I’m trying to say that your little “wave a magic wand” business is irrelevant bullshit, that’s all. When you say “You take what your given to work with and make the best possible informed decision you can,” in response to my posts, it sure sounds like you’re dismissing my arguments.
I can’t be arsed, frankly, to read the badger, brandon, and ywtf menage a trois. Everything I’ve read in there seems to me to miss the larger point, which is that none of it relates to the real world. So I’m only reading stuff responding to me. And your stuff responding to me was also missing the larger point.
Yeah, it’s actually an excellent analogy. Given that Apaches don’t take scalps, a report of an Apache scalping is as plausible as a report of a Comanche scalping, given that Apaches do take scalps as long as we assume any of a whole bunch of unproven stuff. Of course, the fact that we wouldn’t have had to assume anything additional in the case of a reported Comanche scalping doesn’t affect the likelihood of either event.
Concerning the roulette example quoted upthread, I have to say that I’d find no reason to doubt either report. It’s not that implausible to score a direct hit on one number - only 35-1 against (worse on American wheels, of course) and given the number of people who play roulette, it happens many, many times a day, and there’s going to be self-selection going on with the reporting as people who stake a buck on one number and don’t win probably see no reason to mention the fact.
On the other hand, if the keeper at my local zoo leaves a typewriter in the monkey cage overnight and the next morning the paper bears the proud inscription:
To be, or not to be; that is the question
the one fact I am unshakeably convinced of is that the monkey didn’t just type the most famous quote in Shakespeare. It can’t read or write, and it wouldn’t hit that exact character string at random if it typed from now until Doomsday. So no matter how high the stack of Bibles on which the keeper swears that no-one went into the cage overnight, I am utterly confident that a set-up has been perpetrated somehow, and only the details remain unexplained.
Again, I think my problem with all this talk about “more unlikely”, “less likely”, “more credible” stuff is that rarely are we forced to compare the likelihood of multiple, totally independent events when examining ONE specific case.
If I tell you that I’ve been hit by a lime-green car, what does it matter that I’m 30 times more likely to be hit by a black car?
If I tell you I found a dollar on the floor, what does it matter that I’m 100 times more likely to find a nickle on the floor?
If I tell you’ve I’ve been raped by a man, what does it matter that I’m 4 times more likely to be struck by a lime-green car full of dollar bills? Or that a woman five years younger than me is more likely to be raped by a man wearing a pink tutu? Or that a fat woman with camel-toe is more likely to be raped while eating a McRib sandwich? I mean…these other events have no bearing on the likelihood of MY claim. Bringing them up doesn’t make sense…
If we were going based on probability figures based off on the frequency of events, all crime allegations would require a heavy dose of skepticism. An average guy is more likely to be walking down the street, minding his own business, than he is to be robbing a bank or mugging an old lady. An average woman is more likely to be sleeping in bed than she is to be shoplifting or setting fire to someone’s house. If we had statistics on the average proportion of time devoted to crime for each demographic group, you would probably see that 99% of the time–most populations are participating in crime-free activities, on average. Does that mean that if I tell you I’ve been robbed by someone–anyone–you should be automatically skeptical?
I really want someone to answer the above question, because an answer would maybe clear up my own confusion about what people are so vehemently defending.
Okay, so 90% of white rapists rape white women. Let’s say you get this statistic by polling white men who have been convicted of rape. Okay. I’m with you.
So you say 90% of accusations against white men (in general) come from black women. We get this information from the Department of Justice. My question is this: How do you leap to the conclusion that these complaints are disproportionately false? And perhaps more importantly, how do you pin any kind of number on this false accusation rate?
In other words, what kind of assumptions do you have to make in order to say this and should you be making these assumptions?
Your conclusion could very well be right. But I could also conclude that white men are raping a lot more black women than they are white women, and because of a racist justice system, the men who rape blacks are more likely to walk than those who rape whites. There’s no reason to assume that your conclusion is any better than mine. Both rest on unexamined assumptions.
You still haven’t proven your case. The likeihood that I, ywtf, Black Woman Extraordinaire, am being truthful when I say that I was raped by my white boyfriend is entirely independent of what generalized crime and conviction statistics are. Because I’m not a die, I’m not a roulette table, I’m not a coin that you can flip a million times. Whether I’m being truthful is a binary thing. Either I am or either I’m not.
You can not attach some arbitrary estimate of probability to my claim, simply because I’m black and other black women have lied in the past. They are not me, therefore any data about them is irrelevant.
