Is a revolver enough?

[Bolster]

How many data points are needed per round? 10?, 1,000?, 10,000?

I'm guessing you mean 'caliber' rather than 'round'? The best way to determine required n would be with a power calculation. Absent that, we can default to common rules of thumb, such as 'no cell of interest lower than 50' or 'overall minimum n around 200-400, and more is better.' Ellifritz' dataset is a good one for size, his overall n is 1785. There are low cells for .32 and for .44, but that would just caution us to read the .32 and .44 data with skepticism.

What this dataset needs, IMO, is a regression analysis that could tell us the relative importance of caliber (and weight, speed, and/or other variables if available). As is, the results are presented as categorical data, which is fine as far as it goes, but analyzing it as continuous data in a regression would give a lot of additional insights.

Those of you who read Larrin Thomas’ work will recall an analogous regression formula that gave us the relative importance (for TCC) of hardness, edge angle, and various carbides. A similar analysis for the Ellifritz dataset is what I’m suggesting.

[Doc Dan]

The data in that “An Alternate Look at Handgun Stopping Power” article from 2011 is kind of meaningless because it has no accounting for actual cartridges used, the firearm, distance and placement of those shots..

I get your point, but personally I'd not go so far as to say meaningless. There are a number of relevant variables not accounted for (though shot placement was considered), in which case, research relies on the law of averages to arrive at approximations across the unexamined variables. Relying on the law of averages is problematic when the n is low, as it was for the .32 data, for example. Of course, that's not the author's fault, it's just that shootings are a "low-n" sort of phenomenon to study, they're relatively rare events where only partial information may be available.

There's also a problem with accounting for all relevant variables, in the cross-tab-ish analysis used by the article...it would effectively lowering the n of each unique cell, requiring a much, much higher overall n. What I would do, if this were my research, is move away from cross-tabs style to a regression style of analysis. I think that would make the results more interpretable, and it would also give us a statistic (the R-squared) regarding how "trustworthy" the analysis is.

Does anyone know of a stopping-power study that uses a regression analysis?

You are thinking too much.

The one that is taken off real world shootings is more than good enough to get a realistic idea.

Unless one wants to line people up and execute them using the various ammo and guns.

How many data points are needed per round? 10?, 1,000?, 10,000?

That is basically what happened in the "Strasbourg Test". Instead of people they used large closer to human sized goats. They shot them in various places and noted the effect. The goats were plugged into all sorts of gizmos like EEG, pressure monitors, and other things measuring vitals when shot.

[Ankerson]

The data in that “An Alternate Look at Handgun Stopping Power” article from 2011 is kind of meaningless because it has no accounting for actual cartridges used, the firearm, distance and placement of those shots..

I get your point, but personally I'd not go so far as to say meaningless. There are a number of relevant variables not accounted for (though shot placement was considered), in which case, research relies on the law of averages to arrive at approximations across the unexamined variables. Relying on the law of averages is problematic when the n is low, as it was for the .32 data, for example. Of course, that's not the author's fault, it's just that shootings are a "low-n" sort of phenomenon to study, they're relatively rare events where only partial information may be available.

There's also a problem with accounting for all relevant variables, in the cross-tab-ish analysis used by the article...it would effectively lowering the n of each unique cell, requiring a much, much higher overall n. What I would do, if this were my research, is move away from cross-tabs style to a regression style of analysis. I think that would make the results more interpretable, and it would also give us a statistic (the R-squared) regarding how "trustworthy" the analysis is.

Does anyone know of a stopping-power study that uses a regression analysis?

You are thinking too much.

The one that is taken off real world shootings is more than good enough to get a realistic idea.

Unless one wants to line people up and execute them using the various ammo and guns.

How many data points are needed per round? 10?, 1,000?, 10,000?

That is basically what happened in the "Strasbourg Test". Instead of people they used large closer to human sized goats. They shot them in various places and noted the effect. The goats were plugged into all sorts of gizmos like EEG, pressure monitors, and other things measuring vitals when shot.

I would stick with the real shooting data taken from the real shootings in the end as it is real world.

[Ankerson]

How many data points are needed per round? 10?, 1,000?, 10,000?

