Ruined For Life - Velocity Truths
HTML-код
- Опубликовано: 17 сен 2022
- Visit us on Patreon: / winninginthewind
In this video, I describe the bear trap that is chasing your tail on velocity statistics, whether is is extreme spread, standard deviation, or something else.
Reloading can be dangerous. Use proper, safe techniques and slow, deliberate load workups to ensure safe maximum charges are not exceeded. DO NOT USE LOAD DATA FROM THIS VIDEO! Loads shown were safe in my rifle, but may cause dangerous pressures in yours.
Videos on this channel are for informative, and entertainment purposes only. Using any of the information is at the risk of the individual using the information. We (including RUclips) will not be held liable for any injury or damage resulting from attempting anything shown or discussed in any of our videos. By viewing or flagging this video you are acknowledging the above. 
Спорт
I can't tell you how refreshing it is to see someone actually say this. As an engineer I base all load development on data. To prove to myself, I have chronographed every shot for about the last 650 rounds and put it into a single dataset. I have also seen that with a big enough sample size there are no flat spots, just a linear increase. As an exercise I did what you did but from actual data and found that to have any level of SD and ES confidence I need 27 round sample sizes. Funny how close to 30 that is. Of course we can get pretty close by taking a 5 or 10 shot group and multiplying ES and SD based on confidence factor.
When you throw in the accuracy of ±0.1% on a WELL ALIGNED Lab Radar that is almost ±3 fps. Just look at the differences on 100 rounds chronographed with both a lab radar and magnetospeed if you want to look for sources of experimental error. When looking at sources of error and sample size, I couldn't agree more with your conclusion, just pick what shoots small groups. Anymore I do 5 shot groups to establish a mostly linear velocity curve, find the charge where I start seeing pressure signs, then back off .4-.5 gr. Tuning seating depth and fine tuning with a tuner almost always tightens them up regardless of the powder charge. I look at seating depth as the strong force to optimize and powder charge as the weak force.
Super interesting topic. I love anything that bucks the traditional thinking on almost any subject.
Been thinking about a tuner for a while too. Might have to look for one.
Yup, for a normal distribution, theory says that you need at least 30 shots to have a reasonable confidence on the stats that you have calculated; you can calculate confidence intervals based on sample size and it gets bigger with smaller samples. Ladder test...let's someone have a theoretical basis to give some kind of statistical confidence on the results
Great post. Many years ago I spent a couple of years in Quality Assurance for the Air Force Ground Electronic Engineering and Installation Agency (GEEIA). I developed a great respect for MIL Standard 105 (U can download it). It Does all the math for you. It's been 45 yrs, but I seem to remember a 75% confidence level in the mean of a 25 Piece sample. That said, a couple of F-Class posters, get a lot of mileage from 5 shot groups based strictly on group size. But is't typically 10 groups of 5.
One day my data was all over the place. I put the chrony to the side and just shot groups. While cooling down I saw it. There were small beetles flying around the sky screens. Chronys are great tools but we are shooting to hit a target.
CEP - Circle Error Probable - The military has been doing this for over 100 years. Mostly with artillery. You might be surprised to know that nowadays they even add GPS accuracy to CEP for guided munitions. And it's worse that most people think...
Great video! Thank you for doing this.
It's been hard to explain to people that just because you've found a velocity node, doesn't mean you've found a harmonic/"accuracy" node.
Well said!
The way I see it is if you think you've found a velocity node, you probably haven't fired enough shots at each charge level to get a reasonably accurate estimate of the velocities. The discussion in the video covered this.
You just crushed my illusions… all those beautiful flat spots on Ladder tests…
On a serious note, as someone who had to learn biostatistics in college (and hated it), in the back of my head I always knew that 3 was not a big enough “n” for any SD to be significant.
Well said Keith. Not a condemnation of chronographs, but also a good demonstration of how they have been misunderstood.
Just watched this and I had to chuckle! You hit the nail on the head with a sledgehammer. I am an analyst by profession. I study things in great detail. Made a great living doing that. I also been shooting for longer than most people have lived. I used to reload thousands of rounds a year and stopped. Why? Quality off the shelf ammunition has evolved to a point that I (the shooter part of the equation) can't better that performance. Even with extreme spreads that would make reloaders cry they still shoot bug holes. What's up with that?
