Beto 2020

[Bozo_Casanova 26508]

Cool.

So your only reason for keeping assault weapons on the streets is because they are fun to shoot and the microscopic chance you are cornered and attacked by a roaming sounder of hogs?

You realize that (poor) people still have to take off their shoes to board an aircraft, because one person tried and failed to detonate a shoe bomb on a flight? Why don't they just put all of that TSA money into mental health screening instead?

 

 

Angry Beto Bro is angry

[Bozo_Casanova 26508]

I just don't buy that.   Who says it won't stop anything?   It will stop plenty of potential killers who either won't be able to figure out how to get the weapons on the black market or who will be caught in stings while trying to find the black market.   People who could have otherwise gotten the guns from their parents, or at Walmart. 

And if you want to talk efficacy, how many lives do AKs and ARs actually save?   How many people will die from attacks of feral hogs and meth head gangs if people are deprived of those guns and have to make do with shotguns or just running from the danger?   And I'm not trying to be cute about the examples you gave — I get that your point is that these guns have a practical use when a person is under seige from a big group of people or animals.   Yet I haven't seen many "Man mows down 14 attackers with AR-15" stories in the news.   If banning these guns won't cost any lives, why isn't it worth the chance that even one school shooting will be prevented?   Let alone that many might be prevented? 

Agree those situations don’t come up much. I was just trying to be responsive to your question. But I also think mandatory buybacks would stop fewer school/workplace shootings than a more sensible control regime including the kind of tax stamp required for problematic guns you see with sawed off shotguns, and a robust approach to mental health.

 

Well, that and the fact that regulation and improved community response is politically viable and a mandatory buyback probably isn’t.

[Bozo_Casanova 26508]

I agree we should model the French and get our firearms per 100,000 citizens down from 120 to 19.6.
I don't think that's what you had in mind though.

Why would I mind? The high rate of ownership in the states is artificially inflated by fetishists anyway.

[Anastasis 23479]

5 hours ago, Pig Bellmont said:

Sorry your toy is fun to shoot and good only for killing hordes of pigs or hordes of meth addled trespassers - oh and school children. But as a society we have to decide that the correlation to every mass shooting is too strong and the weapon has to go. 

The strongest correlation between weapon type and mass shootings would be the semi-automatic handgun by a pretty wide margin. 

[Biff Tannen 45694]

Re: the mental health argument. I would say the Internet has made it a bit moot. Kinda hard to do anything about these incels that are radicalized in their parent's basement with no supervision. You have to want to get help and I don't think this environment is very conducive for that. 

[Hugo Stiglitz 23367]

8 minutes ago, Anastasis said:

The strongest correlation between weapon type and mass shootings would be the semi-automatic handgun by a pretty wide margin. 

While your statement might be factually accurate, it deteriorates when it comes to intensity of the mass shootings.

Top 20 mass shooting over the last 20 years in the United States by body count:  I’d speculate at least 90% of them are AR-15/AK-47 style semiautomatic rifles.  The Virginia Tech shooter is the most obvious outlier. 

So while we may still have plenty of mass shooting with handguns, the body count would likely be reduced if access to assault weapons is restricted.

[Anastasis 23479]

20 minutes ago, Hugo Stiglitz said:

Top 20 mass shooting over the last 20 years in the United States by body count:  I’d speculate at least 90% of them are AR-15/AK-47 style semiautomatic rifles.  The Virginia Tech shooter is the most obvious outlier. 

So I pulled down the top 22  shootings from the motherjones mass shootings data (1982-2019) and dropped them into a pivot table after a little bit of recoding of the weapon type to collapse the categories in a manageable number.  The categories I used were: handgun(s) only, rifle(s) only, shotgun(s) only, and four mixed categories (rifle+handgun, rifle+shotgun, hangun+shotgun, rifle+handgun+shotgun). The most frequently used weapon type when a single weapon type was used was a semiautomatic handgun (N=7). When you consider that multiple weapon types could be used (rifle+handgun, rifle+shotgun, etc.), handguns show up in N=18 mass shootings, rifles in N=12. I used N=22 cause there were three shootings with 10 or more casualties.  If I drop those and just do the top 19 as a sensitivity check, handguns still predominate as the most frequent single category (N=7), and the most frequent weapon type when considering use of multiple weapon types (N=15).