I'm guessing you mean 'caliber' rather than 'round'? The best way to determine required n would be with a power calculation. Absent that, we can default to common rules of thumb, such as 'no cell of interest lower than 50' or 'overall minimum n around 200-400, and more is better.' Ellifritz' dataset is a good one for size, his overall n is 1785. There are low cells for .32 and for .44, but that would just caution us to read the .32 and .44 data with skepticism.

What this dataset needs, IMO, is a regression analysis that could tell us the relative importance of caliber (and weight, speed, and/or other variables if available). As is, the results are presented as categorical data, which is fine as far as it goes, but analyzing it as continuous data in a regression would give a lot of additional insights.

Those of you who read Larrin Thomas’ work will recall an analogous regression formula that gave us the relative importance (for TCC) of hardness, edge angle, and various carbides. A similar analysis for the Ellifritz dataset is what I’m suggesting.

No, I ment round as I said.

Many different ammo/round types per caliber so.

And like is also said you are thinking way too much, you will never get that type of data off any sort of real world shootings.

All that is really needed is the basics to get a good real world general idea of what works.

Any more than that and we are back to what I said before about executing humans, and a lot of them, that's just not going to happen in the real world.

[RamZar]

If you’re comfortable with tests on ballistic gelatin this link provides tests on many popular self-defense handgun ammo with data on penetration and expansion. It has other advice beyond the ammo so it’s worth reading.

“We fired five rounds of each self-defense load using short-barreled pistols positioned 10 feet from a Clear Ballistics synthetic gelatin block with a four-layer heavy clothing barrier placed in front of it.”

https://www.luckygunner.com/labs/self-d ... onclusions

[Doc Dan]

I get your point, but personally I'd not go so far as to say meaningless. There are a number of relevant variables not accounted for (though shot placement was considered), in which case, research relies on the law of averages to arrive at approximations across the unexamined variables. Relying on the law of averages is problematic when the n is low, as it was for the .32 data, for example. Of course, that's not the author's fault, it's just that shootings are a "low-n" sort of phenomenon to study, they're relatively rare events where only partial information may be available.

There's also a problem with accounting for all relevant variables, in the cross-tab-ish analysis used by the article...it would effectively lowering the n of each unique cell, requiring a much, much higher overall n. What I would do, if this were my research, is move away from cross-tabs style to a regression style of analysis. I think that would make the results more interpretable, and it would also give us a statistic (the R-squared) regarding how "trustworthy" the analysis is.

Does anyone know of a stopping-power study that uses a regression analysis?

You are thinking too much.

The one that is taken off real world shootings is more than good enough to get a realistic idea.

Unless one wants to line people up and execute them using the various ammo and guns.

How many data points are needed per round? 10?, 1,000?, 10,000?

That is basically what happened in the "Strasbourg Test". Instead of people they used large closer to human sized goats. They shot them in various places and noted the effect. The goats were plugged into all sorts of gizmos like EEG, pressure monitors, and other things measuring vitals when shot.

I would stick with the real shooting data taken from the real shootings in the end as it is real world.

These are actual shootings, just not on people. There were goats, deer, and many other things shot in controlled environments so as many variables as possible could be ruled out. They had all sorts of gizmos attached to measure various vitals. They were shot and the results measured. That is pretty good data.

[Ankerson]

You are thinking too much.

The one that is taken off real world shootings is more than good enough to get a realistic idea.

Unless one wants to line people up and execute them using the various ammo and guns.

How many data points are needed per round? 10?, 1,000?, 10,000?

That is basically what happened in the "Strasbourg Test". Instead of people they used large closer to human sized goats. They shot them in various places and noted the effect. The goats were plugged into all sorts of gizmos like EEG, pressure monitors, and other things measuring vitals when shot.

I would stick with the real shooting data taken from the real shootings in the end as it is real world.

These are actual shootings, just not on people. There were goats, deer, and many other things shot in controlled environments so as many variables as possible could be ruled out. They had all sorts of gizmos attached to measure various vitals. They were shot and the results measured. That is pretty good data.

Well, not really.

If one wants data on shooting elephants they wouldn't use squirrels as targets would they?

I would stick with the real shooting data on actual people that is available.

[Doc Dan]

Well, then you wouldn't have empirical data. They used human sized animals and measured the body response. Also, the goats were close to humans in physiology. I don't see how we can do better without using humans. I believe these studies are valuable and when combined with street shootings give a clearer picture (a lot of street shootings used in studies don't show if an arm or leg was hit and the person stopped, nor do they measure those that simply quit after being shot but were not incapacitated. This study gives that information.)

