Looking At The Washington Post's Data On Police Shootings

According to Data from the Washington Post there were 990 fatal police shootings in 2015.  I looked through this data to find patterns.  In terms of race, as I mentioned before whites outnumbered blacks in terms of fatal shootings.  There were 494 whites shot dead, compared to 258 blacks.  This doesn't count Hispanics who are incorrectly counted as a race.  According to the 2015 census data close to 90% of Hispanics in the country are white.  That is an important point because the FBI's crime data doesn't count Hispanic as a race.  I already did an article about how the data doesn't make it clear that police are racist against blacks, in terms of using lethal force.  Blacks are more likely to be killed per capita, but are also more likely to commit crime, and especially violent crime.

For this article I wanted to look at details of the police shootings.  I looked at the unarmed category, in which 38 of those killed were black, and 32 white.  There were a total of 93, out of which 34 involved an attack in progress.  There were some 9 undetermined, and 50 listed as Other.  If there was one category to look at to criticize police officers it would be the unarmed category, and in that category, it would seem` it would be best to look at other, as opposed to undetermined, and attack in progress.  The problem for the BLM supporters is that looking at that category doesn't do much to back up their case, at least not in my view.  There were some accidents.  One involved police in a shootout with a suspect in a vehicle, in which they shot a passenger.  Another involved Felix Kumi , a bystander who was shot by an officer who was trying to shoot someone else that pointed a gun at their head. There were a few instances of police thinking someone had a weapon when they didn't, but those were by far the minority of the 50. There were instances of people reaching in their waistband, and or pretending they had a gun.  In some cases the Post didn't provide much data.  Like the case of David Kassick (who was white btw).  I looked up his name on google, and found an article about charges being dropped against the officer who killed him.  Kassick was on the ground when he was shot, but a video showed that Kassick's hands repeatedly disappeared under his body, even as officer Lisa Mearkle told him to put his hands where she could see them.  In some other instances there was a struggle, or some confrontation, but further information is not given.  In some of these cases the "unarmed" person had a weapon.  One rushed at officers with a large tree branch.  Another refused to drop a broomstick. One man was throwing rocks at a police car.  Another rammed his vehicle into cars of police, and the U.S. Marshal's Service. In another incident Joshua Omar Garcia, who was a 24 year old Hispanic man, was handcuffed in the front seat of a patrol car, and still tried to drive away, before he was shot.  These are those "unarmed" people who aren't listed as an attack in progress, and even those 50 cases don't make much of a case that police are the bad guys, or racist, and certainly none of them look like premeditated murder.  What would happen if I looked at the instances of those who had a deadly weapon and were shot dead by police?  If the 50 instances of "unarmed" in the Others category doesn't paint a clear picture of evil police, why would the other 940 do so?  I am getting ready to look some of those instances of armed people being killed by police.


References

http://www.dailymail.co.uk/news/article-3305916/Cop-fatally-shot-motorist-cleared-charges.html
https://www.washingtonpost.com/graphics/national/police-shootings/

Police And Black People: Going to the Stats

Police shootings of black men have been in the news lately, in case you haven't noticed.  The shooting deaths of Alton Sterling, and Philando Castile have given Black Lives Matter protestors more reason to rally.  There was also the shooting deaths of five police officers at the hands of Micah Xavier Johnson, who said that he wanted to kill white people, especially white cops.  I have seen numerous people picking out instances of a black man being killed by cops, and a white man not being killed for somewhat similar behavior.  Even though most of these that I saw were bad equivalents, that is not the point.  The point I wanted to make was to focus on the totality of police shootings to see if there is a pattern which shows police racism against blacks.  The total numbers give us a better picture than some facebook post, or some video compilation. 