Edited September 24, 2019 by Anastasis

[Hugo Stiglitz 23367]

So you chose a period of time during the assault weapons ban?  No way that would skew the numbers.

[Anastasis 23479]

Another sensitivity check, @ top 30 pattern still holds.

	Row Labels Count of case handgun 11 mixed - rifle, handgun 8 semiautomatic rifle 5 mixed - rifle, shotgun, handgun 3 mixed - shotgun, handgun 2 shotgun 1 Grand Total 30

[Anastasis 23479]

1 minute ago, Hugo Stiglitz said:

So you chose a period of time during the assault weapons ban?  No way that would skew the numbers.

I used the full range of the MJ data, which includes data both pre- and during the AWB.  But as another sensitivity check I ran the numbers for 2004-2019. Handguns still come out on top 16 to 12.  

Row Labels	Count of case
handgun	7
mixed - rifle, handgun	6
semiautomatic rifle	5
mixed - shotgun, handgun	2
shotgun	1
mixed - rifle, shotgun, handgun	1
Grand Total	22

[F250 28791]

Trumpism is borderline fascism and ICE acts like the Gestapo. Trump has pardoned Sheriff Arpaio for fucking civil rights violations and Navy Seals who committed war crimes were supported by the President of the United States while the prosecution was attacked by POTUS. Children are locked in cages and American citizens are being illegally detained for being brown. The President of the United States has removed the legal restrictions placed on the Arms of the State. Not to mention the POTUS provides support and legitimacy to fascist and white supremacist groups that commit acts of violence. He also instructs his DOJ to look the other way when it comes to these domestic terrorists.

I own assault rifles, shotguns, rifles and handguns, there is no fucking way I am handing over my weapons. Long live the Second Amendment. Beto's White Privileged Ass can fuck off.

 

[Anastasis 23479]

One last data point.  Frequency count on the field "Prior Signs of Mental Health Issues".

Row Labels	Count of case
Yes	59
Unclear	34
TBD	5
No	17
Grand Total	115

[Smax 3800]

6 hours ago, Pig Bellmont said:

Sorry your toy is fun to shoot and good only for killing hordes of pigs or hordes of meth addled trespassers - oh and school children. But as a society we have to decide that the correlation to every mass shooting is too strong and the weapon has to go. 

 

sorry you think its fun to throw out insults and talk down to people but we as society ( err crazy leftist who are triggered easily) have decided that is hate speech or you disagree with us you are a racist and must be shouted and shut down immediately 

 

Thats the first two amendments of the bill of rights the left is trying to dismantle, how many more do you all want to eliminate?

[Pig Bellmont 8719]

2 minutes ago, Smax said:

 

sorry you think its fun to throw out insults and talk down to people but we as society ( err crazy leftist who are triggered easily) have decided that is hate speech or you disagree with us you are a racist and must be shouted and shut down immediately 

 

Thats the first two amendments of the bill of rights the left is trying to dismantle, how many more do you all want to eliminate?

🙄 tell me more about constitutional rights and how they are never limited by the courts or statutes. They are always absolute and free from all regulation 

[Brisketexan 105164]

 
sorry you think its fun to throw out insults and talk down to people but we as society ( err crazy leftist who are triggered easily) have decided that is hate speech or you disagree with us you are a racist and must be shouted and shut down immediately 
 
Thats the first two amendments of the bill of rights the left is trying to dismantle, how many more do you all want to eliminate?

Wait....a trumpkin is saying the first amendment is being attacked....by....”the left?” That....is fucking rich, from a fanboi of a man who has actually cooperated with foreign regimes to see members of the press arrested (remember that part of the first amendment?). Awesome.
This has to be a bit, right?

[Smax 3800]

Look throughout history, every tyrannical socialist regime starts this way

[Smax 3800]

Just now, Brisketexan said:

Wait....a trumpkin is saying the first amendment is being attacked....by....”the left?” That....is fucking rich, from a fanboi of a man who has actually cooperated with foreign regimes to see members of the press arrested (remember that part of the first amendment?). Awesome.
This has to be a bit, right?

 

 

Brisket remove your head from your ass and go fuck yourself

[fattyflattie 4639]

Smax has best avatar on thread. I’ll allow it. 

