Hand Length to Gun Fit: A Case Study: Analysis and Conclusion (Part 3)

[RJ]

Not a huge deal, but I stumbled on a nasty bug in the spreadsheet involving a circular reference - ugh. I fixed that. I also tweaked the algorithm for computing Gun Sizes (which are now called "Grip Sizes"; I decided I like that a lot better) and regenned the data. Some guns swapped around sizes, but it is pretty subtle so I am not sure there's value in posting it.

I also remeasured the GC of all four of my samples, and I must have had a brain fart with my LCR, because it is 20mm more than I had at first. Not a clue why. Anyway after all that, my LCR is in the M and the other 3 (P365X, P365XL, Garrison with thin grips) are in M/L camp, for Grip Size, which makes me feel a bit better, since my trigger reach on all of them is pretty good.

I'm up to 48 guns on the Part 2, still looking for more of the smaller end if possible, like G43, G43X/48, Hellcat, in particular.

[RJ]

Or maybe I need to rewicker "Gun" into "Grip". That would disconnect the physical size of the gun (e.g. Large, like a 1911) from the ability to reach the trigger (e.g. Small, like a 1911 with thin stocks and a short trigger pad.) From a word choice point of view, perhaps grip is closer to the mark, in terms of initial fit re: hand size to provide adequate trigger reach.

So:

Gun Size --> Grip Size
Gun Table --> Grip Table

Moar hmmm.

Still pondering. This morning I was thinking, what would be the "elevator speech" graphic for this study, so I came up with this (it confirms why I never got into graphical design, for sure).



I did some more tweaking to the spreadsheet, getting closer to feeling comfortable to upload it for view. I straightened out the concept of using boundary transitions for the sizes, computing them consistently. And rewickered all the Hand Sizing and Grip Sizing entries in purple.

I have aspirations to go to the Fancy Pants range in The Villages today. They have a wall o' guns out on display. I think I will take my tape measure and ask the manager if I can measure some.

[RJ]

I was able to obtain grip circumference on seven additional handguns today. This was with the permission of Shooter's World, The Villages, FL, and I greatly appreciate their help.



I am up to 55 samples now, but definitely appreciate any additional measurements anyone would like to add.

[BN]

Still pondering. This morning I was thinking, what would be the "elevator speech" graphic for this study, so I came up with this (it confirms why I never got into graphical design, for sure).



I did some more tweaking to the spreadsheet, getting closer to feeling comfortable to upload it for view. I straightened out the concept of using boundary transitions for the sizes, computing them consistently. And rewickered all the Hand Sizing and Grip Sizing entries in purple.

I have aspirations to go to the Fancy Pants range in The Villages today. They have a wall o' guns out on display. I think I will take my tape measure and ask the manager if I can measure some.

I was looking at your pictures and realized that you are measuring the hand from the wrist to the tip of the middle finger, but the measurement on the pistol was from the web of your hand to the pad of your trigger finger. Will that make a difference in your research?

[RJ]

I was looking at your pictures and realized that you are measuring the hand from the wrist to the tip of the middle finger, but the measurement on the pistol was from the web of your hand to the pad of your trigger finger. Will that make a difference in your research?

Thanks for the comment. The picture is not, really, very accurate; I'm not much of an artist I'm afraid. The pistol is meant to be measured by "Grip Circumference", basically from the web of hand position, around to the trigger face and then back again. The Red Arrow is not a good representation. I was aiming to try and put the whole thing on one piece of paper, using images, so it was as simple as possible. Sorry for the confusion, I'll see if I can work in a better way to picture this concept.

[UNK]

Thanks for the comment. The picture is not, really, very accurate; I'm not much of an artist I'm afraid. The pistol is meant to be measured by "Grip Circumference", basically from the web of hand position, around to the trigger face and then back again. The Red Arrow is not a good representation. I was aiming to try and put the whole thing on one piece of paper, using images, so it was as simple as possible. Sorry for the confusion, I'll see if I can work in a better way to picture this concept.

