[brycelarson]

Thanks to Chris Nathan for the donation of the indoor testing facility! We made a much smaller mess this time.

Full Data Sheet

Video Intro for test

Barrel test 2 - test in progress





enjoy everyone!

This post has been edited by brycelarson: 06 January 2009 - 02:39 PM

[Iram]

And the results are? I'll take unprocessed data even...

[Troy]

brycelarson, on Jan 2 2009, 06:38 PM, said:

This is just a placeholder and a bit of a tease to make people excited. The accuracy portion of our barrel test is done - data and videos coming up quick.

I EXPECTED DATA!!!

[brycelarson]

soon, be patient.

[cockerpunk]

holy smokes you guys will like this test.

i've been looking at the numbers for a good 4 hours now, very very very interesting stuff.

bryce, we'll talk tomorrow about our stuff.

This post has been edited by cockerpunk: 03 January 2009 - 02:22 AM

[Iram]

cockerpunk, on Jan 3 2009, 02:22 AM, said:

holy smokes you guys will like this test.

i've been looking at the numbers for a good 4 hours now, very very very interesting stuff.

bryce, we'll talk tomorrow about our stuff.

Oh quit teasing and post the damn numbers already.

[lukewarm]

ive checked this thread 15 times in the last 40min

[Iram]

Well, cockerpunk's last post was at 2:22 AM Eastern. My guess is shortly after posting, he went to bed. So, we've probably got several hours before there's any chance of getting the results.

[brycelarson]

I'm looking at the numbers right now as well - there are a few things we need to decide how to attack (mathematically) - so hopefully soon.

[Iram]

Well, here are a couple suggestions from the guy who sucks at statistics:

I'm assuming you have a table that's basically X and Y coordinates for each shot.

I'd suggest:
1. Center the grouping: Average of all of the X values. Add this value to all of the X values. Average all of the Y values. Add this to all of the Y values.
2. For each shot, determin the distance the shot is from the center of the group. I believe the best approache would be taking sqrt((X^2)+(Y^2)).
3. Average the distance each shot is from the center of the group.

[brycelarson]

Iram, on Jan 3 2009, 09:05 AM, said:

Well, here are a couple suggestions from the guy who sucks at statistics:

I'm assuming you have a table that's basically X and Y coordinates for each shot.

I'd suggest:
1. Center the grouping: Average of all of the X values. Add this value to all of the X values. Average all of the Y values. Add this to all of the Y values.
2. For each shot, determin the distance the shot is from the center of the group. I believe the best approache would be taking sqrt((X^2)+(Y^2)).
3. Average the distance each shot is from the center of the group.

you don't need to do that if you use a standard deviation - the SD will tell you how far from the mean all of the shots are.

If you wanted to do analysis of how consistent the pattern was - then you would need to normalize the data. for our needs - the SD works fine.

[Iram]

I would imagine knowing how consistant the pattern is will be important. Any chance I can get a copy of the raw data to work on?

[cockerpunk]

data should be up momentarily.

[brycelarson]

check the first post

[Iram]

Thanks

[Iram]

After looking at the data and manipulating it the way I had suggested, I would conclude that they are all about the same. I'm not seeing any pattern to the data that could be used to conclude that any specifc relationship between accuracy and bore size/length/number of pieces.

I found the center of each group, and then the distance each shot was from the center. Graphed the average and standard deviation. "Two" is for the two piece, "One" is for the one piece. I dropped the Dye/Armson/Empire barrels off the list so it's all from one manufacturer.

This post has been edited by Iram: 03 January 2009 - 12:37 PM

[Capt. Kirk]

i assume you where aiming for the origin? how did you determin that you where aimed in the right spot to begin with?

[cockerpunk]

Capt. Kirk, on Jan 3 2009, 12:41 PM, said:

i assume you where aiming for the origin? how did you determin that you where aimed in the right spot to begin with?

we did not aim at anything in particular.

as long as the shots land on the grid somewhere, we can back calculate the center of the pattern, thats what Iram did.

i prefer to use the standard deviation, which is automatically normalized.

[Capt. Kirk]

so you where looking at the consistancy of the clusters rather than the position on the grid?

[cockerpunk]

Capt. Kirk, on Jan 3 2009, 12:51 PM, said:

so you where looking at the consistancy of the clusters rather than the position on the grid?

yup!