Think about it. What if all those black women who cried wolf were money-hungry hookers looking to make a fast buck through settlement. But I’m not like that at all. I’m an upstanding member of the community, a wearer of sensible shoes, a mother of two cats, and I would never dream of crying wolf to make a fast dollar. There is no logical reason why the credibility of my claim is dependent on what a bunch of gold-digging degenerates do. Do you understand this?
And? What I’m talking about has nothing to do with anything about how rape claims should be investigated, and I never said it should. No-one has. So why are you complaining? Is it not possible to have a discussion about the use and abuse of statistics without being nagged to look at the “bigger picture”? I’m not missing any point; I’m just talking about a different one at present, because it’s relevant to the accusation against Bricker. So what?
If that is the larger point, many of threads on this board have no reason to exist, as they too miss real world application. The only one missing the point here is you. If the goal of this is not real world applicability for a criminal investigation, it seems rather ludicrous to place that restriction on it. I fully understand if you can’t be “arsed” to read our discusion, but then why just throw in your snide two cents here and there? If all your doing is posting random comments willy-nilly that aren’t relevant to the actual debate, that’s fine and I’ll respect that, it is after all a public message board. However it would be appreciated if you would let me know before hand next time so I don’t waste my time trying to engage you with a rational counter response.
The point is that to the degree folks think statistics like these should be part of a criminal investigation, they’re abusing the statistics. To the degree that folks don’t think they should be part of a criminal investigation, they’re discussing the breath weapon damage of a green dragon. If you agree with both these statements, I got no beef with you.
Right. However, in the (admittedly) totally non-real scenario posited in the other thread and carried over here, there* are no other* multiple, independent events. The phrase “in the absence of other evidence” kept cropping up to bolster that point.
That is why the whole exchange seemed silly to me. No one inn their right mind would assign credibility based on race alone in the real world. It was not a real world situation.
I think we’re really getting somewhere. If you look back at my long post in the other thread which had all the equations in it, you’ll see I listed some of the many assumptions that must be made; some more justifiable than others, some with which I took strenuous issue. My main bone of contention with Bricker is his broad “all things being equal” blandishment. He would like us to assume in the absence of data that 50% of allegations against whites are made by each race, a simplification I find insupportable. An approximation I might consider, for example, might be to say that if raped, blacks and whites are equally likely to report it; one can argue with this on many levels (blacks might have lower trust of the judicial system, for example), but it’s one I might entertain since its premise is not inherently biased against a given race. But Bricker assumes a disproportion, which I believe is begging the question, and this is why I think he’s completely wrong. I think there are all sorts of other simplifications involved, but I believe these can be covered with an “all things being equal” hypothetical. However, Bricker’s hypothetical was missing crucial data that rendered the whole thing useless.
Absolutely, and I pointed out these and other assumptions to Bricker in the other thread. As I said, our only approximation to occurrence rates is conviction rates, which I entirely agree are only viewed through the filter of the criminal justice system - I made this very point previously. I absolutely acknowledge that there are a myriad other factors which we would have to carefully examine and which fit into a model far more complex than a series of dice throws. If we were, for example, to examine the prior probability of a black person’s allegation resulting in a conviction vs. a white one’s, I expect neither you nor I would be surprised to discover a discrepancy. I would of course make no inference about a given race’s trustworthiness based on such a result, because of all the other factors involved; rather, my instinct would be to go straight for the justice system. But we’re taking a hypothetical here, to make a broader point about statistics.
In short: no, I don’t think the statistic we’re talking about would be very easy to analyse at all in the real world, but my only point throughout is that, given certain data, we could extract it. It appeared to me that you were denying that such a thing even existed, hence my disagreement with both you and Bricker.
Of course I do; many many times I have said that race would only ever be a proxy variable, masking all the factors you mention and more. It’s like when I mentioned about race affecting one’s chances of getting in to a certain school - you and I would surely both agree that race does not affect intelligence or aptitude, but is in this case a proxy variable for all sorts of things like economic status, access to educational facilities, the likelihood of racism in selection procedures and so forth. If we discover that I’m an affluent black person attending a posh private school, then many if not all of those factors are eliminated at a stroke. Nonetheless, given a person and told nothing about them but their race, one is still able to make a statistical statement about their future, weak though it may be.
But the hypothetical was put forth of having only one discriminating factor to observe (race) with a given data set, and my only point is that this could be used to make a probabilistic observation. It’s a lousy bit of data, and the probative value would be atrocious. I absolutely agree that once we start factoring in your character, standing in the community, etc. and so forth, then race becomes utterly irrelevant, as you’re directly examining those things for which race has previously been used as a proxy. You’re absolutely right to say that once we start factoring in things which are of much closer causative connection to one’s reliability, race becomes useless. And as I’ve said a lot, I think you’d find that race is irrelevant anyway. But statistically; not just by declamation.
