PF margin of error

[tollarja]

Hey guys, I am working on ensuring that I will make minor for the Northwest challenge. I weighed 20 Montana Gold 124 grain bullets, found the standard deviation and average. I then chrono'd 19 rounds found the average and standard deviation of these. To Calculate my PF I subtracted 2 standard deviations from both averages and then multiplied them together for a PF of 128. Assuming the distribution of velocities and weights are normal, as they appear, I feel confident leaving the load as is.

My reasoning is that by subtracting 2 standard deviations from both bullets and velocity I am using two values that I am 95% certain will be the lowest likely. The numbers I used were in fact less than the lowest values recorded.

I wanted 130 PF, but I feel confident that the chrono procedures used are very much in favor of the shooter and that the risk of being below 125PF is very unlikely in my case. I am also shooting clays which makes me a bit nervous as the weather improves and temperature increases. So far there are no signs of overpressure, but I don't feel the need to push it.

Is my reasoning sound and would you feel comfortable with this margin?

Thanks in advance,

Jeff

[Sarge]

In my opinion:

128 is not enough cushion. Try to stay above 130. I like closer to 135. Some even like 140.

You need to chrono alot more than 19 rounds. I would run several strings of 10 or 20. I keep track of all the lowest fps in each string then figure out what my pf would be if the chrono guy were to get them. They should be nowhere near 125 for me to be comfortable.

Remember, at a match you will pull random ammo to put in a bag for the chrono stage. They will weigh one and use that for your weight. MG bullets should be very consistent in weight.

I try to make the chrono a non issue for a match by playing it way safe. It baffles me to see shooters sweating bullets over the chrono.

It is also wise to put up a 25 yd target to shoot at when chronoing to see how the load shoots. A good soft load that can't group worth a crap is useless when a little hotter(probably not even noticable)load that groups 2 or 3 inches at POA is what counts.

[Skydiver]

Jeff is not saying that his Average is 128 PF.

Jeff is saying that his (Average - 2 * STDDEV) is 128 PF, not that his average is 128 PF. This means that 97.5% of his rounds will all be 128 PF or higher assuming a normal distribution. (For the stats challenged like me, it took me a little while to realize that 2 * STDDEV means that 95% of population will clustered around the average. This leaves 5% away from the average. Half of the 5% will be on the low end, and the other half will be on the high end. So that's how I came up with 97.5%)

So assuming a STDDEV of 2 PF, that would then mean that Jeff's average would be 132 PF.

Edit: Adding this link:

http://en.wikipedia.org/wiki/Standard_deviation#Rules_for_normally_distributed_data

The graph also made me realize that based on the assumption the STDDEV is 2 PF, that would then mean that about 99.74% of the rounds would be above 126 PF.

Edited May 2, 2010 by Skydiver

[BayouSlide]

Based on my experience, it is not unusual to see a match chrono vary plus or minus PF 3 from what I get with my chronos. If you use an infrared or chrono box with lighting, YMMV. I would not be comfortable with a PF under 132, but that's just me. Then again, discussion of SD makes me

Edited to add: the necessity of making minor becomes even more important with the realization that, if you fail, you're shooting the match for no score.

Curtis

Edited May 4, 2010 by BayouSlide

[G-ManBart]

I use this example a lot (second time tonight!), but it's absolutely true, and I kept pretty good records.

In 2008 I loaded a large batch of ammo for Nationals, a Sectional and an Area match (along with a couple of smaller matches). In testing (think it was ten different days) prior to any of the matches, it went 173-175PF with boring consistency. At the Sectional and Area match it stayed in that range (I log these in my load book). At Nationals it went 166.1PF If I had been at 170PF, I would have gone well under Major. That same batch of powder, primers, bullets, cases loaded on the same machine without a single change went 173PF at the 2009 Nationals. I'd love to know where that 70fps went! Go figure...crazy, but it happened.

Sub-Minor means shooting for fun. There's no benefit to having a load at 130PF when you can run it at 135PF, have it feel the same, and have virtually no risk of failing to make Minor. My current Minor load is in the 136-137 range and it's not noticeably different from when I was at 129PF working up.

I also do multiple 20-shot strings with new loads. After I've tested a given load a few times I may only do one 20-round string if I want to spot check a batch of ammo. If I change lots of something, it's back to at least two 20-round strings. R,

[Skydiver]

I'd love to know where that 70fps went! Go figure...crazy, but it happened.

