Just for j0e_bl0ggs:


More pictures and detailed analysis later.

Paul

Hokay...
All of the targets have the following in common:
20-grain, hollow point, "Dog Town" bullet.
IMR3031 Powder
Range 100 yards
Temperature: 40 F
Humidity: 50%
Altitude: ~150 ft Above Sea Level
At least 3 minutes between every shot.
Barrel was never more than detectably warmer than ambient.
20x scope - Could see brachial pulse in sight picture.

First load was 25.4 grains. Here's the group:

The "light" lines delineate 1/2 inches. The "dark" squares are 1".

Here's the chrono, and point of impact data table:

Most of it should be self-explanatory except the statistical stuff. The data are sorted by velocity from slowest to fastest, but the "Shot #" preserves the shot string sequence.

The "x" and "y" data values are the position of the center of each shot relative the to the point of aim in rectangular (x:y) coordinates. For example: the point of impact (PoI) of shot #4 (the first one above), was 0.525" to the left (-) of the point of aim (PoA), and 0.128 high (+).

The "Euclid Dist" is the Euclidean distance of the center of each bullet hole from the PoA.  This is how far the PoI was in a "straight line" from the PoA.

The "r^2:x" value is the "R-squared" value of the muzzle velocity correlated to the "x" values of the PoI. The "r-squared" value in this case (for the "x" value of PoI) is the "correlation coefficient" of the relationship between the muzzle velocity and the windage of each shot in the group (25.4 grain group). The exact same is true for the "r^2:y" value, except with respect to elevation. This number is the "amount of variability in the windage/elevation explained by muzzle velocity". So, for this first 25.4-grain group, about 37% of the windage and 53% of the elevation is "explained" by the muzzle velocity of each shot.

If you look at the "Average" Euclidean distance, you can see that the value is 0.503". that means that the average PoI was half an inch from the PoA. This is NOT the groups "size". It is the average "miss" of the PoA.

If you look at the average "x" and "y" values, those are the scope adjustments for windage and elevation that need to be made to have the average Euclidean distance (average miss of PoA) be "zero". In other words, were I to select this charge, I would adjust the 'scope .442" to the right, and .144" down.

The "Std. Dev." (standard deviation - SD), is a measure of the variability of the given variable - muzzle velocity, windage, elevation, etc. It is a function of sample size. For the muzzle velocity, the SD is 89 f/s. Notice that the spread of five shots is  4537 to 4317 or 220 f/s. That's BIG. However, a sample size of 5 is SMALL! Therefore, the standard deviation is multiplied by 2.776 and added and subtracted from the average to get the "95% confidence interval". (If the sample size were "large" - say 121 or greater,  the multiplier would be 1.996.) That means that based on THIS sample of FIVE shots, if I shot 100 more times, 95 of the PoI would land in a an ellipse that was approximated half an inch wide by six-tenths of an inch high. The rest would land outside that ellipse.

Notice that if I want to predict the muzzle velocity of the next 100 shots based on THIS sample of FIVE shots, I would guess they would all fall somewhere between 4166 and 4659 f/s.

If I wanted to predict how far the next 100 shots would be relative to the PoA, I would look at the 95 % Confidence interval for the Euclid Dist. In this case somewhere between 0.005" off to 1.000" off of PoA.

That's it for the data table. All the rest are the same calculations, just recalculated for each charge's sample (group).

Here's the 25.7-grain group and data table:





Here is the 26.0-grain group and data table:



Note muzzle velocity standard deviation. That's more like it.

Here's the 26.3-grain group and data table:





Looks to me like the 26.0-grain charge gets the nod. While I wasn't looking for an average MV that fast, 4539 f/s, I'll take it since there were no signs of bullets coming apart even with the 26.3-grain charges. But what really makes this the "winner" is the standard deviation of the MV. TWENTY ONE feet per second. That's a good number for a 'pedestrian' bullet let alone a 'speed demon'!

I suppose I should includ some "group size" numbers. I don't "do that" like other people (gun-writers ptooey) do because I don't like to fool myself, let alone others. (Not something with which gun-writers (ptooey) concern themselves.) Instead, I calculate the 95% confidence ellipse that surrounds the group. I call the area of that ellipse the "group size" because it represents the size of a "hole" that would be made if you shot 100 shots at the same target.

Charge . . Group Size (in square inches)
25.4 . . . . . 0.924
25.7 . . . . . 0.961
26.0 . . . . . 0.600
26.3 . . . . . 0.748

So, not only does the 26.0-grain charge produce the smallest deviation in muzzle velocity, it also produces the smallest group. One would expect that, but it is not ALWAYS the case.

Paul

PS - Notice that as charge/velocity goes up, the average Euclidean distance increases. As the bullet speeds up, the group center moves to the left.

It shoots! TY!
It'll be worth trying that 26gr without the magnetospeed on board.

I don't think I've EVER shot 20 rounds at 100 yd into as small a group as these. I know this is a yawner to benchresters, but I've never put that many together in one string. The point being; I'm not sure removing the MagnetoSpeed could make a difference. I don't know if I can shoot better than that.

Paul

Worth a try and I bet you could actually still tighten em up a little...