On the FBI's page when looking for statistics on race and crime the latest stats I could find were from 2013.  I did find Washington Post's data on police shooting deaths by race in the time they have been counting it, but that data is about what happened in 2015.  I wanted to get an idea of what went on recently, so I can assume that the race and crime ratios from 2013 are similar to what they were in 2015.  According to the Washington Post's 2015 statistics there was 990 fatal police shootings in 2015.  In most of them the person killed had a weapon.  494 whites were shot dead by police, while 258 blacks were.  That's a ratio of 1.91/1.  There are far fewer black people than white people in the country, and many adjust for population to try and prove that pattern of racial bias by police.  The problem with that is that those that commit crime are most likely to encounter police.  I wanted to look at crime, and especially violent crime.  Going back to the 2013 there were 2.44 times more whites arrested in than blacks. Using data estimates for population based on the 2015 census, that
would mean that blacks are 2.38 times more likely to be arrested than whites.  Now I am using 2013 crime statistics, and 2015 population statistics, but the ratio should be similar if I had 2015 crime statistics. It is important to look at violent crime.  I'll just guess that a person arrested for murder or robbery, is more likely to get into a confrontation with police than one arrested by embezzlement.
In terms of violent crime that white/black ratio drops to 1.51/1.  That is a ratio lower than the police shooting death ratio.  If we just adjust for violent crime whites are more likely to be shot dead in a confrontation with police.  Factoring in other crime doesn't give a clear picture.  Adjusting for the 2015 population again, blacks are 3.84 times more likely to be arrested for violent crime than whites, and 6.68 times more likely to be arrested for murder and non negligent manslaughter. 

My data isn't quite good enough to either show a correlation, or disprove one, since I am using 2013 crime data, and 2015 population data.  I also cannot adjust for the types of crime, at least not in the way I want to.  There is also the fact that the Washington Post article falsely list Hispanic as a race. There was a study done at Harvard that I just found out about today.  According to the study data police are bias to blacks in terms of use of force, but not in terms of shooting.  The fact that police are more likely to use force against blacks they encounter can be explained by that fact that blacks are more anti police, but what about the other finding.  The study doesn't show bias against blacks by police in terms of shootings.  The study adjust for how, when and where they encounter police, it least it does for use or force, and most likely for shootings as well.  This goes along with other data I've found on fatal police shootings, and police killings.  When you get past pointing out an example here or there, we don't have strong evidence of racial bias against blacks by police when comes to police killing.   

The Blacks Lives Matter movement arose behind a false story.  The false 'Hands up don't shoot' Michael Brown narrative that has been proven wrong by the evidence. It also is motivated by an inaccurate narrative of police targeting black men.  BLM is like a religion though, so I don't expect something as awful as the facts to get in the way of their movement. 


https://www.washingtonpost.com/graphics/national/police-shootings/
http://www.census.gov/quickfacts/table/PST045215/00
https://www.fbi.gov/about-us/cjis/ucr/crime-in-the-u.s/2013/crime-in-the-u.s.-2013/tables/table-43

Harvard  Study
http://www.msn.com/en-us/news/us/bias-found-in-police-use-of-force-but-not-in-shootings/ar-BBubfce?li=BBnb7Kz

Old Emmitt Smith Shit

Emmitt Smith outperformed his running back teammates in average per carry during his 15 year career, but not by a bunch.  He averaged 4.16 yards per carry, while the other running backs on his teams averaged 3.94.  If we look at his 13 seasons with the Cowboys his average is 4.24, and his teammates' is 4.01.  The difference is 0.23.  That would be middle of the line for an all time great.  Pro football Insiders has his teammates average as being 4.00 for his 13 seasons with the Cowboys, so my stats are pretty accurate.  Smith outdid his teammates' averages in 10 of his 15 seasons, and outdid his team's number 2 runner (determined by attempts) in 8 or his 15, but those numbers are skewed by what happened when he was well past his best.  In his first 10 seasons Smith averaged 4.31 yards per carry while his teammates averaged just 3.70.  The reason I look at the first 10 seasons is because that was the length of Barry Sanders' career. The surprising thing is that the Cowboys other runners didn't do that good in years when the team was at their most dominant.  The gap between Smith, and his teams' other runners closed when he was past his prime, and his teammates performed better running the ball.  So the .23 difference doesn't give you a good idea what Smith meant to the Cowboys at his best.  When looking at the stats of Smith's running back teammates, it seems to me that the changes in average was determined more by who the runners were
than who the blockers were. Daryl Johnston, Derrick Lassick, and Sherman Williams didn't do that great behind "The Great Wall of Dallas", but Chris Warren who had proven himself in Seattle did pretty well running behind the Cowboys line.  The one anamoly is Troy Hambrick. He backed up Smith for two years with the Cowboys, and one with the Cardinals.  In those three seasons he averaged 4.6 yards per carry, but when he was the number 1 runner for the Cowboys in the 2003 season, he averaged just 3.5.