I think we’re actually pretty close to agreement here; it’d be nice to calm things down somewhat, eh?
Okay, so if we both agree with this basic thing, why do you find yourself unable to reserve judgement in the case of Janice and Marie? In order to do any kind of meaningful comparisions between Janice and Marie, you need the same kind of information that is necessary for assessing the veracity of my claim. It doesn’t matter if you are assigning relative credibility or absolute credibility. For some reason you think that the situation changes when you are asked to view a claim by itself, but it doesn’t.
I, a black woman, claimed that I was raped by my white boyfriend.
Based on just this information and some DoJ stats, how credible is this claim? Rightly, you conclude that you don’t have enough information.
Janice, a white woman, claimed that she was raped by her white boyfriend. Marie, a black women, claimed that she was raped by her white boyfriend.
Based on just this info and some DoJ stats, which claim is more credible? You say that Janice’s claim is more credible. Even though you know nothing about Janice or Marie beyond their race, you feel confident in affording more believability to Janice’s accusation.
There no way you can square this with reason.
Only if you assumed that the white woman is less likely to lie about it than the black woman. That’s the only way you can logically reach your conclusion. The credibility of a specific claim hinges on the credibility of the person testifying to that claim.
I don’t remember seeing him put forth that specific argument, actually. I still think you’re projecting more depth to his bone-headed position than was actually present. Really, if you read the linked thread and the Lying Whore one, you’ll clearly see that Bricker thinks prevalence data can tells us the likelihood that an event has happened. If that number is low, then he thinks that means that a specific event that has already happened is automatically “unlikely”.
At no time did he express an appreciation for the sundry number of underlying factors that could influence what we see in populations. That’s why his “all things being equal” bullshit spawned this thread. But I’m rambling now…
So if you fully acknowledge how inherently tenuous the conclusions you reached from your hypothetical statistics are, why have you been so insistent that I’m wrong about the applicability of stats to the question of “how credible is this claim?” As I stated before, if you go into an analysis with garbage assumptions, you’re going to get garbage results. All along my point has been this: you can not use racial crime stats in assessing the credibility of a specific, individual claim. Why not? Because any estimate you come up will be arbitrary, useless, and therefore wholly irrelevant.
Do you agree with this basic position now?
You haven’t shown why the estimate of credibility would be anything more than gobbledygook. Sure, you could come up with all kind of estimates. But will they be useful? Will they be accurate? How would you know?
Sure you can make a statement, but it would be meaningless. Not just weak but meaningless. A crap shoot basically.
Yes it would be. This ain’t no life and death thing. As long as we both agree that Bricker is dumber than rocks and has failed to own up to it, then all is well.
I would reserve judgement in both cases, however I would find it more likely (credible) that the white woman was raped by a white man. If Case A is shown to happen on a more frequent basis then Case B, than there is a greater chance that Case A will have occured more often if you are to guess which of the two happend at any given time. You don’t have to know anything about their trustworthiness to determine which was more likely to happen which is the essential part of your arguement, honesty is not the issue. The issue is which is more credible (likely) to have happned, and if the numbers show that a white woman is raped by a white man more often than a black woman, then it is more likely to have happened in that manner again. It’s averages and probability, pure and simple, and is taught at all levels of math.
Once you know the trustworthiness and the quality of character of the ones making the accusations the probablity of who got raped changes. The white woman may be a crackhead who’s been thrown this charge of rape at anyone and everyone and the black woman may be a citizen known for integrity and honesty by many, that evidence changes the likelihood of who actually got raped.
If that is not known however, as it wasn’t in the proposed scenario, the probablity that the white woman was raped by the white man is still greater than the probablity the black woman was raped by the white man. Pay close attention that this outcome only applies when no other information is known, as once you comprehend that the only logical conclusion to make is the one I have stated above.
When you say:
The only response possible is I’d need to know more about yourself as a person, how else would one estimate that your telling the truth. What would you expect someone to say to a question like that? The only way to guess about your tendency to tell or not tell the truth is by studying what you have done in the past. Say you’ve lied 10% of the time and told the truth 90%, it would be more likely your telling the truth now, it doesn’t mean that your definently not lying as you have lied in the past 10% of the time. It just seems more credible to believe your telling the truth if statistics show that is what you do more often compared to lying. It is akin to saying it is more credible to believe a white woman was raped by a white man when compared to a black woman being raped by a white man, it’s just more likely to have occured one way as that is what the statistics show, it’s not a statment that the other way has never occured.