[BLKNSTY 289]

 
sorry you think its fun to throw out insults and talk down to people but we as society ( err crazy leftist who are triggered easily) have decided that is hate speech or you disagree with us you are a racist and must be shouted and shut down immediately 
 
Thats the first two amendments of the bill of rights the left is trying to dismantle, how many more do you all want to eliminate?

exhibit a: things racists say...

[Loco 3753]

 

10 hours ago, Anastasis said:

The strongest correlation between weapon type and mass shootings would be the semi-automatic handgun by a pretty wide margin. 

 

10 hours ago, Hugo Stiglitz said:

While your statement might be factually accurate, it deteriorates when it comes to intensity of the mass shootings.

Top 20 mass shooting over the last 20 years in the United States by body count:  I’d speculate at least 90% of them are AR-15/AK-47 style semiautomatic rifles.  The Virginia Tech shooter is the most obvious outlier. 

So while we may still have plenty of mass shooting with handguns, the body count would likely be reduced if access to assault weapons is restricted.

There is another factor to consider ...  lethality   Highspeed rifle rounds deal MUCH more damage than pistol rounds.  That energy combined with easier weapon management make even the wimpiest internet badass a serious threat.  The pistols only advantage is conceal-ability (diminutive size)

 

From a Parkland shooting 

Quote

As I opened the CT scan last week to read the next case, I was baffled. The history simply read “gunshot wound.” I have been a radiologist in one of the busiest trauma centers in the United States for 13 years, and have diagnosed thousands of handgun injuries to the brain, lung, liver, spleen, bowel, and other vital organs. I thought that I knew all that I needed to know about gunshot wounds, but the specific pattern of injury on my computer screen was one that I had seen only once before.

 

In a typical handgun injury, which I diagnose almost daily, a bullet leaves a laceration through an organ such as the liver. To a radiologist, it appears as a linear, thin, gray bullet track through the organ. There may be bleeding and some bullet fragments.

I was looking at a CT scan of one of the mass-shooting victims from Marjory Stoneman Douglas High School, who had been brought to the trauma center during my call shift. The organ looked like an overripe melon smashed by a sledgehammer, and was bleeding extensively. How could a gunshot wound have caused this much damage?

The reaction in the emergency room was the same. One of the trauma surgeons opened a young victim in the operating room, and found only shreds of the organ that had been hit by a bullet from an AR-15, a semiautomatic rifle that delivers a devastatingly lethal, high-velocity bullet to the victim. Nothing was left to repair—and utterly, devastatingly, nothing could be done to fix the problem. The injury was fatal.

https://www.theatlantic.com/politics/archive/2018/02/what-i-saw-treating-the-victims-from-parkland-should-change-the-debate-on-guns/553937/

[SquishMitten 10421]

15 hours ago, Hugo Stiglitz said:

While your statement might be factually accurate, it deteriorates when it comes to intensity of the mass shootings.

Top 20 mass shooting over the last 20 years in the United States by body count:  I’d speculate at least 90% of them are AR-15/AK-47 style semiautomatic rifles.  The Virginia Tech shooter is the most obvious outlier. 

So while we may still have plenty of mass shooting with handguns, the body count would likely be reduced if access to assault weapons is restricted.

 

14 hours ago, Anastasis said:

So I pulled down the top 22  shootings from the motherjones mass shootings data (1982-2019) and dropped them into a pivot table after a little bit of recoding of the weapon type to collapse the categories in a manageable number.  The categories I used were: handgun(s) only, rifle(s) only, shotgun(s) only, and four mixed categories (rifle+handgun, rifle+shotgun, hangun+shotgun, rifle+handgun+shotgun). The most frequently used weapon type when a single weapon type was used was a semiautomatic handgun (N=7). When you consider that multiple weapon types could be used (rifle+handgun, rifle+shotgun, etc.), handguns show up in N=18 mass shootings, rifles in N=12. I used N=22 cause there were three shootings with 10 or more casualties.  If I drop those and just do the top 19 as a sensitivity check, handguns still predominate as the most frequent single category (N=7), and the most frequent weapon type when considering use of multiple weapon types (N=15).

From your source, based on the top 20 shootings since 2004:

Shootings WITHOUT an assault-style rifle: 10 events, 133 fatalities, 112 injuries.

Shootings WITH an assault-style rifle: 10 events, 245 fatalities, 765 injuries.