Well, because simple velocity stats are much less telling a factor in precision than other more mundane, albeit much more time consuming and expensive problems. In short, the mechanical relationships of how a bullet and cartridge are put together and then interface with the delivery system (aka the rifle). What that tiny bullet meplat is looking at when it leaves the barrel is everything.
Don't get me wrong. I'm a pretty good shooter, but the factory ammo is just that good. We humans are the third leg of the stool, and with all our built-in biases probably the worst part of the triad.
The ammunition manufacturers have in great part solved the first part of that first mechanical duality and get better by the day. Rifle systems are greatly improved, but still not available at mass market prices. And the last leg - the shooter I won't comment about that leg.
Thanks again!
I think it was Glen Zediker (RIP) that said: Results on target are all that matter [for target shooters]
I live his books. I didn’t know he passed. When???? I talked to him once on the phone. Great guy!
Correction. I love his books. Living them would make me crazy lol
Thank you. I've never believed those flat spots from a single shot ladder, because I've also shot 5, sometimes 10 shot, ladders to look to see if they persist and, as you've shown, they evaporate.
This is making a lot of sense. I all need now is a place to a 20 shot test at 600 yards. The only rage in my area that goes that far only has steel plates past 100 yards.
After 28 years of statistical work in production control and quality release I can confidently say you are on to something here. 30 years ago I commandeered a corporation’s computer to “shoot” groups all night which which taught me a lot about the comparison between 3, 5 & 10 shot groups. Keep up the good work!!
I spent one morning with a chronograph and a brick of a popular brand of 22 LR ammo. What I discovered was a distribution of velocities that was not different from one predicted by a normal distribution in a statistically significant manner.
I don't own a chronograph. I have always done as you said- I focus on the group size. I don't care how fast is goes. I stay within the charge weights in the manual of choice, of course, for safety. My reloads are stone cold white tail killers. I would love to someday get in to F-Class. I really enjoy the technical aspects of your videos. Keep it coming!
Well said. I have a 280 Remington that shoots 140 Nosler Accubonds into 5/8” groups at an average of 2940 fps. I have a buddy who is convinced that I need a different rifle since I’m not getting at least 3000 fps. I have to constantly remind him of the mule deer and pronghorn that have been killed by myself and others with that rifle and load.
I love to see some clarity on this. I have beat my head against a wall for a year now using a Labradar chasing all of the "flat spots", graphing every single group at the range. All I have concluded in all that time, is that 90% of the time, my best shot groups are among the most terrible ES/SD groups, so I have begun to err on the side of groupings like you are stating.
I’m glad I found this video. It is exactly what I needed to hear. Thanks..
You sir just earned yourself a new subscriber. This is just excellent info and it makes so much sense. Upload a video yesterday on how you do load development using this method.
Great video. I basically only reload for my hunting ammo and just shooting for fun. I have tried using several methods of load development and have never really felt like they were all encompassing. I've just been trying to do the development techniques others use who shoot a lot more than I do. Thank you for making my process simpler now!
Excellent, thanks Keith !!!
Great video, glad you did the math for us! Looking forward to more Mythbusting
Thanks for doing this video!! I am seeing a lot more discussion about the inherent problems using small groups for zero, group size, velocity nodes, and etc. The information you presented based on normal distribution random number generator matches closely with Hornady podcasts #50 and #52 regarding small group sizes. The Podcast shows how group size grows with sample size, and also touched on velocity being linear with charge weight. Their data was from actual shots fired in their test rig, so it's great to see the statistical analysis line up well with real data.
Well said. To summarize my reply to smoakngun, accuracy at a reasonable velocity kills critters.
What a terrific video; the best I’ve seen on the topic. Five Stars.
I get caught up in chasing small SD’s because when I get that ugly shot, I don’t know whether it's outlier or a flier. I lack the confidence/experience to “know” whether it’s bad luck, a quirk in my ammo, a failure to properly account for environmental conditions, poor technique, or all of it. But I try to learn from shooters like yourself; shooters who are willing to share their knowledge and who are just plain good teachers. Thanks.
==========
This may be too fine a point but ES’s and SD’s are not the same kind of critter. For normally distributed velocities and for a given group size, the mean ES is directly proportional to the population SD, the proportionality constant depending on sample size. This goes for percentiles too. That’s just math. Tell me your SD, your sample size and I know everything about the statistical distribution of ES’s, its mean, variance, and all its percentiles.