According to Appendix C2, 18.b, match chrono's are allowed to have -/+ 4% variance from the daily average.

Doing the math on your example above, it looks like you had 100 gn bullets going between 1730-1750 fps. 4% if 1750 fps is exact 70 fps and 4% of 1730 is about 69.2 fps. Assuming worst case of 1730-70 = 1660 fps with a PF of 166 PF. Definitely within the realm of possibility to lose the 70 fps.

[CocoBolo]

Minor is certainly a different story than major. With major if you go minor you are still in the game, but if you fall short with minor you are out, for score. My last 10 major matches including the Nationals I have found that match crono's varied about 20-30fps and that is + or -. The same load in my Limited gun was 189pf at the Gator and 171 at the Nationals. Temp was a factor at the Gator.

Shooting minor with a non compensated gun I try to run at least 130pf and with a compensated gun I push that to 135/138pf. The numbers I try to get are at temps I expect at the match. That is why it is very important to record the ambient temp when you crono. I don't think I have found any powder that is not some what sensative to temp extremes.

I shot my backup gun in the last major just because the primary was having trouble getting to 175pf, I wasn't risking it with a gun that was only capable of 168pf, I have two new barrels for it and it is going in the shop next week. My 175 turned into 172.25 at the match maybe the other gun would have been 165.2 too close. On Saturday they dropped the crono, but you can't count on all of the crono's getting shot. A number of level II matches have dropped the Crono during some part of the match, but never had a Level III without crono.

[Sarge]

While I admire all of the science and engineering skills demonstrated in this thread I just want to say that when you get to a major match the guy running the chrono could care less. He is going to load your gun and shoot your ammo. It's pass or fail!

Speaking of percentages of error, margins etc. Loading on a progressive machine(550 here) the margins start early. The powder throw can vary by .1 then the scale can vary what it weighs the charge at by .1 Then the guns velocity can vary on any given round. Then the chrono can vary by as much as several FPS from shot to shot.

Try this test. Some day when you are bored throw about 50 strings of 10 powder charges each and see how frustrating it can be to get the same EXACT reading every time. Then ask yourself, "was the charge off by .1 or was the scale off by .1, or were they both off?"

Then go chrono and see the fluctuations in velocity and try to figure that out. Was it the chrono, the gun, or the round that caused the change?

Then go to a match that is 20 degrees warmer than when you chrono'd at home and start wringing your hands on what should be the easiest stage of the match.

Trust me I am not knocking the techniques being discussed by the OP. But if I were running the chrono and you showed me your graphs and calculations to explain your rounds not making PF. I would very politely say, "Those are nice, you just made 124 PF." "NEXT SHOOTER!"

Load to 135+ for the conditions you expect at the next big match and relax. Of course you still might go sub minor!

Edited May 2, 2010 by Sarge

[G-ManBart]

I'd love to know where that 70fps went! Go figure...crazy, but it happened.

According to Appendix C2, 18.b, match chrono's are allowed to have -/+ 4% variance from the daily average.

Doing the math on your example above, it looks like you had 100 gn bullets going between 1730-1750 fps. 4% if 1750 fps is exact 70 fps and 4% of 1730 is about 69.2 fps. Assuming worst case of 1730-70 = 1660 fps with a PF of 166 PF. Definitely within the realm of possibility to lose the 70 fps.

If that were the case I wouldn't have been allowed to declare Major PF USPSA has a minimum of 112gr for Major in Open. I'm running 115gr bullets at around 1515fps. Still, it's within the realm of possibility...a couple of unusual low rounds, some weather change, a different chrono etc...any and all could have been part of it. R,

[BritinUSA]

We've had similar discussions here before; The problem is not so much the variation on the chrono that you test your rounds on, the problem is the variation between different makes/models of chrono and even the same chrono set up in series like they have at Nationals.

There was a 5 page thread on it (LINK)

[tollarja]

Aye Ca rumba! There appears to be a lot of error possible from chrono to chrono. Maybe I'll shoot for a minus 2 standard deviation PF of 132. I really noticed the extra 0.2 grains of powder, it made the gun more unpleasant to shoot. I should probably stop being a wuss and bump it up some more though. For further discussion my average PF is 131 at 54 degrees Fahrenheit.

Thanks for all the replies guys.