[Ankerson]

Well, then you wouldn't have empirical data. They used human sized animals and measured the body response. Also, the goats were close to humans in physiology. I don't see how we can do better without using humans. I believe these studies are valuable and when combined with street shootings give a clearer picture (a lot of street shootings used in studies don't show if an arm or leg was hit and the person stopped, nor do they measure those that simply quit after being shot but were not incapacitated. This study gives that information.)

Well they were/are real world police Shootings for the most part so it's not like they were willing test subjects.

Humans are easier to kill etc than when compared to other animals as has been shown over time.

Also doesn't take into count on drugs like PCP etc or the ones who give up after a paper cut or something.

Taking into count Police are trained to shoot center mass assuming they hit what they are aiming at.

Not sure how the goats would give up or the goats would know to give up or not, not really quite the same thing as shooting a human.

Even head shots aren't 100% on humans per the real shooting data.

A lot of guess work in the end really.

The only smart thing Law Enforcement ever did was get away from those lead round nose bullets and or round nose FMJ bullets and started using expanding bullets.

[RamZar]

Even head shots aren't 100% on humans per the real shooting data.

True. You have to go for the T which is the area of the eyes to the bottom of the nose.

[Ankerson]

Even head shots aren't 100% on humans per the real shooting data.

True. You have to go for the T which is the area of the eyes to the bottom of the nose.

Yeah, the center line is the kill zone so hitting it as close as possible is best.

[RamZar]

Even head shots aren't 100% on humans per the real shooting data.

True. You have to go for the T which is the area of the eyes to the bottom of the nose.

Yeah, the center line is the kill zone so hitting it as close as possible is best.

Yes. Best to miss below the T than the hard skull above.

[benben]

Even head shots aren't 100% on humans per the real shooting data.

True. You have to go for the T which is the area of the eyes to the bottom of the nose.

Yeah, the center line is the kill zone so hitting it as close as possible is best.

Yes. Best to miss below the T than the hard skull above.

I’ve always read and heard straight through the center of the mouth is the most devastating, you’re dead standing, dead before you hit the ground.

[RamZar]

True. You have to go for the T which is the area of the eyes to the bottom of the nose.

Yeah, the center line is the kill zone so hitting it as close as possible is best.

Yes. Best to miss below the T than the hard skull above.

I’ve always read and heard straight through the center of the mouth is the most devastating, you’re dead standing, dead before you hit the ground.

A bit higher than that with a center line through the nose between the eyes.

[benben]

Yeah, the center line is the kill zone so hitting it as close as possible is best.

Yes. Best to miss below the T than the hard skull above.

I’ve always read and heard straight through the center of the mouth is the most devastating, you’re dead standing, dead before you hit the ground.

A bit higher than that with a center line through the nose between the eyes.

I go to church with a SWAT sniper, he’s talked about the upside down triangle covering the same exact area as your T.

[Ankerson]

True. You have to go for the T which is the area of the eyes to the bottom of the nose.

Yeah, the center line is the kill zone so hitting it as close as possible is best.

Yes. Best to miss below the T than the hard skull above.

I’ve always read and heard straight through the center of the mouth is the most devastating, you’re dead standing, dead before you hit the ground.

Right between the eyes, dead center is the aim point.

Death would be so fast they wouldn't even know they were hit.

[RamZar]

Yeah, the center line is the kill zone so hitting it as close as possible is best.

Yes. Best to miss below the T than the hard skull above.

I’ve always read and heard straight through the center of the mouth is the most devastating, you’re dead standing, dead before you hit the ground.

Right between the eyes, dead center is the aim point.

Death would be so fast they wouldn't even know they were hit.

Aim points for head shots from different angles.

[Ankerson]

Yes. Best to miss below the T than the hard skull above.

I’ve always read and heard straight through the center of the mouth is the most devastating, you’re dead standing, dead before you hit the ground.

Right between the eyes, dead center is the aim point.

Death would be so fast they wouldn't even know they were hit.

Aim points for head shots from different angles.

Like I was saying.

[Doc Dan]

[SpyderEdgeForever]

With revolvers in particular what grips do you all prefer? Synthetic polymers and composites or stuff like wood or stag? The very old ones used Vulcanite which was an early form of treated rubber.