A single group only tells you so much. For a single 10-shot group (say SD = 5), 90% of the ES’s will fall between 9.3 and 22.4 fps; averaging over five 10-shot groups, the bounds shrink to (12.5, 18.4) fps. Over many, many groups the bounds close in on the expected value of 15.4 fps (for n = 10)
Similar statistics are applicable to shot placement groupings (usually modeled as bivariate normal, equal variances, correlation = 0). A goal of 10-shots in 2” at 600 yards (~ 0.318 MOA) only goes part way. Since ES is a random variable, you need to state which statistic, e.g., the average? The 95th percentile? It makes a difference.
Take two 10-shot groups. If you wanted 95% to be under 2” at 600 yards (5% fail), you’d need to be shooting 0.255 not 0.318 MOA. For 95% of five 10-shot groups, 0.236 MOA. At the shorter testing ranges I have access to, the differences between 0.318, 0.255 and 0.236 MOA would be hard to see on paper and take way too many rounds to reliably estimate.
As you remind us, with small samples, there is a real danger of over-interpreting differences which may well be random fluctuations. Three and 5-shot ladder tests ought to be prime suspects. If you could shoot more, many of the differences we thought we saw would dissolve.
As Box quipped, all models are wrong but some are useful. The challenge is to decide how best to use our limited data to guide us toward meeting our goals. In the end, just show me the holes.
This is such a good point. Almost all shooters on RUclips treat the extreme spread on 3 shot groups as if it will tell you something meaningful about how a gun shoots. I highly recommend reading the book Statistical Measures for Riflemen and Missile Engineers (It's only 30 pages). It's written by Frank Grubbs who has a PhD in Statistics. It is quite eye-opening. For instance, it lays out the case that extreme spread (which pretty much everyone uses to assess if a group is "good" or not) is one of the least efficient metrics for gaining useful information about how your rifle shoots. It throws out all the data for all but two shots in a group. Mean radius, for instance, is a much more efficient metric, meaning you can shoot less to gain the same confidence in your data. Mean radius is also less punishing than extreme spread when it comes to "fliers" (i.e. shots shooters ignore because it means their $350 hunting rifle isn't "sub-MOA all day")
Plain and simple, thanks!
As an engineer I've always struggled with the "node". It just doesn't make sense in the physical world. Thanks for saying it to the rest of the world.
Not intended to downplay your video, another ballistics group has concluded the same basis truth and published their findings recently, which backs up your thoughts. Great job 👏
There's more going on than just velocity. Too many people in the shooting world don't have a good handle on what actually happens when they discharge a firearm, yet they like to go on forums and spew half-baked nonsense. Unfortunately, over time, that nonsense gets repeated enough that it gets treated as though it's "fact". It takes many good vids/posts (like this one) to finally start to dispel the forum/gunshop dogma.
That is the cold hard truth!!! Too many variables in trying to pin down and understand how to make things work consistently! In the end, that is what I typically do now, know the velocity of the load just for the sake of ballistic calculations! Most of everything else when you miss the shot is more the mechanics of the equipment and fundamentals of marksmanship in my opinion!🙌😄
I could not say that any better - Thank you!
WOW..amazing information.
Holy shit you just blew my mind. Thank you so much!
Thank you!
The funny thing is in your last vid you were talking about the ammo novel....
Then a few hours later gavintoobe had a 4 min vid on one hahaha
I was wondering if that was a coincidence, or shots fired haha.
Supposedly Speedy had a hummer of a barrel on a rifle he shot in the infamous Houston Warehouse that shot low teens group size relatively consistently. The groups from 0.010-0.020” for 5 shots. When they finally chronographed the load it had an ES of 70fps.
Targets >>chrono.
Higher velocities create more drag, lower velocities have less drag; so it makes sense that as long as your ES isn't massive (100+FPS), that it doesn't matter too much, so long as the bullets are leaving the barrel on the same trajectory every time. At the muzzle, the ES might be 70fps, but what is it at 100 yards? 200?
one of the benefits of a stiff, heavy barrel is that they are less sensitive to slight variations in pressures (velocity) than a thinner, more flexible one. Pursuing a small ES and SD is more important for featherweight hunting rifles, than it is for heavy bench guns.