[pjb45]

I have loads that go back to 05 Nationals, Area 2 and WSSSC matches. I never chrono a large number of rounds. I chrono three rounds at a time. That is how they run the Chrono at those matches. I try and duplicate their procedures. FWIW: 1.18 AOL, mixed brass (once fired), mixed primers, Zero 180JHP, and 4.85-.489 grains of VV320. Just last weekend I chrono'd at 171 at SSNats. At home the chrono was 169PF.

I generally weigh 10 rounds of powder then average it. But if you record each powder charge you notice a variance in weights.

I always keep a control group of ammo that I know how it measured at Nationals/A2 then run three rounds through the chrono. Then I run my new match ammo. If there is much of a difference, I make changes. Fortunately, my recipe has not changed and it always make PF.

Again, I only do three. If you do twenty at a time you smooth out any deviations that may occur, imo. So that is why I duplicate the established chrono process.

Another reason I use VV powder, even thought it is more expensive, is they keep excellent records for how each lot compares to their master lot. If the variance it out of their tight norms, they send it back. So for the last five years, through various lots I get very little variance in my recipe.

Quite frankly, the variance in bullet weight has had more of an impact on my PF than anything else. (This is also true for my MG bullets).

I have talked with many GMs and most try to do minimum PF + 5 as a safety margin. A few try come to close and have ended up shooting minor. The +5 PF seems to be the norm for most.

[WDB]

I usually try for an extra 7-10 pf. Must be my chrono I don't know, but I last summer I was getting 171 and at NC and VA-MD matches I was 166. Once that buzzer goes off you won't know the difference between 128 and 132 or so. What's the old saying, It's better to have and not need than to need and not have.

[G-ManBart]

Aye Ca rumba! There appears to be a lot of error possible from chrono to chrono. Maybe I'll shoot for a minus 2 standard deviation PF of 132. I really noticed the extra 0.2 grains of powder, it made the gun more unpleasant to shoot. I should probably stop being a wuss and bump it up some more though. For further discussion my average PF is 131 at 54 degrees Fahrenheit.

Thanks for all the replies guys.

There is actually so much variation from shot to shot that it makes it hard to tell how much error there is in the machine. I do know that I've frequently seen the higher velocity on one chronograph (big match) for one shot and then the next shot the other chrono have the higher velocity.

If you shot a couple hundred of the load with the extra .2gr it wouldn't be something you'd even think about...it sort of just goes away. R,

Edited May 4, 2010 by G-ManBart

[tollarja]

I have loads that go back to 05 Nationals, Area 2 and WSSSC matches. I never chrono a large number of rounds. I chrono three rounds at a time. That is how they run the Chrono at those matches. I try and duplicate their procedures. FWIW: 1.18 AOL, mixed brass (once fired), mixed primers, Zero 180JHP, and 4.85-.489 grains of VV320. Just last weekend I chrono'd at 171 at SSNats. At home the chrono was 169PF.

I generally weigh 10 rounds of powder then average it. But if you record each powder charge you notice a variance in weights.

I always keep a control group of ammo that I know how it measured at Nationals/A2 then run three rounds through the chrono. Then I run my new match ammo. If there is much of a difference, I make changes. Fortunately, my recipe has not changed and it always make PF.

Again, I only do three. If you do twenty at a time you smooth out any deviations that may occur, imo. So that is why I duplicate the established chrono process.

Another reason I use VV powder, even thought it is more expensive, is they keep excellent records for how each lot compares to their master lot. If the variance it out of their tight norms, they send it back. So for the last five years, through various lots I get very little variance in my recipe.

Quite frankly, the variance in bullet weight has had more of an impact on my PF than anything else. (This is also true for my MG bullets).

I have talked with many GMs and most try to do minimum PF + 5 as a safety margin. A few try come to close and have ended up shooting minor. The +5 PF seems to be the norm for most.

When you take more samples you find out more values and where they are around the average. When you calculate the standard deviation multiply by 2 and subtract you determine the lowest value with 97.5% chance that a value will be higher. The chance that you will get three values that are lower is 0.0015625%. So yes there is some smoothing, but you get a very precise and conservative calculation of where you stand. In my case the values I used for my calculations were lower than the lowest values I measured. My average measured PF is 131, but the one I am 97.5% certain will be the lowest is 128. I do not know how precise the chrono I used was so my conservative PF should probably be lower.

