Savage 12FV in 223 Rem Accuracy Experiences

[SlowSqueeze]

HANDLOADED FOUL ROUNDS

I will be cleaning my rifle after 50 shots, including my 5 foul shots (after the study I’ll only need 1-2 foul rounds, or so I think before I’ve seen the results). Since the main study requires 266 groups with 5 shots each group, that means I’ll need about 150 fouling rounds ((266 groups * 5 shots per group)/45 groups per trip = 30 barrel cleanings, 30 x 5 = 150 foul shots) during the shooting. I want a few groups of foul shots to make sure I limit environmental impacts and get decent, meaningful, statistical results. Hence, I’ll collect more than 1 ‘foul’ target over the course of the shooting.

One of the big questions I have right now is what load to use for my foul shots. I want to pick something decent, and not just randomly pick some charge weight and hope. So, since I’ll be using the ILDM and OCW I figure I’ll use those results to quickly point me in the right direction. While I’m shooting for the ILDM and OCW, I’ll use some of my existing PMC Bronze ammo as foul rounds- I might even be able to glean some results by comparing these foul rounds to my past performance with PMC Bronze.

 

[Evlshnngns]

This is great! I'm grabbing the popcorn

 

[SlowSqueeze]

RIFLE CLEANING APPROACH

It’s been a while since I posted anything to the thread- the new year has me cranking away at life and I haven’t had much time to sit and type out the stuff I want to share… I’m still chipping away though, so keep an eye out every week or so...

After the last few posts I figured I’d tell y’all about my rifle cleaning approach. It’s not anything fancy, and it’s largely based on the approach Ryan Cleckner went over the NSSF video I mentioned way back in the beginning of the thread (LINK HERE), so I’m not going to go into super detail. But I’ll give you enough to understand what I do in general after the video.

Basically I run through 3 rounds of brushing and cleaning. I start out with a brush soaked in solvent and give the barrel 10 strokes or so and let it sit for about a minute. After that I use wet patches until they lighten up, which usually takes about 6-7 patches. I brush 10 more strokes and wet-patch again with about 4-5 patches. Last brush of 10 strokes followed by ~3 wet patches finishes the cleaning. Once I get a clean wet patch after the 3rd brushing, I use 2 dry patches to remove most of the solvent, and finish up with a patch that has 3-4 drops of oil.

I do not regularly use a copper solvent however- in fact I’ve only done it one time on this barrel back in 2015. I might do an experiment later to see if we can see some real effects of copper fouling on the barrel, but for now I don’t think it does much to the accuracy at the level we are discussing.

Here’s a picture of the results from a recent rifle cleaning of my Savage 12 VLP in 243- I use the same approach for all my rifles. I started with the 10 brush strokes mentioned above, and then ran the patches you see on the left, starting with the most dirty one on the bottom. On this day it only took 3 patches to get a relatively clean stroke, BUT my VLP has a stainless barrel while my FV does not- the FV usually takes a few more patches to get clean on each cycle (stainless does make a difference here). Next, I ran 10 more brush strokes, and then got the 3 patches you see in the next column over from the left. A final 10 brush stokes yielded the last 3 patches toward the right (column 3 if you will). Notice how each round of brushing loosens up a lot of gunk in the barrel- that’s why I use the brush 3 times. The last two patches you see on the right are the dry patch (second to the left from the right), and my oil patch with 5 or so drops of light oil (far right). I’ve got my process to the point that it only takes about 15 minutes to clean from start to finish.



Oh, one little tidbit- after I clean my rifles I store them barrel down so anything in the chamber will flow to the muzzle (and not the action). I usually put a small piece of paper towel or a napkin to catch stuff, and I usually get a little (maybe a 1/10 of a drop). Better to come out the muzzle than go into my action I guess!

 

[SlowSqueeze]

WHY I USE 5 SHOT GROUPS

A lot of people have been over and over the number of shots to use in groups (this is the best one I found).