@@wildrangeringreen Agree completely. I also subscribe to the idea of "positive compensation where you can find a node such that the barrel is moving up at bullet exit and slower rounds leave slightly higher than faster ones and this suppresses some of the resulting vertical.
And because faster rounds have slightly more drag, the ES in speed only decreases as the bullets travel.
@@wildrangeringreen ES only decreases with distance because faster bullets experience more drag. In Speedy’s case, it was just a short range benchrest gun and didn’t have enough distance for ES problems to show.
Great vid!
like half way through the vid i was saying "whats the target telling you ?" then you confirmed my thoughts
EXCELLENT !!!!TOPIC AND REVIEW. I also have chased my tail and pounded my head on the table with this issue and Yet the shot groups really tell the story. No doubt if you have an example of a long ES speed , it makes one think how can that look like, say at a 1000 yards. But unless you shoot at a 1000 yards and see what groups your getting, You really don't know what your load is doing. How many times have you not had a consistant velocity and your dope says use so many clicks at 500 yards and you follow it. But yet on your target down range the group is great. And its actually 4 inches low, Despite you followed the tragectory projected by the computer. You just don't know until you shoot and see what's on target.
Good information
TRUTH!!!!!, or at least the truth to me!! It took me a lot of time and a bunch of components to understand this and only after I became so frustrated I just quit for a while. Then a long, long time shooter told me, numbers only matter to math folks and bankers. Trust what is happening on paper.
I’m going though this right now. Two open groups (1/2 moa), 6 small groups (0.3s - 0.1s) in a row, 2 half moa group. Velocity chart looks like a smooth ramp.
I've been reloading 35 years before you were born and I figured that out back then.
Have a cookie.😁
Thanks!
Well said. It’s eye opening to plot this pattern on a graph.
Generate 5 samples from a normal dist, then calculate the SD of these 5 samples and plot it. Repeat this 100 times. Then repeat that using a sample size of 10, 15, 20, 30, 50, 80, 100. The bands narrow around the true population SD wayyyy slower than you want them to.
Excellent video another good recommendation is stay away from the es if you were using velocity but pay attention to what the bullets are doing on target truth I don't know if it's true I don't think it's true about the bullet settling in I've heard that for years but I do my three shot groups at 300 yards gives me a little more confidence also if you have a bad group in the middle of good groups go back and look at it again because even with the best components I believe there is component failure or defects at times sometimes a bad load can turn out to be your best load LOL keep shooting
Awesome video
You are keeping it simple . So I can choose my tightest groups. Shoot at diffrent ranges and I should have my drop. If I want range every 25 yards out to 500 yards. This hunter will only need the range to put the bullet on target.
Woah! Mind. Blown!
My Dad taught this to me about 30 years ago. This video is golden.
.....Same here Sir, except it's been 54 years for me....saying to true EVERYTHING, NO MATTER WHAT !!
People were reloading using chronographs in the 1970’s???
Really good information. I know why the ES becomes larger as you increase the sample size but maybe you can use a bell curve with random points to visually demonstrate how this happens for everyone. I would however have some rules of thumb around special cause variation. If you have hundreds of data points for a given load, you can assume the population and the standard deviation to be accurate. Multiply your standard deviation by 3 and add this to your average for your upper control limit and deduct it from your average for your lower control limit. Any round with a velocity outside this control limit has a special cause. See if you can identify what it is (maybe you cooked until it was crispy in the chamber). If you have eight or more rounds in a row above or below the average without crossing it, you have a special cause (maybe your rounds are all really cold due to bad weather conditions). If you have 4 out 5 rounds at more than one standard deviations away from the average, you likely have a special cause. The same goes if you have 2 out of 3 rounds at more than two standard deviations away from the average. There are more tests for special causes like increasing / decreasing velocities for 6 or more shots in a row, or any other recognizable pattern.
I have personally documented those flat spots and 60% of the time they correlate with the best group in set. A little bit better than flipping a coin, it helps to make an informed decision. Always trust what the paper tells you at multiple ranges.
You've documented those flat spots with how many shots at each charge level?
@@jacquesshellac5311 three, then 5, at multiple temperatures, splitting the grain of charge from 0.2 to 0.5 grains depending on case capacity to verify the shelf. After that, I choot it again to verify primary and secondary zero. After all, hunting and practice ammo doesn't have to be 1/4moa to take deer at less than 300 and I get a stiffy when I see 1/2moa at 500.
Been handloading for 46 yrs. Never had a crony until 5 yrs ago. My load development was based on one parameter. What did the bullets do - at the target. For hunting - killing was the success. And for all those yrs - I wondered if I was missing out. Would knowing the velocity of each shot - make me a better shooter. OR Would it make me gaze at my navel trying to figure if meditation would magically make the Bullet Fairies herd my bullet to the 10 ring better than the next bloke.
5 yrs - from knowing what my numbers were - Has all but silenced my inner question. - Why - ? - Because my grandfather told me that sometimes you are right - and the world is wrong. ----- Be contrary. Feel your truths - and knwo they are right.
Some 46 yrs later - I am using my Chrony - to just give me a glimpse into what my handloads are capable of. But it is within me - and the fairies - to nudge my bullets to the paper.
Thank you Sir! Today - After 44 yrs as an RSO - I took another course - Because - I want to remain teachable - that was my point. For a white haired guy! Cheers!
Thanks a lot.😩 After watching your previous videos about load development, ES & SD, and having spent ALOT of money for a
Labradar, you now tell us that velocities are not important, just look at groups. 😮😬😬
Labradar is still great, it really simplifies velocity and BC calculations (BC changes depending on the velocity and rate of twist, the ones offered by MFG's are usually overly optimistic) , which helps with determining trajectory. Otherwise to calculate BC, you'd need two conventional chronographs and a measuring tape (or accept more error and shoot a few shots with the chronograph close, then move it out to a measured distance and shoot the same number of shots).
As an engineer, this makes a good deal of sense since 3/5/10 shot groups can be bit misleading when it comes to ES and SD. I have a question: how do I find the proper powder load since the flat spots in the powder load plot are essentially meaningless?
Hello, is there a book recommendation you have for load development procedures?
I would ❤️ to see a video about how to correctly Tune ! Steps, Nodes, Shots per group, Velocity, Seating, Harmonics. How many shots before I can begin to tune a new barrel? How can I tune quickly/accurately before the barrel is half worn out? A complete How to Tune correctly video series 👊🏻. I once read a post Alex Wheeler made about how most cut rifle barrels r hummers if tuned correctly.
Saterlee,ladder,OCW, ......which methods are best according to most champion shooters? 🤷♂️
Any F class/Benchrest publications on properly tuning?
Easton arrows used to print a tuning bulletin for archery, it was just a few detailed pages in a pamphlet.
Timing of this vid is prophetic. I just knocked a load up for 600yd BR using my Sako S20 for factory class. Because if lack of time/components I did a powder charge test of 2 bullets each found the best group did a seating depth test shot 3 good groups of 3, 2 being one hole at 100m that'll do me for next weekend
I was just about to spring for a lab radar. 23 year old shooting chrony still works ok but I always wondered about the accuracy of the readings.
Do I understand the ramification that a 5 or 10 shot group of group size at a ladder of ranges might be a better gauge?
Thanks great content. I’ve looked for a Labrador for past few weeks can’t find one anywhere.
I've always enjoyed the statistical nature of the sport. So many people don't have a grasp on that side of things.
🤯🤯🤯🤯🤯
I thought velocity node,low es, flat spot etc, so, less elevation varieties at longer range, then seating depth..
Not too sure if you can, should or are excluding this aspect...
Less load development is attractive though ...
Always interesting..
Thanks .
i started with the satterley method to 'find the velocity node' then the sample size reality punched me in the face. so i moved to the newberry ocw method, to find the barrel harmonic/load combo.
Great video. I am relatively new to reloading, but if I am understanding you correctly. You are suggesting do your "ladder" test with 10 rounds of each charge weight. Pick the load that has the best group and run with it. Then get your velocities and use them for your ballistics calculator. Am I tracking?
Not quite, the 10 round comment was aimed at dispelling the idea of velocity flat spots. I’m working on a video now on the load development process that I use.
Where is the link to the charger magnetic swivel you recommend for the labradar, , have looked everywhere but heard you know what one to use. thx so much, love all your videos.
I didn't recommend one. I've never tried one. Might have been another RUclipsr?
You're gonna hurt some feelings with this one.
This all makes sense BUT if you get a flatter spot in the velocity isn't this better for temperature variations having a wider powder charge to keep the velocity in your range for ballistic calculations? Think I've missed something in your video. Great content as usual just trying to understand your take on this.