Proves the quality of your work.

You have them cooking for speed as well as grouping!

Shotgun patterns at best. Speeds are also unimpressive.

RJ

That's a shooter!!!!

Quote from: recoil junky;137544Shotgun patterns at best. Speeds are also unimpressive.

RJ

Paul

By the way, did 'you' notice that the average velocity on the 25.7 and the 26.0 charges were essentially the same? Another indicator of "optimal" charge weight.

Paul

What did I ever do without a chronograph? While working up the load for the .222 many years ago I stuffed some IMR3031 (cuz that's what we had) in a case a nd tromped in a bullet. Loaded 5 that way and Dad and I went out and shot the mess into a pinky nail sized group out of his .222 over 20 gauge Stevens. I figured "If this is all the harder this is going to be" . . . . . . .

Little did I know!!! I worked without one until about ten years ago when I got the 300 RUM. Seeing if a 180 grain Accubond really was reallybgoing over 3200 fps was worth buying one!!

RJ

Now if you just knew where there were some ground dwelling rodents.

RJ

I know where some are, but they are "ground-dwelling" right now for sure.

While there are many ground-dwelling rodents in Alaska, the ones that are similar to the ones shot for sport Outside are the arctic ground squirrel (Urocitellus parryii) and the hoary marmot (Marmota caligata). The arctic ground squirrels ("Siksik" in Inupiaq), are ubiquitous in Alaska. They are related to prairie dogs, but a bit smaller. They do not form "towns" like prairie dogs. The hoary marmots are considerably smaller than the "groundhog" of the eastern Lower 48. A big hoary will go 7 lb and are found in rocky outcroppings of mountains.

While both can legally be hunted, and there is no closed season, if you 'take' one, you are legally supposed to 'salvage' either the meat or the hide. No one "hunts" either of them as far as I know. Natives have long made kid clothing and some ornamental pieces of clothing from siksik skins. I don't know of anything made by natives from marmot skins. Both are deep in hibernation at this time.

I'd get after the 'dogs' - fox, coyote, and wolf - with this rifle, and it is soon 'that time of year' for them. Enough daylight to shoot with, and warm enough not to be too uncomfortable. Not much hope for wolves this year. I might be able to get out after coyotes and foxes, but not likely. I'm afraid this rifle will be relegated primarily to punching paper. Which, makes me lean heavily to selling/trading it for something I would actually "use".

Paul

Here is a graph of all of the shots aggregated into one 'electronic' target. I subtracted the average "x" and "y" values of each group from each "x" and "y" value of each group, which moves the center of each group to the rectangular coordinates of "0,0" - the center of the graph. This is how one can make a 'cumulative' graph of one's targets shot during different range sessions, or as in this set of data, at different points of aim.



I attempted to make the "holes" approximately the correct diameter - 0.172" - relative to the scale of the graph. I also tried to make the graph "square", so the windage and elevation LOOK was correctly proportioned.

While one would rarely do this for groups shot with different charges or bullets, all of these groups were sufficiently similar that I decided to go ahead and mix 'apples with pears'.

In this graph the width ("windage") of the 95% Confidence Ellipse is greater than the height ("elevation"). While that is not uncommon in 'normal' (one-charge/one-bullet) aggregate graphs, the reason it appears here is probably because the trigger on the Axis is most certainly not a "target trigger". This is not particularly noticeable in the sight picture until the 'scope's magnification gets up above about 16x.

Note the area of the 95% Confidence Ellipse - 0.428 square inches. That's SMALL. If you took the largest windage and elevation values of the ellipse (0.820" x 0.664") and created a rectangle instead of an ellipse, the area of that rectangle would only be 0.55 square inches. To get some 'reality' to that, draw some circles and squares with these dimensions on a piece of paper.

Also note that while a few of the "holes" 'cut' the line of the ellipse, all of their centers are within the ellipse. That means that the ellipse is actually larger than the "point-to-point" spread of the widest two points. The "point-to-point" spread is how most people measure "accuracy", (correctly, precision). Assessing precision by taking point-to-point measurements is "OK", because it is easy. However, not only does it provide a "biased" result, that bias is toward "smallness". In other words, if one relies on that measurement of precision, one is fooling one's self with respect to the actual precision the rifle.

Why do we care about "group size" anyway? Besides the ubiquitous "mine's better than yours" crapola, we want to predict (know) where the "next" shot is going to go. Predicting that from "point-to-point" measurements - say 1 MoA - means expecting the point of impact to fall within a MoA will be wrong about 50% of the time. And that 50% would be if the point-to-point MoA measurement was from something more than about 20 shots. For three-shot groups, the likelihood of hitting that MoA again would be about 30%.

Of course people shoot within the point-to-point spreads quite often. That's to be expected. Every time they do they up the "sample size" and the size of the 95% Confidence Ellipse gets smaller. However, folks have a tendency to only "count" the small groups, and ignore the "big" groups and fliers. Going through the exercise to produce the above type graph (and the other algebraic gymnastics that I do), is my attempt to keep from fooling myself about how "good" I and my rifle really are.

Paul

amazing groups!!!! Way to go!!! Do you have a wolf pelt? God Bless.