Now another problem seems to be that the standard that is required by the rules is more precise than can be accurately measured which is a huge no-no in math, science and engineering (my major) and I really think this should be changed. All it would take is to multiply the PF by say 1.04 if Chrono's tend to be off by 4%.

[G-ManBart]

Again, I only do three. If you do twenty at a time you smooth out any deviations that may occur, imo. So that is why I duplicate the established chrono process.

Paul,

You can't take the most anything in a vacuum and only measuring three shots leaves a lot of vacuum! Long shot strings certainly remove the influence of a couple of low or high shots, but they do give you an average of what to expect on any given shot. By taking only three, you have no idea if you happened to get three high shots, three low shots, three in the middle, or some sort of mix. The longer strings let you see patterns that aren't visible with three shot strings. There isn't a need to duplicate their procedure if you've already run a long string, just take the first three and get a number. What I think the smart thing to do is to shoot the long string, take the three lowest, and make sure that even those low three will make PF plus a reasonable safety margin. That way you're never surprised by a coincidence a couple of low shots in a row at a big match. By testing those long strings you already know what the lowest of the low should be. With three shots....it's much more of a guess because the sample size simply isn't large enough. I'm sure someone with a statistics background would do a better job of making it clear, but that's the general idea.

My most recent chrono session with the Open gun and load that generated the numbers I mentioned above is a good example. I ran two 20-shot strings. The first averaged 1506fps (173PF). The second averaged 1498fps (172PF). The lowest three in the first string were 1482, 1489, and 1493 (Avg 171PF). The lowest three in the second string were 1478, 1486, and 1489 (170PF). The first three in the first string were 1501, 1509, and 1501 (173PF). The first three in the second string were 1497, 1499, and 1494 (172PF).

At the very worst I averaged 170PF....but I know that's the worst since I put enough rounds through it to be sure I'm not likely to see anything lower. If I was only going off the first three I'd think that my load was really 172 or 173PF when it really could be 2 or 3 points lower if I got unlucky at the chrono.

Granted, with a big enough margin (like 173+PF) it doesn't make much difference, but for the folks trying to get right in the 169-170 range, a point or two might make a big difference if they get unlucky with a few low rounds in a row. R,

[tollarja]

Oh, just so you guys know I'm listening, I am going to test another batch with a little more powder.

[Sarge]

Oh, just so you guys know I'm listening, I am going to test another batch with a little more powder.

I think that is a wise choice. There will come a point even with a minor PF load where you will start to feel the softness start to get more harsh and therefore more difficult to get control of and score good hits. I had loaded up some ammo for Area 6 that turned out to be just too low for my liking (126 ). I was in a time crunch so I loaded up some loads I used to use in a totally different brand of gun. I knew they were safe and I knew they made a reasonable pf. Turned out to be 141 in my G34 at Area 6. I know some will say it is a puss load anyway but a 15 pf difference is a handful especially at a major match for the first time. I know it hurt my performance at least a little.

[tollarja]

Just tested another from 4.0-4.2 up from 3.8 grains. I can't say that I like it any better. Not too bad except for weak hand recoil movement is significantly more.

Edited May 4, 2010 by tollarja

[Skydiver]

When you take more samples you find out more values and where they are around the average. When you calculate the standard deviation multiply by 2 and subtract you determine the lowest value with 97.5% chance that a value will be higher. The chance that you will get three values that are lower is 0.0015625%.

Based on Linda's advice that shooting the last bullet is always better, to fail chrono you'll need to have 5 out of 7 bullets go below the Power Factor floor. The probability for that would be 2.5% raised to the 5th power, or 0.0000009765625%.

Of course, all this depends on the match chrono measuring velocities close to what you were measuring at home. And therein lays the problem because of the variance is chrono accuracy.

Now another problem seems to be that the standard that is required by the rules is more precise than can be accurately measured which is a huge no-no in math, science and engineering (my major) and I really think this should be changed. All it would take is to multiply the PF by say 1.04 if Chrono's tend to be off by 4%.

I agree. I wonder what it will need to be done to have this added to the USPSA chrono rules.

[G-ManBart]

Now another problem seems to be that the standard that is required by the rules is more precise than can be accurately measured which is a huge no-no in math, science and engineering (my major) and I really think this should be changed. All it would take is to multiply the PF by say 1.04 if Chrono's tend to be off by 4%.

I agree. I wonder what it will need to be done to have this added to the USPSA chrono rules.