First? :)
I've never had a chronograph. So I always shoot a ladder test in .1 gr jumps at 200 yards and record where each shot lands. I've actually thought that F Class shooters could do the same and probably get good information. But they would do it at 600 yards.
That is exactly how I do my own testing. :)
@@winninginthewind good to know I can tell my buddies who don't believe it's a good test :)
.5 inch groups at 200 yards from a factory Savage 12 223
Consultant results don't lie.
Damn good advice. Load development should be results-oriented. Velocity is just a statistic.
Find the speed you want, tune seating depth. That’s what I’m getting out of this.
Thank you! You just made shooting more enjoyable…. 😊
Ouch, that hurt my head
so far I've only used factory rifles however I watched lots of videos and kind of listen to what some accomplished shooters were doing find my powder charge don't even worry about a Target I do still shoot one just to get the practice but I don't pay attention to groups finding my powder charge and I've noticed after doing that sometimes I move seating depth over 100,000 before I find a seating depth that shoots good and is repeatable and I've accomplished some really good loads that shoot ridiculous however it seems extremely difficult to get them to shoot good doing it that way example the 6.5 PRC I'm currently working on powder node was right toward the very top however poi and group size drastically changed right before that charge my thought would be pick a powder charge based on group size and poi but I was listening to what everybody said don't pay attention to groups you make it shoot with your seating depth maybe it works better in custom rifles I think I'm starting over and I'm going to try exactly what you're saying I'm going to watch what's going on with my ladder test three shots each charge when I have three .3 increase in a row poi the same that seem to group okay I'm going to pick the middle of that and then start with seating depth I'm guessing I get my load a lot faster this time and I'm extremely cautious that every powder charge is perfect two scales including a beam scale every time I will remove or ADD one kernel I know I don't need to do that but I do and can for the amount I shoot so my SD and es should be fairly good no matter what if I've done everything else correct
You should try using punctuation. Maybe proof read and edit.
Good grief now my brain is a scrambled mess. I thought I remembered seeing something about ES and SD being a indication of reliable ignition and combustion and focus on these to fix those before groups. Think Cortina said this but honestly I haven't a clue. I've tried to get good 5 shot groups with low ES and SD . If I'm understanding what you're saying we don't have a large enough sample size to really determine what matters so if the bullets hit the target then who cares. Am I on track? Thanks for video
Very interesting. I wonder if you did the same theoretical test but with an Thousand round SD of 10 and an SD of 20? What I’m gathering from your video and my own extensive chronographing is that a 5-10 round SD of 0- 2 can actually be an SD of 5-7 and vice versa with occasional round with a high or low velocity out of the norm. This is a great video keep up the great info 👍.
Changing the SD to 10 in my software resulted in bigger numbers, but the same effective outcome. It is very difficult to tell which 3, 5, 10, or 20 shot sample is better or worse... There is still a 50% overlap in the ES on 20 shot groups and a smaller overlap on 20 shot SD. For 10 shots, you would only have about a 50% chance of identifying that the SD had doubled for the population. In other words, you might, or might not get lucky and have it show you the truth. If you shoot enough big samples, then yes, you will see the difference, but a 5 to 10% change in SD will be very hard to see.
Thanks for the response 😀
Bottom line for me has always been trueing my loads just as trueing my dope. - Period,.....even though I do log velocity/es & sd's for comparison in weather conditions & other variables for the day. I have been laughing at these young-gen,-know-it-alls/johnny-come-lately's for the last 1/2-dozen years when one says something like "your es & sd's have to be this or that", or "get them es & sd's down to gain accuracy" .... to some dude that doesn't know better but is shooting as good as the know-it-all......LMAO.....guess all those increase in vax's ARE taking a toll......Thanks fer generously sharing your updated insight, experiences & experiments....."A Master never stops being a student".
and ur EXACTLY CORRECT Keith....stated simply : velocity is ONLY used to help you figure out where the "Pressure Peak" is, (or where you WANT it to be), and the only thing that matters when shooting, is where the bullet WENT...everything else is beside the POINT.....OnWard.....
👍I do agree with Keith's analysis about not to draw conclusion on small sample size. We all know the cold bore shots should not be counted. 👍 However, the value of knowing ES and SD should not be understated. ❓Question is: What is the acceptable max ES and SD to begin with for a match? No mater how big is the sample size, if ES is >40, then maybe it is no starter. Thoughts?