[EDIT] I found an article over at PrecisionRifleBlog.com that does a much better job of explaining what I'm doing here, and how shots per group impacts things: check it out here. For comparison purposes, I'm actually using 35 round groups for the statistical calculations (keep reading to understand how). [END EDIT]

It’s ultimately a matter of opinion about what you think is best for you. From my experiences, I have noticed 3 shot groups are too sensitive to error for my needs- the standard deviations tend to be higher overall, and a single flier (or lack of one) has a real impact on the results of individual groups.

NOTE: For hunters and hunting rifles I do see logic in practicing 3 short groups- if you need more than 3 shots at an animal there are bigger problems than the ammo and rifle.

I have also used 10 shot groups at times, and 7 shot groups as well, but with these larger groups it becomes hard to see individual bullet holes in a group that are close to the center. While individual 7 & 10 shot groups would be better from a statistical confidence perspective, it’s a trade off to accurately identify where bullets hit.

So I find that a 5-shot group is a solid balance between the more sensitive 3-shot groups and the bigger 7 or 10 shot groups. Later, when I present my statistical results, I’ll show you some math about limitations that 5 shot groups present- basically the statistics will help answer a basic point: if you gave me 1,000,000 rounds I can all-but-guarantee I could hit something at an absurd distance like 1,000 yards, but could I do it with 5 rounds (more specifically, what is the probability I could hit something with 5 rounds)? There’s no right answer which ever way you go however...

One final note about 5 shot groups- 5 is good even multiplier of 20, 50, and 100, which makes it convenient to divide a box of ammo with 20 rounds into 4 groups. 3 shot groups means I’d have 2 bullets leftover (might be good for foul rounds!), and 7 shorts has a similar problem. Outside of the reasons discussed, there’s nothing special with 5 shots in a group.

 

[SlowSqueeze]

WHY 7 GROUPS @ EACH CHARGE WEIGHT

Unlike the 5-shot groups, there is a LOT more importance on the number of groups I collect as it pertains to statistical methods. The total number of rounds for each charge weight, or number of shots per group times the number of groups (5 x 7 = 35, statistically the number of ‘observations’) directly determines the mathematical confidence of my results. I’d love to be able to tell you that I’m 100% confident that my results will be perfectly accurate, but in reality there is always some unknown- I just might have had 7 really good days when shooting 22.8 grain loads, even though they are going to be shot on different days. And so it goes that if you dig really deeply into the statistical math (and don’t fall asleep), there are some special things that happen as you get more than 30 ‘observations’. Having 35 shots for each charge weight pushes me above that magical level.

Later on I might also play some games with OnTarget later on and see if I can visually simulate a single 35 round group by combining all the individual 5 shot groups into one ‘super group’. Doing that changes how I can use the numbers a little, though I’m hopeful it won’t change the results! If the results do change that’ll likely signal an error somewhere... Mathematically I’m already combining them by taking an average of an average (averaging the Average To Center, ATC).

To sum it up, I wanted a lot of shots/observations for statistical reasons, but there’s a direct cost for each group added- it takes more cash and a lot more time to shoot a lot of groups, so getting just over 30 shots per charge weight fits the bill for my needs.

 

[SlowSqueeze]

PROBLEMS WITH MAX GROUP SIZE MOA

To this point, I have been using the common measure of MOA for my results- that is, I have been taking the MOA measurement of the two shots furthest apart for each group, or Max MOA. A few years back, when I started using OnTarget, I saw another measurement under the Max MOA that was labeled ATC (“Average To Center”). I read up a bit on ATC, but kept using Max MOA because it’s a lot more common.

There is a problem with Max MOA however.

Max MOA is very sensitive to your WORST two shots (and by very sensitive I mean COMPLETELY DEFINED). In other words, Max MOA is a worst case scenario for the precision and accuracy of a given group because it ONLY considers the two shots with the greatest distance between them. The following figure shows what I mean.



The two arrows point to the only two shots that count for the Max MOA calculation: the other three shots are ignored. Notice in the above how the ATC measurement is just under half of the Max MOA 0f 1.091. I’ll talk about that in my next post. One additional note about Max MOA for now- if you are a competition shooter, or a hunter, the Max MOA may be a better metric for you because it shows your worst case accuracy between the two widest shots. I’m only hunting paper with this study, and am very focused on statistically meaningful results, so Max MOA’s limitations are, uh, undesirable...