Therein is my point, this is not an estimation of who’s more likely to be lying, it’s an estimation of which case was most likely to have occured going off nothing but the race of the parties involved and information that shows the history of race in crimes such as this. Once more, honesty and propensity to fib are not factors in discovering which is more credible.
If a black person says a brick was thrown at him and was seriously injured I’d be more apt to believe him over the white person who said he was seriously injured when a pillow was thrown at him.
Note that of course one of them could be lying and they weren’t hurt at all, however im fairly certain if there were statistics correlating the difference in injuries when being hit by a brick versus a pillow they would show that bricks tend to be more harmful. That is not to say injuries don’t occur as a result of airborne pillows, it just it seems more likely (credible) that the brick would result in an injury. I wouldn’t somehow be disinclined to believe the white person, as people have been been seriously hurt by pillows, but if statistics show an injury from a brick happens more often, then that is the one more likely to have happened. Thus, the black person’s story seems more credible barring other evidence. That is not an automatic judgement that the white person is lying, just that his particular scenario has been shown to not happen as frequently and is thusly less credible than the black person’s.
There’s no rational reason why Janice’s story is more likely just because she is white. None. You might as well be fixated on the fact that Janice’s name start with a J and Marie’s start with M.
Ooooh, more rape victims have “J” names than “M”. So this difference must be significant in some way. Never mind the fact that more women in general have “J” names than “M”. Janice’s claim must be more likely, eh?
Until you understand that the above is an illogical way of thinking, you’ll continue to disagree with me. Apparently nothing that I’m saying is getting through. And frankly I have a headache with all this repetition.
you with the face, I tend to agree with Zakalwe and Brandon. I also think *most *of the points you have made have been correct, but one event seems to have a greater likelihood of having a occurred—and occurring—than the other. Let me put it this way, keeping in mind the “givens” in the hypothetical, two women have claimed being raped by a white man, one woman is black, one is white. Only one claim is true. If you can tell me which one I will give you a million dollars. Which one do you go with? Why?
You’re basically asking me to flip a coin, with black being heads and white being tails. Any guess I make will be a completely random one.
I don’t know anything about the black woman’s tendency to lie in relation to the white woman. You’re asking me to tell which claim is more likely to be true, or to phrase it another way, which person is most likely to be telling me the truth.
You and Brandon apparently do not see that the credibility of a specific allegation depends on the credibility of the person making the allegation. You can not separate one from the other. Your question requires me to know something about the claim-maker’s trustworthiness. In the absence of that information, there’s only one thing I can do to answer your question.
Flip a coin and pray that I’m right. Because a $1 million would be hella good right about now.
Credibility is not in any way, shape, or form, reliant on likelihood. Credibility requires plausibility and believability, not likelihood.
As long as white men raping black women is plausible (it is, and it happens all of the time), then barring any other reason not to believe the claim, a claim of rape by a black woman against a white man should be just as credible as a white woman against a white man.
Credibility stands alone, and requires only plausibility. To use the examples given earlier, if an honest person claimed that they hit a number on roulette, I’ll find them more credible than a habitual liar who claims he hit on red, even though statistically, the latter happens more frequently. If two honest people made the above claims, I’d believe both of them equally, as they are plausible (hell, even the rare one is extremely common). Both events are plausible, thus credibility isn’t hurt by the somewhat rarer nature of the former.
I’ll give you another hypothetical to explain you with the face’s argument.
I do a study and determine, among other things, that in households where domestic violence occurs on a monthly basis, I tend to find a half-gallon of ice cream in the freezer. Doing a random sampling, I find that 75% of violent households have ice cream in the freezer. Only 50% of non-violent households have ice cream in the freezer.
Those are the facts.
Two women claim to have been beaten by their husbands. One likes ice cream and the other doesn’t. Only one claim is true. If you can tell me which one I will give you a million dollars. Which one do you go with? Why?
I’m so very happy that these hypotheticals are, well, hypothetical. They are so abstract that they make my head hurt.
I flip a coin and am as likely to be correct as someone who selects the white woman, as credibility isn’t dependent on likelihood. On the other hand, if you say “Guess the color of the next woman to be raped by a white male.”, I’ll choose white, as frequency of the occurrence actually does make a difference in that case.
No. I’m asking you to take into account ALL of the little information you have. Let’s try a simpler one. I’m waliking down a street in Manhattan. I pass a 10-year-old boy who says " I saw a dog today". I walk further down the street in Manhattan and happen upon another boy who says “I saw a falcon today.”
Now either boy could be lying. Manhattan has both dogs and falcons. So, again, for one million dollars, only one kid is telling the truth, which one?
Which guess increases your odds of gettinig the cash?