Interesting video. Your going to get some good and bad comments now lol. Keep up the hard work and thanks
That explains a lot 😩
What!? Your telling me those little holes in the paper are more telling than the almighty chronograph? Well said, and I’m guilty….
Ok, so where does this leave us on load development? Do we pick a charge that provides a velocity we desire and then do a bullet seating ladder test, or say screw it and pick a charge seat to recommended OAL and then use a tuner?
Targeting a certain velocity is sort of useless. You could be in a bad spot harmonically in the barrel and will drive yourself nuts. Look for shots with the near same vertical at a distance of say 300 to 500 yards. If you are doing a ladder test and have a 0.5 grain spread with the same near vertical impact on target, and it's repeatable, that is your harmonic happy place. I usually use three shots per charge and shoot them round robin style. Also, use the same point of aim for every shot! Use seating depth first to tighten group size, then play with the tuner. This may be near the velocity you want or may be 50fps different but you will have an accurate reliable long range load.
Reminds me of a write up in an older Speer manual. "Why ballisticians go gray".
How many groups where sampled per population size?
My understanding is and was: If I have found a V0 range in which the barrel harmony is very good and the hits are good, the fluctuation of the V0 does not matter. in that range On the other hand, a V0 fluctuation that occurs can also be an indication that I am working imprecisely when manufacturing ammunition. I would also like to monitor / determine the weather conditions / temperature that influences the V0 and ballistics. That's my point to messure V0 each time and write it down - not to be the SD chamion. Am I wrong?
I left a message of the recent Hornady podcast and since Eric Cortina had Jayden on, it would be great if somehow you and Hornady had a mathematical talk it out. Since I'm fairly sure you have reached a similar style conclusion via different methods as Hornady has. Your experience and manufacturing tolerances shorten testing methodologies making what you see more instantly valid than those that are just starting. It would make some people rage and people like me happy. I hope it happens.
The proof is on the target. In the 1970's I had a 6mm Remington 788 I reloaded for that shot 1/2 inch groups a 100 yards with neck reaming, Seria MachKings, weighed powder changes from the book, and little else. Statistics is a tool to help us understand what is hidden in the data. Few of us know statistics well enough to fully use it. Small groups are easier to understand. If we had chamber pressure data to go with velocity and group size, the uncertainty would be smaller but still have the same problem with flat spots in load ladders of three and five-shot groups. I think there are too many factors in accurate loads for the velocity you be the only answer. I believe every powder, primer, bullet, cartridge, and rifle has an ideal load. The barrel needs to be tuned harmonically for the best accuracy. The most accurate ladder load is likely to fall on the most accurate harmonic node of the rifle barrel.
When you find a load that shoots a half minute of an angle, it doesn't make much difference why if it is repeatable.
Just get it said you are moving to the statistical concept of the Central Limit Theorem (CLT)
I understand what your practicing at one said distance, but at the end you mention ballistics, so would this load development work if used fo random shooting distances?
I have not found tune to be terribly range specific. In my experience, if it shoots good at long range, it will also shoot good at short range.
Amen.!.!!
Statistics are only descriptive approximations, not reality. The holes in the paper are real even if its only three shots. The odds that a three shot group is representative is far greater than it being not representative.
thats how the guys that I shot with used to develope our loads, chronographs were rare and expensive back then, so you would pick a load from the reloading manual for the bullet\cartridge you were using load 5 send down range if they grouped you stuck with that load otherwise you tweeked it up or down till you got it to group. Not very scientific but effective.
So this concludes there isn't a magic powder amount that settles velocities down. Obviously consistent velocity is important and the farther the target the more it matters. What are the factors that help achieve consistent velocities?
That's a later video... :)
Finally! Someone speaks the truth about velocity ladder testing. It is just an exercise in idiocy, performed by people who have no understanding of random processes. I did this same thought experiment with a spreadsheet populated with 10,000 gaussian random numbers a few years ago and figured out that you need a much larger sample size to have a reasonable confidence in the result. I use 100 shots. I am waiting for a proponent of velocity ladder testing to answer the following two questions: 1) If adding (or subtracting) .1 gr of propellant increases (or decreases) velocity at one point on the curve, why doesn't it add (or subtract) velocity at a different point on the curve? and 2) Why do I care about what happens with a charge that is +/- .1 gr from what I want when I measure to an accuracy of .01 gr? I will NEVER drop a charge that is +/-0.1 gr off. Thanks for the great video, Keith. Yours is the only shooting channel I have found that does anything to dispel the many myths around about shooting/reloading.