 

[SlowSqueeze]

ATC/MEAN-RADIUS GROUP SIZE MOA DETAILS

Over the last month I reread some material about Average To Center, or ATC (also called “Mean Radius”), and will be adopting that as my precision metric moving forward. ATC uses ALL the shots in a group to determine a measure, not just the two with the greatest distance like Max MOA. To illustrate the difference between the two methods, look at the following figure…

Which group would you rather shoot, the one above, or the one in my previous post? (HINT: If you said the one in my previous post, make an appointment with the optometrist!)

Turns out both groups have similar Max MOA measurements (1.091 compared to 0.933). So you can’t really tell how good an entire group is with Max MOA, but you can tell how bad the worst two shots are. The ATC MOA tells us, with numbers, that the above group has an ATC MOA of 0.276, which is a LOT better than the one in the previous post of 0.484.

If you don’t like math, save yourself some time and skip to the last paragraph…

To level the comparison out a bit more, we can take 0.164 off the first group size (1.091 - 0.933 = 0.164) so both would mathematically be 0.933 Max MOA. Knowing that ATC is less than half of the Max MOA, we can take half of 0.164 off of the ATC for the first figure and we see 0.402 MOA (0.484 - 0.082 = 0.402). In other words, with simple correction the group above is over 0.1 MOA BETTER than the group in the previous post.

Also as a general point, the closer the ATC:Max MOA ratio is to 1:2 (i.e. 0.5) the more uniform the group should be- the above case the ATC/Max MOA ratio is 0.295 (0.276/0.933 = 0.295). Since 0.295 is pretty far away from 0.5, that means something went wrong with that group, which we can all see as the one shot out of five total that shot high. The group in the previous post had a ratio of 0.444 (0.484/1.091), which is pretty close to 0.5- meaning the flier that defined the Max MOA wasn’t THAT far off from the rest of the group...

Because ATC as a measurement agrees with the common sense preference we saw above, I’m going to switch to using that going forward. But, ATC has one issue that I don’t like...

 

[SlowSqueeze]

2 ATC AS MY STANDARD MEASURE OF PRECISION

Because ATC measures the average distance of a shot to the center of a group (hopefully the point of aim), it can only tell us HALF of the size of what we can expect to be able to hit. Put a different way, if I have a 0.5 MOA ATC rifle, I can confidently say that I can hit a 1 MOA target. That’s because ATC does not consider where the shots are landing around the center of a group- and that’s why the ATC number in OnTarget is usually about ½ the value of the Max MOA. You can hit half an MOA low and half an MOA high for a distance between the shots of 1 MOA, but the ATC would only be 0.5 for that two shot group. See the following illustration of a 2-shot group- you can see that the Max MOA is exactly twice the ATC MOA (accounting for significant digits) because the ATC calculation finds the center point between the two shots (the large “+” sign) to get the average of each shot to that center.



In order to get results that are a little more comparable to the far more common Max MOA data that people routinely publish online, and to give more useful measurements of targets we can expect to hit, I’m going to use the ATC times 2 (or 2 x ATC) as my standard measurement of precision and accuracy. Or, instead of Mean Radius, I’m going to use Mean Diameter (diameter is 2 x the radius). Doing so eliminates sensitivity from Max MOA on the two worst shots in a group and aligns better with my statistical methods and results. In the above example, 2ATC is 2 x 0.505 = 1.010 MOA which is REALLY close to the measured/displayed 1.009 Max MOA.

 

[SlowSqueeze]

UPDATED INITIAL RESULTS

Looking back over my historic data and loading up the OnTarget calculation for ATC yields the following data. This time I color coded all the different days of shooting, so each color is a different day. I also found a day that I didn’t load into the last set, so this table has more data, and smoothed things out a bit for the lower charge weights.

NOTE: The data in the table was entered directly as ATC MOA from the OnTarget results. I multiplied the average of the ATC MOA results by 2 to get my 2ATC column (labeled “ATC Ave*2”) that I used for the plot below.



