. If you said something like “On line 16, Malacandra cancelled the unknown variable from both sides of the equation, which is fallacious as it ignores the solution x=0”, then you’d be pointing out something worth paying attention to. Y’all know the difference between an argument and a contradiction, obviously.
“You’re wrong. you with the face is right. Hah! Pwn3d!!!111!!one!”. Do you actually know enough about the subject to be entitled to advance an opinion?
blah blah ad hominem blah blah you’re ignorant blah blah. Fuck off back to kindergarten and leave the field to ywtf who is at least advancing some arguments - that I think them wrong does not detract from this fact - and not just saying without elaboration that I am talking like a flat-earther or a Creationist, 'cos, like, dude, everyone knows they’re full of shit and if I say you sound like one then, like, you must be full of shit too, hyuck hyuck. :rolleyes: