Number of shots used for Load Development

[Triffid]

Hornady have produced a series of podcasts about various aspects of shooting & reloading. I linked one on Barrel Tuners which has prompted some interesting comments.

 

Delving further, I've come across a series looking at the impact of sample size on load development, which pretty much rubbishes the normal ways I've used (eg Ladder / OCW). The podcasts are long and detailed, but my takeaways from it are:

1. To get a representative idea of group size, you need to fire about 30 shots.

2. Conclusions based on smaller number of shots (eg 3 and 5 shot groups/velocities) are invalid. Using exactly the same load, you are likely to shoot very different 3 and 5 shot group size, velocities and/or zero position.

3. You can use small samples to identify 'bad' loads, but not 'good' ones; if a three shot group measures 3MoA, it cannot get any smaller. But a group measuring 0.5MOA can only stay the same or get bigger.

4. Their load development process consists of throwing different combinations of components at the rifle to see which ones work. No powder charge work-up or seating depth changes; with large sample numbers they find that there are no 'nodes'.

 

I'm going to try shooting ten five-round groups of my favourite load . . .

 

Triffid

 

 

 

[jcampbellsmith]

https://eley.co.uk/importance-of-eley-batch-testing/

Regards

JCS

[Big Al]

On 1/23/2024 at 10:24 AM, Triffid said:

Their load development process consists of throwing different combinations of components at the rifle to see which ones work. No powder charge work-up or seating depth changes; with large sample numbers they find that there are no 'nodes'.

 

There is an old saying, "if you put shite in you get shite out"

If someone tries to randomly develop a load as you suggest then they end up drawing the wrong conclusion that there are no nodes.

Listen to people who shoot really small groups regularly, dont waste your time or expensive ammo on stuff like this.

[Popsbengo]

On 1/23/2024 at 10:24 AM, Triffid said:

Delving further, I've come across a series looking at the impact of sample size on load development, which pretty much rubbishes the normal ways I've used (eg Ladder / OCW). The podcasts are long and detailed, but my takeaways from it are:

1. To get a representative idea of group size, you need to fire about 30 shots.

2. Conclusions based on smaller number of shots (eg 3 and 5 shot groups/velocities) are invalid. Using exactly the same load, you are likely to shoot very different 3 and 5 shot group size, velocities and/or zero position.

3. You can use small samples to identify 'bad' loads, but not 'good' ones; if a three shot group measures 3MoA, it cannot get any smaller. But a group measuring 0.5MOA can only stay the same or get bigger.

4. Their load development process consists of throwing different combinations of components at the rifle to see which ones work. No powder charge work-up or seating depth changes; with large sample numbers they find that there are no 'nodes'.

 

I'm going to try shooting ten five-round groups of my favourite load . . .

Your points 1, 2 & 3 are valid summaries from the videos.  Point 4 is completely unfair   to suggest "throwing different combinations etc"

Statistics and the way they apply them in the videos is sound.  It points up the nonsense of people quoting SD from a five shot string (or even 3 shot as occasionally posted on this forum).  I don't think they are suggesting that load development can't improve the consistency of ammo, it clearly does - it's just how shooters fail to grasp the reality of small sample statistics error bars and say "oh cool I've got a great shooting load" after a good five shot or even five by five shot NRA (US) grouping.  For sure 5x5 is realistic - it's what I do, it's just not a great predictor of possible divergence.

The thing that gets glossed over in these discussions is the variability of the shooter him/her self.  I don't know about others but I can't hold point of aim like a bolted down bench gun.  And that variability is effecting group size but not chrono data.   Good consistent well made ammo on the chrono, wobbly old me at the trigger.  Group size not relatable to just internal ballistics effects but muddied by inherent inaccuracy of the nut behind the butt.

[furrybean]

Do you ever stop through the life of a barrel 

[Big Al]

4 hours ago, Popsbengo said:

The thing that gets glossed over in these discussions is the variability of the shooter him/her self. Group size not relatable to just internal ballistics effects but muddied by inherent inaccuracy of the nut behind the butt.

 

As a young apprentice some 40yrs ago my mentor and still dear friend today told me about the 5% rule.

As an apprentice of the year he explained to me that only around 5% of people who take part in anything end up being truly good at it, some 40yrs later after many different experiences I am a firm believer in this concept. You only have to take golf as a good example, statistics show very few players end up scratch golfers and the handicap system worldwide clearly shows the average, the good etc etc. Without any doubt this also transfers to shooting of all disciplines.

The idea being that 5% end up very good then a slightly larger percentage end up good and average, poor, choose a new hobby etc.

I have certainly found this to be very true when it comes to shooting rifles and load development. We might all be able to shoot a good 3 or 5 shot group now and again and blame poor ones on 'flyers' but far far less of us can repeat this with such regularity that we can then trust our results enough to truly draw accurate conclusions from our testing. I agree that lots of people wont consider sample size and its value, one small group and its called good when statistically its far from it.

That said, people can only do what they can and as such not everyone ends up a winner.

[Big Al]

4 hours ago, furrybean said:

Do you ever stop through the life of a barrel 

 

If you want to be truly competitive you cant, barrels and loads change throughout the barrels life and the component batch size. Environmental factors could also change a load month by month. Of course it depends on how competitive the shooter wants to be.

[furrybean]

Exactly that 

[Popsbengo]

40 minutes ago, Big Al said:

That said, people can only do what they can and as such not everyone ends up a winner.

Success at competition certainly is a good measure but I'm sure you'd agree that not everyone that's good at something competes.   Also, not everyone that competes is necessarily good at it either.

[Big Al]

1 hour ago, Popsbengo said:

Success at competition certainly is a good measure.

Yes I think it is but I would say I would be more likely to listen to people who might have had continued success over a reasonable period of time. Those are the people I would be listening to.

1 hour ago, Popsbengo said:

I'm sure you'd agree that not everyone that's good at something competes.   Also, not everyone that competes is necessarily good at it either.

I would agree 100%.

Of everyone who competes in most things the 5% categorisation I mentioned earlier will apply. So very often though its the same names at the top of many leaderboards which seems to confirm that.