- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Coaching Changes
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
1
Posted on 6/3/20 at 5:08 pm to Korkstand
quote:
"45X" figure is bogus
You're the party of science. Why don't you like statistics? What do you think all of your precious Covid models were?
Posted on 6/3/20 at 5:11 pm to WildManGoose
He won’t engage your post that’s what I was saying earlier. Anything that goes against his ideology he just ignores it.
Posted on 6/3/20 at 5:18 pm to Madking
2 downvoters now. It looks like ANTIFA is growing
Posted on 6/3/20 at 5:21 pm to WildManGoose
The guy is a fraud 100%. He’s here to push his propaganda and that’s it. His statement “ideology of truth” lol every ideology is built on the belief that it’s true.
This post was edited on 6/3/20 at 5:24 pm
Posted on 6/3/20 at 5:30 pm to WildManGoose
quote:
We want to know how much more likely a black person is to commit a crime on a white person than a white person is to a black person.
Prob of being white: P(W) = 197MM/235MM = 0.84
Prob of being black: P(B) = 1-0.84 = 0.16
Prob of attack: P(A) = (550K+60K)/235MM = 0.0026
Prob of being white given you're attacking: P(W|A) = 60K/610K = .098
Prob of being black given you're attacking: P(B|A) = 1-.098 =.902
Prob you're attacking given you're white: P(A|W) = (.0026×.098)/.84 = 0.0003
Prob you're attacking given you're black: P(A|B) = (.0026×.902)/.16 = .01466
P(A|B)/P(A|W) = 49
A black person is 49x more likely to attack a white person than vice versa.
Great, now let's do Asians:
We want to know how much more likely an asian person is to commit a crime on a white person than a white person is to an asian person.
Prob of being white: P(W) = 197MM/218MM = 0.90
Prob of being asian: P(A1) = 1-0.90 = 0.10
Prob of attack: P(A) = (79K+44K)/218MM = 0.00056
Prob of being white given you're attacking: P(W|A) = 44K/123K = .358
Prob of being asian given you're attacking: P(A1|A) = 1-.358 =.642
Prob you're attacking given you're white: P(A|W) = (.00056×.358)/.90 = 0.00022
Prob you're attacking given you're asian: P(A|A1) = (.00056×.642)/.10 = .00360
P(A|A1)/P(A|W) = 16
An asian person is 16x more likely to attack a white person than vice versa.
Does this number have any meaning whatsoever in regards to the relative danger posed by either race to the other? What does the magnitude of that number accurately reflect?
quote:We have specifically chosen a subset of data that is biased by demographic distribution in OPPOSITE DIRECTIONS for each race. The minority data is skewed toward the majority, and the majority data is skewed toward the minority. Then we have normalized that data again, which has the effect of skewing the data even more in those same opposite directions.
There it is. No squaring or doubling of population or any of that other b.s. you keep spouting.
Posted on 6/3/20 at 5:32 pm to Madking
quote:It is absolutely relevant in pointing out how the real statistics are manipulated and deceptive.
I couldn’t care less about Asian crime rates and how you spin em. They’re not relevant at all.
Posted on 6/3/20 at 5:32 pm to Korkstand
quote:
If you think it makes any sense to "normalize" the data in the OP to produce "45X more likely",
The graphic didn't say "more likely," it said "more violent toward." It simply creates a "violence factor" with a static rate of violence towards a particular group and changing the hypothetical total population. It's simplistic and almost certainly would prove incorrect if it beared out in reality, but it's no more manipulative than all of your examples.
Haven't read the rest yet, but did I just solve your 18 page confusion?
Posted on 6/3/20 at 5:37 pm to WildManGoose
quote:I love statistics.
You're the party of science. Why don't you like statistics?
quote:I honestly could not name a single covid model. I haven't looked at any of them. I couldn't tell you which ones I agree with and which ones I don't, and I don't care to. They don't matter to me, because they are all biased in some way.
What do you think all of your precious Covid models were?
Posted on 6/3/20 at 5:44 pm to GRTiger
quote:That is probably a good way to put it. They have created a "violence factor", a number with no application aside from exaggerating a more realistic probability into something scarier.
The graphic didn't say "more likely," it said "more violent toward." It simply creates a "violence factor" with a static rate of violence towards a particular group and changing the hypothetical total population. It's simplistic and almost certainly would prove incorrect if it beared out in reality, but it's no more manipulative than all of your examples.
quote:You'll have to ask those who are confused.
Haven't read the rest yet, but did I just solve your 18 page confusion?
Posted on 6/3/20 at 5:57 pm to Korkstand
You can't really provide a figure and call it more or less realistic. There are variables you cannot account for, and probably even more you wouldn't even think to try to account for.
The applicability, aside from I guess scaring whites, is to show how far outside reality the current narrative is, namely, that blacks are the most at risk group of race based violence.
The applicability, aside from I guess scaring whites, is to show how far outside reality the current narrative is, namely, that blacks are the most at risk group of race based violence.
