Baltimore riots

[Rupert Pupkin]

Originally Posted by Danzig

There has been study after study discussed in the press regarding bias against blacks. From kindergàrten to high school, stop and frisk in NY where whites more often had contraband, but blacks were stopped far more often...same with Ferguson, then there's disparity in court, sentencing, bias in who gets the death penalty. Bias is in every part of the criminal justice system.
Google it, look it up.
I have read plenty on the subject, which is why I know all that. And I am sure to read studies by viable groups with rock hard stats, not crap where someone has no data on 25% of his supposed subject, and then narrows it down with a subset, all so he can get the result he sought. Reminds me of the guy who did the vaccine ’study' which has since been debunked.
Look up harvards racial bias study for a start. Or Stanford, the professor earned a MacArthur fellowship award.

I don't disagree with most of that. I think there is plenty of evidence that racial profiling exists in many areas. I don't deny that. There has been a history of bias in the criminal justice system in many areas. I don't deny that. But when it comes to shooting violent criminals, I don's see any evidence of a bias.

You keep repeating yourself with regard to some missing data. We could do an experiment on anything. It could be on any subject. It wouldn't matter what we were analyzing. If we had a big random sample, the study would be valid. We would not have to have complete data. I will give you an example. Let's say there is a traffic light that takes over 5 minutes to change. It takes so long for it to change, that many people get fed up and jaywalk when the light is red. If we wanted to figure out what percentage of people jaywalk at that light, do you think we would need to count every person for a year? Of course not. If we counted for just a couple of days and counted 1,000 people, that would probably be a plenty big enough sample to get a very good idea of what percentage of people are jaywalking that light. If we saw that 50 people jaywalked out of 1,000, that would tell us that 5% of the people jaywalk. That would a big enough sample to have fairly accurate data. If we counted 5,000 people, we would still get the same result. If we counted 5,000 people but then we lost the data on the final 1,000 people, would that change our result? Of course not. As long as you have a good size random sample, you don't need complete data. The percentages are not going to change.