When I plot the average the ATC measurements multiplied by 2 we see the following.



The first take-away from the plot is the Extreme Spread (ES) of the data is reduced to 0.266 MOA (above max of 0.641 - min of 0.375) from 0.3998 MOA in the original data. So, as intended, there is a reduction in variation by including more shots from the ATC measurement. This makes a lot of sense because there is less sensitivity to fliers, and it’s counting over two times the number of shots in each group. [Max MOA counts 2 out of 5 shots, while ATC counts all 5: 5/2 = 2.5 more shots for ATC]. Next, but not intended, the 2ATC results are lower than the Max MOA results. 23.2 gr is now showing 0.375 2ATC MOA, while it was 0.49175 Max MOA. BUT, 23.2 gr is the best load again, so there’s comfort in seeing the same ‘right’ answer.

For the record, I’m not particularly pleased about the lower numerical values for the 2ATC results because it might look like I’m trying to skew my data to make things look better, which I’m not. Rather, I’m trying to include the maximum data I can in my dataset and get meaningful statistical results. In short, you won’t find me saying that my rifle is a 0.375 MOA (ATC or not Max) at 23.2 gr of IMR4064, because it is not.

 

[SlowSqueeze]

DOES ATC HAVE LOWER STANDARD DEVIATION- YES!

As another feat of mathematical drudgery, I wanted to see just how much lower the ATC standard deviation (that’s just ATC StDev from the above table) is than that of the Max MOA StDev. So I plotted the two SD columns into the following graph.

MATHEMATICAL NOTE: I’m not using twice the ATC Standard Deviation because (I’m pretty sure) that’s not mathematically viable. The Standard Deviation is always a distance from the ATC (toward and away from the center), and the orientation of the hits around the center wouldn’t change the amount of deviation. When I look at the implications on the target shape (in my next post) you will see where I doubled the standard deviation contribution, but this real world application doesn’t apply for the following comparison to the Max MOA Standard Deviation. If there are any Stats guru’s out there please shoot me a PM and let me know if I’m totally screwed up...



What you want to look for is 1) a downward shift in the blue line, which tells you that the ATC SD is lower (duh- and a good thing), and 2) you want to see where it deviates from the orange line. For example it looks like at 22.0 grains the two curves are wide apart- the orange Max MOA curve has a spike, but the blue ATC stays really flat.

1. Downward shift of Standard Deviation from Max MOA to ATC: first off the ATC StDev is lower than the Max MOA StDev, and it’s over 2.6 TIMES LOWER. That makes a lot of sense because ATC includes 2.5 times more shots.
   
   
2. Max MOA and ATC StDev Divergence: if my data was a little better, and I was really looking to dig into this I would present more analysis here (I did spend a few hours playing with that graph). Since I can’t really experiment and dig into what drove differences, I’m not going into detail outside of pointing out that 22.0 grains show a LOT more variation with Max MOA than ATC does- that’s probably due to a) a lot more fliers, b) the fliers that were there are a lot worse, or c) both.

Also of note, between 22.6 and 22.8 grains there is very little variation in both curves. There is a second ‘quiet’ range between 23.2 and 23.5 grains. Or, if I fip things a bit, perhaps my groups at 22.9 and 23.1 are unusually bad and caused the peaks you see (maybe I had some anomalies and a lot of bad shots?). As good as my old data set is, it’s still limited and can’t give me the resolution that I need to be able to answer questions like that, and since I’m using better methods now, I won’t be able to fill in any missing data- but we’ll get better and more thorough data that will answer those and far more questions soon!

 

[SlowSqueeze]

2ATC AND THE REALM OF (CONSERVATIVE) PROBABILITY

Okay, lots of math coming (Oh God, more you say!?), so if you don’t care about the stats, skip to (or wait for) the next post… I must be going through statistical withdrawal!

So what CAN I say about my rifle after all this statistical crap? It did perform well at 23.2 grains over four 5-shot groups: 0.491 Max MOA average, or 0.375 2ATC MOA.