Admittedly, that is a higher number of fatalities than I would've expected for the non-assault rifle shootings, but many of the weapons used in those included after-market shit that also serves no purpose other than fun (extended magazines, suppressors, etc). Either way, since this happened to work out to be 10 events of each type, we'll use that as a baseline for some rough math. If every shooting that included an assault weapon was instead carried out with handguns/shotguns, there would have been 112 fewer deaths, and 653 fewer injuries. 

The existence of these guns results in almost doubling the death toll, and increasing the number injured by 700%

Edited September 24, 2019 by SquishMitten

[Anastasis 23479]

2 minutes ago, SquishMitten said:

If every shooting that included an assault weapon was instead carried out with handguns/shotguns, there would have been 112 fewer deaths, and 653 fewer injuries. 

Your recategorization of the weapon type is a reasonable approach. But your projection overestimates the impact of an AW because your projection in based on a selection from the extreme end of a highly skewed distribution. LV in particular skews the data, but also Orlando. Both events involved a AW, but I think that it is fair to say that there were other factors that influenced the fatality rate in those cases, and in the LV shooting in particular (bump stock, number of weapons involved, etc.). Same basic reason that the handgun rates are higher than your expectation (VT and Lubys skew the HG data). 

To run your projection, I think that it is better to look at the fatalities across the full sample set. Also, the average is probably not the best measure, but it's what the pivot table I have puts out. I recoded using your classification of AW/no AW and using the full data set generated averages both with and without the two extreme outliers (LV, Orlando).  I get a mean difference ranging from ~1-3 per case using this approach, versus the ~11 increment that you get. BTW the difference in the distributions (full data set, including both LV and Orlando, AW vs. No AW, Mann Whitney U test) are not statistically significant.

[BradInATX 9208]

30 minutes ago, Anastasis said:

Your recategorization of the weapon type is a reasonable approach. But your projection overestimates the impact of an AW because your projection in based on a selection from the extreme end of a highly skewed distribution. LV in particular skews the data, but also Orlando. Both events involved a AW, but I think that it is fair to say that there were other factors that influenced the fatality rate in those cases, and in the LV shooting in particular (bump stock, number of weapons involved, etc.). Same basic reason that the handgun rates are higher than your expectation (VT and Lubys skew the HG data). 

To run your projection, I think that it is better to look at the fatalities across the full sample set. Also, the average is probably not the best measure, but it's what the pivot table I have puts out. I recoded using your classification of AW/no AW and using the full data set generated averages both with and without the two extreme outliers (LV, Orlando).  I get a mean difference ranging from ~1-3 per case using this approach, versus the ~11 increment that you get. BTW the difference in the distributions (full data set, including both LV and Orlando, AW vs. No AW, Mann Whitney U test) are not statistically significant.

 

"Well, if you remove the most deadly mass shooting in history, that was performed with an AR, then it sure makes the AR numbers look better!"

But of course he certainly could have killed 50+ people from hundreds of yards away out of a hotel room window with a Glock 19, I'm sure.

Edited September 24, 2019 by BradInATX

[SquishMitten 10421]

37 minutes ago, Anastasis said:

Your recategorization of the weapon type is a reasonable approach. But your projection overestimates the impact of an AW because your projection in based on a selection from the extreme end of a highly skewed distribution. LV in particular skews the data, but also Orlando. Both events involved a AW, but I think that it is fair to say that there were other factors that influenced the fatality rate in those cases, and in the LV shooting in particular (bump stock, number of weapons involved, etc.). Same basic reason that the handgun rates are higher than your expectation (VT and Lubys skew the HG data). 

To run your projection, I think that it is better to look at the fatalities across the full sample set. Also, the average is probably not the best measure, but it's what the pivot table I have puts out. I recoded using your classification of AW/no AW and using the full data set generated averages both with and without the two extreme outliers (LV, Orlando).  I get a mean difference ranging from ~1-3 per case using this approach, versus the ~11 increment that you get. BTW the difference in the distributions (full data set, including both LV and Orlando, AW vs. No AW, Mann Whitney U test) are not statistically significant.

I don't think it's reasonable to remove those two as "outliers" since they're a fundamental point of this conversation. The AW makes those types of high-casualty events possible. So, including LV and Orlando and recategorizing as I previously did, there would have been 144 fewer deaths and 890 fewer injuries. 