One of Smith best qualities was his consistency.  Pro Football Outsides' DYAR(Defense Adjusted Yards Above Replacement), and DVOA (Defense Adjusted Value of Attempt) statistics reward consistency.  Comparing Smith's DYAR to his teammates is not a good comparison, because he got a lot more carries, but the DVOA comparison is a good one, and in those comparisons Smith does better than he does in yards per carry comparisons.  Even in some years where a backup was close to him or ahead of him in yards per carry they weren't close in DVOA. In 1995 Smith averaged 4.7 yards per carry, while Sherman Williams averaged 4.3, but in terms of DVOA, Smith's was 21.5% (higher than average), while Williams' was -17.0%.  In 1996 Smith averaged 3.7, while Williams averaged 4.0, but Smith outdid Williams in DVOA with -0.9% to Williams' -21.1%.  In 1998 Chris Warren averaged 4.9 yards per carry, but had a DVOA of only -21.0%, while Smith produced 9.9% DVOA, while averaging 4.2 yards per carry.  The two were close in yards per carry the next season when Smith averaged 4.2, and Warren averaged 4.1, but they weren't close in DVOA.  Smith's was 9.5%, and Warren's was -14.5%   So Smith helped the Cowboys by averaging higher than average yards per carry, while at the same time carrying a heavy workload, being consistent, and being a great short yardage, and goal line runner. 


Below is a year by year look at Smith's yards per carry, and those of his running back teammates.  Next RB, means the running back with on the team with the most carries outside of Smith.  Next NFB means next Non Fullback, or the runner with the most yards  other than Smith who isn't listed by pro football reference as a fullback.  Those runners are the same person in the last 13 years of Smith's career, and different people in the first two.  OtherRB is simply the yards per carry for the other running backs on his teams.  The Next NFB on the Cowboys got just 6 carries in 1990, and 15 in 1991. 

            Emmitt  Next RB  Next NFB  OtherRB
1990     3.9        4.0           1.0              3.4
1991     4.3        3.2           5.3              3.7
1992     4.6        3.6           3.6              3.5
1993     5.3        3.6           3.6              3.5  
1994     4.0        2.8           2.8              2.9
1995     4.7        4.3           4.3              4.4
1996     3.7        3.9           3.9              4.0
1997     4.1        3.9           3.9              3.8
1998     4.2        3.4           3.4              4.0
1999     4.2        4.1           4.1              4.1
2000     4.1        4.3           4.3              4.0 
2001     3.9        5.1           5.1              5.6
2002     3.8        4.0           4.0              4.6
2003     2.8        3.6           3.6              3.6
2004     3.5        4.5           4.5              3.8 


The averages for Smith and his teammates' through his first 10, and 13 seasons, and then for his full career.
                   Smith  OtherRBs
10 seasons    4.31   3.70
13 seasons    4.24   4.01
15 seasons    4.16   3.94


Rushing Attempts-Effective Yards-Avg (Using Football Outsiders)
First 10 seasons
Smith       3243-16737-5.16
NextRBs    642-2023-3.15
First 13 seasons
Smith     4052-19748-4.87 
Teammates 834-2882-3.46
Career
Smith    4409-20745-4.71
NextRBs  1125-3822-3.51



Data from Pro Football Reference, and Football Outsiders