Basically, an average like either of the two numbers I just mentioned tells us that 50% of the time you’ll be higher than that number, and 50% of the time you’ll be lower than that number. Those aren’t really betting odds- at least not where I’m from! What if I wanted better odds than flipping a coin? What if I wanted to be 99.99966% confident (i.e. guaranteed money) I could hit a target- how big would the target have to be given the above data?

The number 99.99966% is special because it represents something called “six sigma”. If you remember the bell curve from earlier in my thread, each standard deviation was one ‘sigma’ value, so taking 3 sigma (3 x the standard deviation) on either side of the average (3 for the outside/high side, and 3 for the inside/low side equals 6 total) means you will have 99.99966% of the shots in that space. This is a really important point- I only care about the probability of a shot outside of the average line. Put another way, the shots inside that average are better than 0.375 MOA, which I’ll take as gravy (a good thing). I want to know about the other part- the shots that fall outside, so I only use 3 standard deviations for just the outside ‘wide’ shots.

So our average ATC at 23.2 grains is 0.1875 ATC MOA, and the Standard Deviation (sigma) at 23.2 grains is 0.046 MOA. So, if I want to be 99.99966% sure that I can hit a target with my rifle using my past hand loaded rounds I would need a target that has a 0.1875 + (3 x 0.046) = 0.326 MOA radius. BUT, that is Average To Center MOA (i.e. radius), so we need to double the target to 0.651 MOA to get a diameter. In other words, I can statistically say that my rifle and I have a 99.99966% probability of being 0.651 MOA, or better, with my 23.2 grain handloads.

• 99.99966% means that 34 out of 10,000,000 shots would miss a 0.651 MOA target (not really, because I didn’t account for the diameter of the bullet, but let’s keep it simple). I don’t need to tell you 10,000,000 rounds is a LOT, and probably overly conservative.
  
  
• What if, instead, we wanted 90% confidence, or 1 miss out of 10? That seems a little loose to me, and would mean I’d miss 2 rounds out of a standard 20 round box… Not good enough for a ‘precision’ shooter, but might be worth a bet...
  
  
• How about 99% confidence for 1 miss out of a 100? That’s probably the most reasonable, no? Well, 4 sigma represents 99.4% probability, so doing the above math over with 4 sigma instead of 6 sigma means that our target would need to be 0.559 MOA in diameter (0.1875 + (2 x 0.046) = 0.2795 MOA radius, then 2 x 0.2795 = 0.559).
  

 

[SlowSqueeze]

HOW DOES BULLET THICKNESS CHANGE THE PROBABILITY OF A HIT?

But what about the bullet thickness, does that matter? Of course it does, because if we just graze the side of the target we essentially missed the target with the center/tip of the bullet. In other words, half of the bullet can still hit a target if we just barely miss. Such a miss means that our target could be a full bullet width smaller (half a bullet width on top, and half on the bottom). So, 0.559 - 0.224 = 0.335 MOA. This is helpful for something like a steel target that you want to “ping” but that’s about it…

I only include that point because of the way I posed the question above- “How big would a target have to be…” If you want to know how big something would need to be to hit it with a bullet, in real world terms, then accounting for the bullet diameter makes some sense.

Before leaving this post I want to say two things:

1. Taking off the bullet diameter is a bit cheesy, but technically its correct- you could hit a paper/steel target with the side of a bullet if you miss on the center/tip of the bullet. But for precision shooting that’s a miss in my book. Also, competition shooters break ties with the closer shot, so it’s more conservative to just count center of bullet to center of target.
   
   
2. A statistical method caution- technically using the StDev the way I did in my last post is wrong because I didn’t have enough shots at 23.2 grains (only 20 shots), and hence I would need to use ‘skewed distribution’ math to tickle out probabilities. In the future I’ll have more than 30 shots, so the above analysis will work for my final data. Also, there is some room on the inner side of the above target beyond 3 sigma- so the probability of a hit is higher than 99.4%. There is a chance I could shoot smaller than 0.0.0955 MOA (0.1875 - 2 x 0.046 = 0.0955), but it aint a big chance!
   

 

[CharlieNC]

2ATC AND THE REALM OF (CONSERVATIVE) PROBABILITY

Okay, lots of math coming (Oh God, more you say!?), so if you don’t care about the stats, skip to (or wait for) the next post… I must be going through statistical withdrawal!