And that is based on the assumption each of those attacks would've been carried out with lesser weapons though. That's likely a flawed assumption. A lot of those shooters wanted the big event for whatever perceived glory they imagined. Many are cowards (I mean, they're all cowards, but many of them probably wouldn't have had the balls to carry through with a face-to-face attack), and they wouldn't have done anything at all.

Regardless, we can somewhat quantify the additional destruction caused by the existence of these weapons, and the exponential impact large scale attacks have on those in the community, the families, and everyone around the country wondering if their town will be next. It's irrefutable that we are at a higher risk due to their existence. Yet we choose to keep our toys, and we try to come up with arguments for doing so.

I don't WANT to give mine up. But I know that we all SHOULD. Not gonna throw it in the garbage to prove a point because that bitch was expensive, but I'll be happy to line up if Beto wants to come buy it from me.

[fattyflattie 4639]

51 minutes ago, SquishMitten said:

Not gonna throw it in the garbage to prove a point because that bitch was expensive, but I'll be happy to line up if Beto wants to come buy it from me.

So you care, but not $1500 care?  I don’t like buy backs, but voluntary ones could help, maybe. It would be harder for those that don’t sell back to get ammo etc., and that might deter some nutbag from shooting up the place.  Pricier for everyone else, anyhow. 

[zork 629]

rut roh raggy.

[Anastasis 23479]

 

14 minutes ago, SquishMitten said:

I don't think it's reasonable to remove those two as "outliers" since they're a fundamental point of this conversation. The AW makes those types of high-casualty events possible. So, including LV and Orlando and recategorizing as I previously did, there would have been 144 fewer deaths and 890 fewer injuries. 

They may be a fundamental point of the conversation, but they result in your original projections overestimating the incremental difference in fatalities and injuries per shooting. Your new number of 144 incremental deaths over 36+ years is probably a better forward looking projection than 112 over 10 events, although I would argue we are still working with some real rough estimates with a large margin for error. 

As far as trimming the data, I do think that LV in particular is an outlier, both in terms of the victim count and the type of weapons being used. The use of the bump stock married to a 100 rd capacity magazine dumping into a crowded outdoor event was a major factor responsible for the extreme number of victims involved. Well, that and the guy had 20+ rifles to cycle. The total victim count is an extreme outlier (604 total victims vs. ~15 on average for all other mass shootings vs. ~44 in the rest of the top 20). The Orlando shooting is less clearly an outlier, both in terms of the victim count and the type of weapons cored to the discussion.

   

[BradInATX 9208]

32 minutes ago, Anastasis said:

 

They may be a fundamental point of the conversation, but they result in your original projections overestimating the incremental difference in fatalities and injuries per shooting. Your new number of 144 incremental deaths over 36+ years is probably a better forward looking projection than 112 over 10 events, although I would argue we are still working with some real rough estimates with a large margin for error. 

As far as trimming the data, I do think that LV in particular is an outlier, both in terms of the victim count and the type of weapons being used. The use of the bump stock married to a 100 rd capacity magazine dumping into a crowded outdoor event was a major factor responsible for the extreme number of victims involved. Well, that and the guy had 20+ rifles to cycle. The total victim count is an extreme outlier (604 total victims vs. ~15 on average for all other mass shootings vs. ~44 in the rest of the top 20). The Orlando shooting is less clearly an outlier, both in terms of the victim count and the type of weapons cored to the discussion.

   

 

This is up there with some of the rationalizations that aggies throw out when they lose. This is horrible. You don't just get to call something an outlier because it doesn't support the point you're making. That's not how things work.

Edited September 24, 2019 by BradInATX

[Anastasis 23479]

13 minutes ago, BradInATX said:

This is up there with some of the rationalizations that aggies throw out when they lose. This is horrible. You don't just get to call something an outlier because it doesn't support the point you're making. That's not how things work.

I provided my rationale for flagging it as an outlier. The rationale is sound statistically and conceptually.  But since you don't trust my intuition, I also dropped the dataset into an online calculator (https://www.graphpad.com/quickcalcs/grubbs2/) which also supports the obvious conclusion that the LV shooting datapoint is a statistical outlier.

And beyond that, even if you don't exclude it, the points that I have made in response to Hugo's claim and Squish's original overestimation of the incremental fatalities/victims attributable to AWs are still correct and valid. 