I wonder what would happen if you also add in the error distribution from the chrono ... 🤣
I'm not sure that I follow your concern over trying to get a standard deviation, especially at a velocity nearing 3,000 ft per second less than 30 ft per second. I don't see that that's an issue. I have seen 30 ft per second, and even greater than that with loads that shoot a consistent quarter inch group. My very first Centerfire rifle, a thin Barrel Ruger Model 77 in 7 mil mag I shot consistent 20 round groups into a center to Center spread of 049 in on a number of occasions. I wasn't even concerned about standard deviation or extreme spread. When I started looking at those numbers, because I was shooting 140 grain Spire point at some 3,200 ft per second or more. I'm looking at less than 1% if it's at 30 ft per second. And that's not going to be enough difference to worry about. I might be concerned if I'm shooting something like a pistol that is running maybe at 12:50 and I've got a 30 to 50 ft per second spread, but with a pistol I'm usually looking for tactical accuracy. So as long as I can keep a 4-inch Group at 100 yards with my 4-in 357 I think I was doing pretty darn good. Admittedly I got spoiled really bad with my first Centerfire rifle. But I was given the load data by at the time a 200 yd bench rest champion, I believe his name was Russ Hope. He said to use virgin brass and load it with I believe it was three grains over the max load of h4350 and he said the advantage to that particular powder in 1979 was that you couldn't overload it because it was the slowest powder available at the time I believe. And it should be quite accurate. And then when I reload it just neck size it or if I full length size it do not push the shoulder back at all or just barely bump it. But I neck sized it and it worked great. My next projects are to take my four 308s, I have a sr-762 with a 16 inch barrel. An M1A. A 308 Ruger Model 77 with a heavy barrel. I believe it's manufactured date was 1972. And a 308 lever action Sako Finn wolf. A rifle that Winchester copied to make the model 88 as soon as the patent run out, within 10 days they had the model 88 model 100 heading out on the market. I want to try a 150 to 155 grain bullet that is supposed to be being manufactured right now, but otherwise I think I'll stick with the 168 grain match King hollow point. Especially for the M1A because that bullet was Taylor Made for the M14. See how they react with that particular projectile and start working them from there. If I can find a single load that will work on all four of them I will be pretty happy. I don't expect all four of them to be particularly good with a single load however but you never know I might get lucky. I'm looking for a 1200 yard capability with the sr-762 and about a 1400 yard capability with the rest. Possibly more. I know that the transitional range on the 16-in barrel will probably be around 600 yd and the rest of them should be a transitional range somewhere around the 800 yard mark. And the Finn wolf I believe favors 150 grain projectiles or somewhere in that vicinity. So I may be stuck with something but there is supposed to be 150 to 155 grain projectile being made with a better BC than the 175 grain Sierra match King. But I don't know that they ever got it out I have to start doing some more searching before I start my loading constructs.
My world is falling apart.
IRL you have mixed distribution more often than not
You just saved me a lot of money. Im just going to get back to shooting. I've won more matches just keeping it simple.
Well I hate to agree but I think I do
Now you are where I arrived over 30 years ago.
Been with you a while, but had to follow the herd. I decided that I'm a donkey among zebras...
So you're suggesting to not focus on velocities, but on the effects we achieve on the target, as we do in the times when there were no chronographs and other measuring devices.
Back to old school methods 😉
Many shooters won the match by reloading ammunition without these advanced electronic devices. It seems that someone "convince us" that these devices are necessary for us to live... just to make a good deal out of it 🤔
When you said, the sample size was 3, 5, 10, 20 what was the sample mean Velocity and SD? Why didn't you compare it to MV and SD of the random sample from the population?
Firstly congratulations doing this video. Sadly it’s a modern trait to ignore the wealth of written information readily accessible (for those who know what a book is) regarding statistical probability as it relates to ballistics & in particular rifle performance - knowledge of which would have shown the futility of most of the latest ‘fad’ load developments methods.
But doesn't velocity help with finding your powder load
Yep. The holes in the paper are the only things that matter in the end.