So what CAN I say about my rifle after all this statistical crap? It did perform well at 23.2 grains over four 5-shot groups: 0.491 Max MOA average, or 0.375 2ATC MOA.

Basically, an average like either of the two numbers I just mentioned tells us that 50% of the time you’ll be higher than that number, and 50% of the time you’ll be lower than that number. Those aren’t really betting odds- at least not where I’m from! What if I wanted better odds than flipping a coin? What if I wanted to be 99.99966% confident (i.e. guaranteed money) I could hit a target- how big would the target have to be given the above data?

The number 99.99966% is special because it represents something called “six sigma”. If you remember the bell curve from earlier in my thread, each standard deviation was one ‘sigma’ value, so taking 3 sigma (3 x the standard deviation) on either side of the average (3 for the outside/high side, and 3 for the inside/low side equals 6 total) means you will have 99.99966% of the shots in that space. This is a really important point- I only care about the probability of a shot outside of the average line. Put another way, the shots inside that average are better than 0.375 MOA, which I’ll take as gravy (a good thing). I want to know about the other part- the shots that fall outside, so I only use 3 standard deviations for just the outside ‘wide’ shots.

So our average ATC at 23.2 grains is 0.1875 ATC MOA, and the Standard Deviation (sigma) at 23.2 grains is 0.046 MOA. So, if I want to be 99.99966% sure that I can hit a target with my rifle using my past hand loaded rounds I would need a target that has a 0.1875 + (3 x 0.046) = 0.326 MOA radius. BUT, that is Average To Center MOA (i.e. radius), so we need to double the target to 0.651 MOA to get a diameter. In other words, I can statistically say that my rifle and I have a 99.99966% probability of being 0.651 MOA, or better, with my 23.2 grain handloads.

• 99.99966% means that 34 out of 10,000,000 shots would miss a 0.651 MOA target (not really, because I didn’t account for the diameter of the bullet, but let’s keep it simple). I don’t need to tell you 10,000,000 rounds is a LOT, and probably overly conservative.
  
  
• What if, instead, we wanted 90% confidence, or 1 miss out of 10? That seems a little loose to me, and would mean I’d miss 2 rounds out of a standard 20 round box… Not good enough for a ‘precision’ shooter, but might be worth a bet...
  
  
• How about 99% confidence for 1 miss out of a 100? That’s probably the most reasonable, no? Well, 4 sigma represents 99.4% probability, so doing the above math over with 4 sigma instead of 6 sigma means that our target would need to be 0.559 MOA in diameter (0.1875 + (2 x 0.046) = 0.2795 MOA radius, then 2 x 0.2795 = 0.559).

You're on the right path but need to read a bit more. The atc is not the standard deviation, not even close. SD is calculated using the difference squared and will result in a larger answer.

 

[SlowSqueeze]

You're on the right path but need to read a bit more. The atc is not the standard deviation, not even close. SD is calculated using the difference squared and will result in a larger answer.

Hi Charlie- I had to go back and reread my posts carefully to make sure I didn't state something poorly leading to a misunderstanding (I didn't find anything). I know ATC and not a standard deviation (as the "A" in ATC means average). I did use the standard deviation OF the ATC results, but that's ATC StDev, and I only used that to determine some probabilities of a given accuracy. So, in short, ya, ATC is an average, not a Standard Dev...

 

[CharlieNC]

I believe you should actually use the SD of the individual shots from the center point if you want to predict the % of where individual shots will fall. That works well to infer Fclass scores, which are the % of shots inside various moa rings. All boils down to what you want to infer.

 

[SlowSqueeze]

HOW ACCURATE WAS SLOWSQUEEZE’S RIFLE (WITH SIMPLE RELOADING METHODS)?

Everything I presented to you so far is based on my past shooting data from shots taken well over a year ago. My methods weren't great, and I was still learning a lot each time I went out shooting, or reloaded some ammunition. So, I can only tell you how accurate my rifle WAS with those methods- take the following with a grain of salt, because I’m confident my new methods will improve things a bit...