[BradInATX 9208]

51 minutes ago, Anastasis said:

I provided my rationale for flagging it as an outlier. The rationale is sound statistically and conceptually.  But since you don't trust my intuition, I also dropped the dataset into an online calculator (https://www.graphpad.com/quickcalcs/grubbs2/) which also supports the obvious conclusion that the LV shooting datapoint is a statistical outlier.

heh. That's not how these things work. You're cherry picking just as someone who isn't versed in stats typically will do to prove a biased point .

1) That outlier calculator assumes a normal distribution, which your sample is not.
2) You cherry picked "injuries" as your test, because of course you did. The reason 600 were injured was because of the stampede, not the shooter. The fatalities, while the highest of any mass shooting, were well within the normal distribution. So your experiment was flawed. There is absolutely no reason to think this is an outlier, and of course anyone with Stats 101 level knowledge would not treat it as such. You should have used fatalities, not injuries. And in a professional setting, we'd remove fatalities that weren't a direct result of the gun itself, though there's no way to do that here.
3) Your sample is bad. You used shootings that happened during the ban. Your "study" completely fails to compensate for trends. There is very clearly a shift towards AR style rifles and semi-auto rifles as they become trendy and as they're seen as weapon of choice in other shootings.

 

Here is your analysis when you actually look at recent history since it's become plainly obvious that the AR style rifle is becoming trendy and people are copycatting it in these shootings. From 2010 to the end of your data set:

 

And here are trend lines of fatalities by weapon type since 1990. Do you want to guess what the red and yellow lines are? Your options are handgun, semi-auto rifle, mixed without semi-auto rifle, and mixed with semi-auto rifle.

 

I'm sure you're very good at your day job. I'd focus there.

 

Edited September 24, 2019 by BradInATX

[Bama Chick 28596]

Well after that thorough pantsing by Brad (high five!) - back to Beto in the Beto thread.

This is the overflow for his town hall at Ohio State this afternoon.

He spent the morning with striking GM workers and calling for impeachment.

[Anastasis 23479]

53 minutes ago, BradInATX said:

1) That outlier calculator assumes a normal distribution, which your sample is not.

I log transformed the data before I ran the outlier test, Brad.

53 minutes ago, BradInATX said:

2) You cherry picked "injuries" as your test, because of course you did. The reason 600 were injured was because of the stampede, not the shooter. The fatalities, while the highest of any mass shooting, were well within the normal distribution. So your test was flawed and of course biased. There is absolutely no reason to think this is an outlier, and of course anyone with Stats 101 level knowledge would not do so.

I ran my original outlier test above on "total victims".  But you get the same results when you run the transformed count of total fatalities.  The LV shooting is a statistical outlier based on all three outcomes (fatalities, injuries, total victims).  When you remove the LV observation from the data, and rerun on fatalities, the Orlando shooting is an outlier. When you drop the Orlando shooting and rerun, there are not more outliers identified using that tool I linked.

53 minutes ago, BradInATX said:

3) Your sample is dubious and terrible. You used shootings that happened during the ban. Your "study" completely fails to compensate for trends. There is very clearly a shift towards AR style rifles and semi-auto rifles as they become trendy and as they're seen as weapon of choice in other shootings.

My "sample" is the most comprehensive mass shooting dataset that I have access to.  I didn't do a lot of digging, but if you have another one to use, go ahead and point to it. I ran a couple of different cuts using the full dataset and restricting to post-AWB earlier.

I have not conducted any "study" here.  I have simply addressed the comment that Hugo made that 90% of the top 20 mass shootings involve AR/AKs.  This is not an accurate statement, whether you use the full dataset of the post AWB period. Not close. Further, I pointed out why Squish's projection of the incremental fatalities and injuries associated attributable to AWs was overestimated.  These are ad-hoc analyses to address those points, not a study by any sense.

53 minutes ago, BradInATX said:

From 2010 to the end of your data set:

 

2 hours ago, BradInATX said:

This is up there with some of the rationalizations that aggies throw out when they lose.

 

Speaking of aggy-like cherry picking of data, this is one of their favorite tactics.  Picking an arbitrary time frame to massage numbers to drive a pre-determined dogma. BTW, I can't replicate your numbers. There is no sequential combination that gets me to 291 starting from most recent data.  Maybe you are filtering on something or leaving categories out? As far as the trend lines, I'd rather not just guess as to the interpretation.  That would probably lead to wrongheaded conclusions such as your conclusion that I didn't log transform. 

