Marshall, Sanow and statistical analysis

[Robin]

Doc, I've heard you speak favorably of Marshall's "one shot stop study" a couple times. I invite you to read http://www.firearmstactical.com/marshall-sanow-statistical-analysis.htm and post your thoughts regarding that article.

[tscentral]

Robin,

Word of warning, some people are going to start thinking you are a Troll for bring the Morgue Monsters vs Jello Junkies debate here.

A little critical analysis of the article you mentioned. Anytime a supposed scientific or statistically based analysis uses words like bad joke or gagged in its text and puts quotes around data and data base of the subject, it becomes suspect. There is more than analysis going on. This piece is also over decade old, so its age reflects badly on it, as well.

For one of the best write ups on this subject. Get yourself a copy of Massad Ayoob's The Gun Digest of Combat Handgunnery. He has a chapter on ammo selection that covers a real world analysis of this. Better written and newer.

Hunter

[Robin]

Hunter, this was not intended as a troll, nor a debate about ballistic gelatin vs. actual shootings. I also do not think the age of the article especially relevant since it refers to studies and methodologies that are presumably still in use by Marshall.

The article states that Marshall's data is suspect because of statistical improbabilities within the dataset itself. It says nothing about the morgue/jello debate. Given Doc's profession as a research chemist w/a PhD, it seemed appropriate to ask his thoughts regarding the validity of the article's claim. I took several statistical analysis classes back in college but my eyes glaze over when I read about it now.

[Doc Wesson]

Robin,
First let me clarify.  I am not a staunch advocate of stopping power numbers.  In the limited amount of statistical data "pooling" we do not have much to go by in the ballistic world.  What I mean by this is that there are multiple calibers, multiple weights, unlimited distances, unlimited angles of incidence, unlimited placements (such as shooting center mass vs. a shoulder), etc...
So if you consider each one of those parameters a "pool" they become unlimited.  In the article you posted, they used permutations.  So lets look at permutations based on the pooling.  It becomes exponential and astronomical.  In my work (my real job), I shoot ballistic panels of various makeup all the time.  I can limit my data to a "pool."  Same angle of incidence every time, same bullet speed, same caliber, same weight, same relative humidity, same temperature, same pressure.  With this method, I get a tiny standard deviation and a very narrow confidence level for the data I am trying to collect.  Even then, I still get data scatter.  All that control, and I still get scatter.  Why would that be?  Because you cannot control everything and make everything the same.  Maybe there was a very slight variation in the tensile of the metal jacket, maybe there was a microscopic variation in the surface area of the projectile, maybe the crimp was just a hair more on this bullet to the last, maybe there was a molecular variation in the kevlar in the area i shot on the panel.. it is impossible to know.  However, I can get a scientific average and a high confidence based on the "pool" I am in control of. In other words, the results tell me something.  As for Sanow and Marshall testing... it is just a broad spec to go by... a tool.. touted to be based on "real world stuff."  Maybe so... but that is a "pool" that is the size of an ocean, is that large a pool of data worthy of consideration, "we report, you decide"... So we relegate to other numerical means.. such as instrumental energy or knock down power and the like.  Great.. so the bigger and faster and heavier the better right?  Big numbers... not necessarily, as we all know.. otherwise we would all be walking around with .44mags... or whatever weight handgun we could stand on us all day..
At the end of the day... I am for shot placement.  Do I like .357 mag and .45 acp and .40 over a .38 or 9mm.  Yes I do, based on my own presumptions and concepts and testing.  But those calibers and speeds not placed are nothing compared to a well placed shot of .38 special.  
Ultimately, it is better to have something than nothing.  So we use the data we have been given and take it for what it is....
Hunter gave good advice in my opinion.  Check out Mass' writings.. he covers the real world view in one of the best ways I can find.  Not just saying that because he is a GRRNer... he is a pro and a great student of the topic.  As are the other ProArms folks... I particularly liked what they discuss on wound characteristics when shooting hogs.  It gives you a good look at what these tools do.  Does it mean that the 115grain +P+ 9mm is a death ray because it created a massive wound channel and liquified lots of tissue?  No.  But it does give us something to talk about.
As for that article specifically... it is fodder.  I add the info I glean from it to the mound I get from everything else.  Take out and keep the parts I find merit in and relegate the rest, based on my own thoughts.  Just like I do with "stopping power" data and Taylor Knock Down numbers and lb-ft of energy... just me thinking out loud.. gather your info, make your decisions, follow your own path... post them here... we can add it to our "database."
Think I need to go lay down now.. that was the longest post I have done on the forum..LOL!

[NavyBlue99TA]

Think I need to go lay down now.. that was the longest post I have done on the forum..LOL!

And one of the best, well reasoned posts I've ever read.

By the way, you must have the coolest job ever.   

[tscentral]

Robin,

I believe your intend was not to troll or bring the debate up. But any reference to Marshall and Fackler will have the unintended consequent of doing just that. Kind of an intellectual equivalent of knowing your back stop, worry about where all your bullets are going, and where all your words may lead.

Hunter

[Robin]

As for that article specifically... it is fodder.  I add the info I glean from it to the mound I get from everything else.  Take out and keep the parts I find merit in and relegate the rest, based on my own thoughts.

Thanks for the long response. I would be very interested in learning which parts you considered with/without merit and why. Knowing that would help me (and possibly others) when evaluating other material.

[RichardInFlorida]

I know I am a bit late in posting to this topic, but if you want to really evaluate Marshall's work and his opinions, ask him yourself.  His forum is at http://www.stoppingpower.net/  Anyone asking honest questions (and keeping a civil tone) is welcome there.

Pros like Dave Spaulding, Tom Givens, and Jim Higginbotham all are regular members, so you can get a wide range of opinions on 'stopping power' if you are interested.

[BikerRN]

Forgive me for replying to an old topic that many may have thought had been laid to rest, but in case anyone does a search and comes up with this thread I felt some more information was in order.

I'm a member of the Stopping Power Forum and have found Evan Marshall to be an honest and upright individual that I would be honored to be in the presence of but I certainly don't speak for him or anyone other than myself. The caliber of people there are first rate, like many here.

With all that said, Evan never intended, from what I know in my reading of his work and post at his forum, for the stopping power evaluations and printings to be a tactical methodology, but rather as a way of providing some way of measuring what the results were, at a time when not many were attempting to do so.

It was in a sense part of the process and evolution, if you will, at measuring how a bullet performs using human test media. Jello, and the F.B.I. test protocols have expanded on that, but neither is the "be all end all" of a very difficult to measure subject matter.

We can find failures and sucesses with any caliber, even rifles. Some calibers and loads just fail less often than others. There ain't no magic bullet folks. You have to put the little pill where it will do the most good. Even with the different testing methodology you will consistantly see certain calibers coming out near, if not at, the top no matter what the testing protocols are.

Instead of discrediting Evan Marshall and Ed Sanow I think the defensive handgun world owes them a debt of gratitude. They have provided one piece of a very complex puzzle, as has Thompson, Massad Ayoob, Fackler, and many others. It all needs to be compiled and then one needs to take all of the results in to consideration when making their caliber and load selection.

That's my $0.02 on the matter.

Biker

[Devereaux]

OohRah, Biker! 1+1! Doc gave a great scientific answer, but you have responded to a more important issue - the character of honourable men, trying to shed light. The old adage, No Good Deed Goes Unpunished, is often in play "in the real world". Thanks to guys like you, we get reminded of what is truly important - not a bunch of numbers of theories, but real men.