I will say (with some mathematical qualifications noted in my previous posts) based on past data, that my Savage 12FV and hand-loaded rounds with 23.2 grains of IMR4064 with 69gr SMKs, using PMC Bronze brass and CCI400 primers has over 99% probability of being accurate to 0.559 MOA (hitting a target that is 0.559 MOA in diameter) at 100 yards when I shoot it.

If I were plinking and hitting targets, my Savage 12FV with my best load should be able to hit a common ballpoint pen on end (not that I have a reason to shoot a pen, but it’s a common object that we can all understand), which is just over 0.335 MOA at 100 yards.

And all of that is with basic hand loading and a poor powder selection for this caliber. I can’t wait to see what better methods can do in the near future!

 

[gpoldblue]

If you are not using wind indicators(wind flags,etc) and enough of them and learn to read them,you will never realize the true potential of you and the rifle. Without flags,you can't determine if the verticle and/or horizontal is due to load,shooter,or conditions. I think wind flags will do you more good than anything. A chronograph would help your load development also.

Stats are good,but just like football,stats don't put numbers on the scoreboard.

Good luck and good shooting.

 

[SlowSqueeze]

If you are not using wind indicators(wind flags,etc) and enough of them and learn to read them,you will never realize the true potential of you and the rifle. Without flags,you can't determine if the verticle and/or horizontal is due to load,shooter,or conditions. I think wind flags will do you more good than anything. A chronograph would help your load development also.

Stats are good,but just like football,stats don't put numbers on the scoreboard.

Good luck and good shooting.

A while back I considered doing the shooting at an in-door range (100 yards only), and using a nice heavy rifle stand. Ultimately this isn't the type of shooting I do, so I decided to simply accept the contributions (i.e. error) from those two sources (wind, and rifle position <-shooter).

I'm with you on the stats comment! Accuracy, and more so precision, are all I really care about. I'm doing a lot of learning about what really drives those two along they way. The only real stat I care about is the ones measured from tight groups!

 

[SlowSqueeze]

CASE PREP OVERVIEW

It’s time to start working up my cases for the rounds I’ll use in the study. I mentioned earlier that I’m using Lapua cases, and a general process based on this article about case prep. I hope you have done your homework and read that article the required number of times, because we’re about to put it all into practice.

Before I hit the garage (where I do my loading) though, I want to share my approach to preparing new unfired brass for loading, which involves three phases:

1. Measuring the cases to determine initial state, and long-term how I need to modify things
2. Machining/Modifying the cases to bring them all into defined tolerances (the article)
3. Verify & Clean the cases before loading

I will treat all 100 of my cases in each step of the process, including measurements- my sample size is 100. If you are following at home, you DO NOT need to measure all of your cases- a sample of 7 or so cases should give you what you need in most circumstances. I am treating all 100 cases for the following 2 reasons:

• Potential outlier cases that have characteristics that are not within assumed ranges- my results (I hope) will be around for a LONG time, so extra care at the beginning of the process will make those results stand up over time.
• To be able to determine root causes of accuracy and precision issues that might be attributable to cases (I’ll need all the data I can get to perform real Root Cause Analysis (RCA)).

For each step I perform below, I’ll include the total time it took me to perform that step on all 100 cases. That should help you to plan if you are doing similar work, and give you a nice ball park for the total investment in time. Since I went through most of my measurements previously, I’ll save you the details for those steps and present a summary table with the results. So, without further delay, let’s get into the first phase...

 

[SlowSqueeze]

PHASE 1 MEASURE: LAPUA FACTORY CASES

Over time I want to build up a detailed library of factory cases that I can use to really drive my case selection and preferences. I know Lapua cases are commonly considered the premier brand, but Norma cases aren’t far behind and are less expensive. Ultimately the case quality really boils down to the amount of time that is required to bring the cases into tolerance. In order to really understand that time investment, we need to be able to compare the final case quality in quantitative terms.

Rather than bore you with individual posts for each measurement, I have packaged everything up into a single table, which is included below. Each measurement also includes the standard deviation, extreme spread, and the total effort it took to measure the 100 cases.