 

Edited September 24, 2019 by Anastasis

[Anastasis 23479]

2 minutes ago, Bama Chick said:

Well after that thorough pantsing by Brad (high five!) -

[BradInATX 9208]

I think she's telling us to stop nerding up her thread. I think it's topical enough.

[Bama Chick 28596]

(There’s like three different gun threads.)

[Anastasis 23479]

Just now, BradInATX said:

9 minutes ago, Anastasis said:

I think she's telling us to stop nerding up her thread. I think it's topical enough.

Yeah, we are derailing this thread.  Sorry Bama, but Beto made firearms a pivotal aspect of his campaign. Most of the last page and this one could get moved elsewhere.  

[BradInATX 9208]

 

 

 

 

Speaking of aggy-like cherry picking of data, this is one of their favorite tactics.  Picking an arbitrary time frame to massage numbers to drive a pre-determined dogma. BTW, I can't replicate your numbers. There is no sequential combination that gets me to 291 starting from most recent data.  Maybe you are filtering on something or leaving categories out? As far as the trend lines, I'd rather not just guess as to the interpretation.  That would probably lead to wrongheaded conclusions such as your conclusion that I didn't log transform. 

 

 

Why would you log transform a data set that runs from like 5 to 60? Lol.

 

58 is not an outlier of a data set where the average is in the 20s and the 2nd highest number is in the high 40s. C'mon now.

 

I picked 2010 because it's this decade. And it's about where I estimate ARs became the cool kid mass shooting tool. Was just a random stab. Either way, those trend lines don't lie.

 

I don't know how you segmented your data. I had to do the same thing you did, classify the free text weapon field into larger groupings. We probably did it differently for a couple.

 

Regardless there's clearly a trend no matter how you want to slice it. But yeah, tag me on a gun thread if you want to continue. Either way

 

[Anastasis 23479]

1 hour ago, BradInATX said:

Why would you log transform a data set that runs from like 5 to 60? Lol.

A: The data point is an outlier based on this statistical test.

B: That test assumes a normal distribution, the data are not (they are lognormal).

A: I log transformed the data before I ran the test.

B: Why would you do that, look at the range?

A:

 

The range is less important than the distribution.  The data are lognormal. I transformed to bring the data more in line with the distributional assumption underlying the test.  This is very cool and very standard practice. 

 

1 hour ago, BradInATX said:

58 is not an outlier of a data set where the average is in the 20s and the 2nd highest number is in the high 40s. C'mon now.

It's also an outlier using the IQR method; however, it flags the top 9 events as outliers, so I don't think it is too limiting for these data.  

 

1 hour ago, BradInATX said:

Regardless there's clearly a trend no matter how you want to slice it.

I don't think that I have argued otherwise. I actually consider gun violence in general, and mass shootings in particular a public health issue.  I think that multi-modal research should be carried out to try to try to increase our understanding the many factors associated with the trends we are seeing. Remember, when had an exchange on that in another thread related to DHS research related to gun violence. You went all gestapo in bastropo on me. 

 

@hayden_horn, if you think its appropriate to declutter the Beto thread, could you move this exchange to the following thread or one of the others ones where it may be a better fit:

 

 

Edited September 25, 2019 by Anastasis

[zork 629]

How has the Beto gun talk affected his polling?

[Bama Chick 28596]

Well it’s increased his crowd turnout.

[Bama Chick 28596]

That’s a lot of fucking Ohioans.

[Post Oak 20254]

Thoughts and prayers for those couches

[Bama Chick 28596]

Little dick spotted at Kent State today.

[Hugo Stiglitz 23367]

I still like Beto, a lot. 

[bad_teammate 13219]

All the Dem candidates will spend the next 5+ months saying, "No it's totally cool that Joe's son made $50k/week on the board of a Ukrainian oil company while Joe was VP and deeply involved in internal Ukraine politics that is good and proper. Trump, though? So corrupt. Everyone is saying it more and more."

You love to see it.

[Grade of D as in David 13253]

BT is going to be insufferable when Bernie gets knocked out of the primary again. 

[bad_teammate 13219]

53 minutes ago, Grade of D as in David said:

BT is going to be insufferable when Bernie gets knocked out of the primary again. 

Joke's on you, I'm already insufferable.

[RPM 26327]

Happy Birthday Beto!

[TexLonghorn 5671]

 

[zork 629]