This is not the level of eork we expect from you Rich. Do better! 😁

[RJ]

First chance to pull some metrics today.



Hmmm.

35% of the population surveyed falls into the "Medium" hand size, by far the biggest number. In contrast, looking at pistols, the largest number with 35% have a "Large" Grip Circumference.

Not sure what to make of this. It has been in the back of my mind, looking at the hand length data, we may have a couple outliers at the top end, which is skewing the data larger than the general population. This would make the overall span of small to large hand sizes broader than it might actually be, and thus skew the Hand Sizes calculated in the Study smaller as a result? Not sure.

One thing I am mulling over is to simply discard one or more of the large Hand Size values, postulating these represent something bigger than the 95% centile (maybe 100% centile? 110% centile?). I just don't think that if you surveyed glove sales across the US, you would not get 35% of them being Medium; I would bet it would be more like 35% Large. I'll have to look into this some more.


The Pistol data makes more sense; 62% are L or XL. Given the demographics of Pistol-Forum, it seems like having duty or full size guns would be common, and the sizing data supports that.

[RJ]

So as a thought experiment, I moved four of the top hand size values out of the "study" population counts. A portion of the Hand Size data table is below with how that looks:



To me what's interesting is that now, the study population and the MIL-STD-1472D population are MUCH closer matched in terms of the 95% centile subjects. I take this to mean the new subset (22 vs. 26) of data more closely resembles gen pop 5 to 95% centile.


Now let's go look at the new metrics, using this data:



I think I can convince myself that the Hand Sizes with this new assumption better reflects typical US Glove Size sales, in that now we are talking 27% are Size L (45% if you lump in M/L), and only 9% are M. I am going to continue to ponder this but I think this makes more sense to me than what I had. Even so, we do seem to have a lot of folks with large hands on this forum.

If anyone is wondering why the grip size percentage changes, it's that the Study Spreadsheet is all computationally driven now, so it recalculates every cell as anything changes. Assumptions remain the same, but all the linear interpolations are automatically regenerated.

[UNK]

So as a thought experiment, I moved four of the top hand size values out of the "study" population counts. A portion of the Hand Size data table is below with how that looks:



To me what's interesting is that now, the study population and the MIL-STD-1472D population are MUCH closer matched in terms of the 95% centile subjects. I take this to mean the new subset (22 vs. 26) of data more closely resembles gen pop 5 to 95% centile.


Now let's go look at the new metrics, using this data:



I think I can convince myself that the Hand Sizes with this new assumption better reflects typical US Glove Size sales, in that now we are talking 27% are Size L (45% if you lump in M/L), and only 9% are M. I am going to continue to ponder this but I think this makes more sense to me than what I had. Even so, we do seem to have a lot of folks with large hands on this forum.

If anyone is wondering why the grip size percentage changes, it's that the Study Spreadsheet is all computationally driven now, so it recalculates every cell as anything changes. Assumptions remain the same, but all the linear interpolations are automatically regenerated.

I know at one time the average height in USA was 5’10. If youve been to developing countries you see the height change dramatically with generations. Maybe the avg height here has changed. Or perhaps its the professions/hobbies on PF lend their selfs to occupations where size is a plus (military/LE/outdoorsmen/hunters)
This is a pretty broad generalization and may have absolutely no merit. Its the first thing that came to mind.

[RJ]

Prototype of a "page 2" of the Graphical Aid, including mockup versions of Table 1 Hand and Table 2 Grip. This is meant to be "stand alone" i.e. give this to someone at the gun counter, along with a tape measure, and they should be able to come back with a gun or guns that fit...maybe? (I call it a "hand out"...get it..."hand"..."out"...heh I crack myself up. Sorry. I'll get my coat.)

As a bunch of "words", this would go with the previous graphical mockup, on the other side. Data is up to date, and reflects discarding those biggest four numbers. Anyone who is a picture-oriented learner could use the graphic side, but anyone who is an analytical and needs the data could use this side.



At this point I'm going to pause and reflect on this for a while.


(If you wanna better look at it, use the attached PDF.)