Posted on 6/3/20 at 5:59 pm to GRTiger
Don’t ever forget a black person can not be racist bc they are black. Only white people are racist.
This post was edited on 6/3/20 at 6:00 pm
Posted on 6/3/20 at 6:22 pm to Caplewood
Most people are not racist
This post was edited on 6/7/20 at 3:56 pm
Posted on 6/3/20 at 6:57 pm to Korkstand
quote:
You are looking at the crime that whites commit against 13% of the population, and comparing that to crime that blacks commit against 61% of the population. In other words, you are looking at 13% of white crime and 61% of black crime, and then multiplying black crime by 5 times again.
This is actually a valid point.
This post was edited on 6/3/20 at 6:59 pm
Posted on 6/3/20 at 7:04 pm to Korkstand
quote:
You'll have to ask those who are confused.
Same weak response
Posted on 6/3/20 at 7:11 pm to ctiger69
It definitely isn’t a one sided issue. Some just don’t want to accept responsibility for their actions.
Posted on 6/3/20 at 7:12 pm to Korkstand
quote:
Does this number have any meaning whatsoever in regards to the relative danger posed by either race to the other?
The number does not reflect relative danger. It reflects relative victimization. And your numbers with Asians show that Asians are 14x more likely to victimize whites than whites are to them. Is that crazy to you because you have some racial bias towards Asians; that they can't be violent? Or that whites are statistically much less violent than you assumed?
quote:
We have specifically chosen a subset of data that is biased by demographic distribution in OPPOSITE DIRECTIONS for each race.
That's not what bias is. That's disproportionality, given each singular attack includes 1 white and 1 black. You're using your own assumption to discredit the data instead of letting the data show you what's going on. You're assuming each person has the same probability of attacking the other race. More white people should result in more WoB attacks. That's what bias is. And that's not actually what is happening. Each white person has a 0.0003 probability and each black person has a 0.0147 probability.
You can't assume more white people means more attackers because it could just as easily mean more victims. You have to use the data to determine what it means.
Posted on 6/3/20 at 7:13 pm to Korkstand
quote:
I love statistics.
Only when it’s your “interpretation” of them.
This post was edited on 6/3/20 at 7:20 pm
Posted on 6/3/20 at 7:58 pm to DemonKA3268
Just stopped by to say 45x.
Posted on 6/3/20 at 8:03 pm to WildManGoose
quote:Precisely! That is the problem with portraying such a figure as "45X more violence against whites". That's not what it means. It also doesn't mean that "a black person is 49X more likely to attack a white person", as you put it. What you have said means that if you put 1 white person and 1 black person on an island, it is 49X more likely that the black person will attack the white person than vice versa. That is what "more likely" means, and that is clearly not true! Given the real crime data, the black person is a little over 2X more likely to attack the white than vice versa, not 49 fricking times more likely.
The number does not reflect relative danger. It reflects relative victimization.
Relative victimization necessarily depends on demographic ratios, and if you must produce a number to reflect this, do not dishonestly portray it to mean "more likely" or "more violence".
quote:Again, not "more likely". This misrepresentation is exposing your bias.
And your numbers with Asians show that Asians are 14x more likely to victimize whites than whites are to them. Is that crazy to you because you have some racial bias towards Asians; that they can't be violent?
quote:Whites are about as violent as the actual statistics show, which is to say about twice as likely to be violent as asians, and about half as likely to be violent as blacks. Not 1/49th as likely.
Or that whites are statistically much less violent than you assumed?
quote:That's exactly what bias is. We have chosen subsets of data from two different populations based on two different criteria!
That's not what bias is.
quote:Yeah, disproportionality which when selected for specific criteria results in a biased dataset.
That's disproportionality
quote:The data shows us what's going on. You are manipulating data to tell a story that you want to hear.
You're using your own assumption to discredit the data instead of letting the data show you what's going on.
quote:No, I'm not.
You're assuming each person has the same probability of attacking the other race.
quote:But what you fail to realize is that more white people should ALSO result in more BoW attacks. Each additional person, regardless of race, is a potential attacker as well as a potential victim. This is exactly why you can't take the raw number of attacks each way, and then only multiply one subset by a multiple and not the other. That's not how it works, and the result is meaningless because of that fact.
More white people should result in more WoB attacks.
You must understand the reason for doing the calculation in order to understand the result. In this case, simply scaling one number up doesn't "normalize" the data... it blows it out of proportion. This is due to the fact that if the black population were scaled up 5X, the number of BoW attacks might scale up 5X with it, but the number of WoB attacks would certainly scale up as well! Maybe it wouldn't scale up by exactly 5X, but it would definitely increase maybe by less, or maybe more. In other words, you are adjusting the numerator without regard to how doing so impacts the denominator. If you really want a result that in any way reflects the actual likelihood as you say you want, you must consider the real underlying rates of total crime. You don't seem willing to do that.
Page 18 of 23
Popular
Back to top
Follow TigerDroppings for LSU Football News