Not going to happen. If someone made 159PF and they multiplied it by 1.04 they'd just get over 165PF to make Major. That would mean the new PF floor was really 159PF, not 165. Some competitors are going to look for every "advantage" they can find and if they can run 159PF rather than 165, they will. It would do the same thing for Minor, but the new number would be 121PF rather than 125PF.

The answer isn't changing the rules, or using sophisticated processes to predict what your load should be doing, it's simply loading with a margin of error sufficient that it's not a problem. If you load 7-8PF higher than the limit, you'll be fine...easy, quick, simple and it works regardless of weather, bad luck, phase of the moon etc. Even simply lowering the limit won't help because the same folks will try to cut it too close and sometimes come up short. R,

[pjb45]

Chris,

You are absolutely correct.

I do multiple three shot strings. I record the reading of each shot and the string average/pf.

Knowing thtat there is a variance between shots in a long string, my goal was to observe if I ever got a group of low velocities in the same short string. By doing multiple small strings I gain confidence that for any three rounds my PF was always acceptable.

I have also done a variance between means analysis for the multiple strings-more out of self-inflicted pain-than for statistical validation.

For me, the bottom line is that I have always passed chrono on the first three shots. If by random chance the first three did not pass and had to shot a second three, the odds of hitting major were excellent, if the moons and planets were mis-aligned then the high three of six shots would have made major.

Having shot multiple matches where people did not make minor or major and their resulting change in attitude encouraged me to always hit my goal. I was at WSSSC once when more than one person did even make minor. Bummer way to shoot a match.

So knowing that my using of Pat's chrono reads slightly lower than the chrono at A2 or Nationals, having my control group of ammo, I never worry about not making major. If the control ammo reads 166 on the local chrono and 171 at Nats and my new ammo reads reads 166 on the local chrono and am confident that it will read 170+ at a major match.

[tollarja]

Now another problem seems to be that the standard that is required by the rules is more precise than can be accurately measured which is a huge no-no in math, science and engineering (my major) and I really think this should be changed. All it would take is to multiply the PF by say 1.04 if Chrono's tend to be off by 4%.

I agree. I wonder what it will need to be done to have this added to the USPSA chrono rules.

Not going to happen. If someone made 159PF and they multiplied it by 1.04 they'd just get over 165PF to make Major. That would mean the new PF floor was really 159PF, not 165. Some competitors are going to look for every "advantage" they can find and if they can run 159PF rather than 165, they will. It would do the same thing for Minor, but the new number would be 121PF rather than 125PF.

The answer isn't changing the rules, or using sophisticated processes to predict what your load should be doing, it's simply loading with a margin of error sufficient that it's not a problem. If you load 7-8PF higher than the limit, you'll be fine...easy, quick, simple and it works regardless of weather, bad luck, phase of the moon etc. Even simply lowering the limit won't help because the same folks will try to cut it too close and sometimes come up short. R,

We're just saying that there should be a + or - value for the PF. I am totally fine with people trying to hang at the very limits if they think it is an advantage. If they want to load for a PF of 122 hoping to barely pass with the uncertainty at the chrono that is fine. I don't like the fact that they have a very accurate requirement that they measure with an inaccurate system. If I felt that they actually wanted a pf of 122 and 162 and already accounted for error I would be fine, but I don't think they did. This is just something I'll have to get over, but the engineer/mathematician/scientist in me is weeping.

[G-ManBart]

We're just saying that there should be a + or - value for the PF. I am totally fine with people trying to hang at the very limits if they think it is an advantage. If they want to load for a PF of 122 hoping to barely pass with the uncertainty at the chrono that is fine. I don't like the fact that they have a very accurate requirement that they measure with an inaccurate system. If I felt that they actually wanted a pf of 122 and 162 and already accounted for error I would be fine, but I don't think they did. This is just something I'll have to get over, but the engineer/mathematician/scientist in me is weeping.

If you have a + or -, it's not really a limit any longer is it? It becomes a range. Once it's a range, the lowest of that range becomes the target, so that's really the new limit...etc, etc. The fact that you have six, (or seven) shots to make PF is the + or - you're suggesting we need. If someone's load isn't capable of getting three out of 7 shots to make PF, even with the error in the system, they really were shooting a load that was too light and they've got nobody to blame but themselves. Load to 173PF+ and you'll never have to worry about it! R,

[Chris iliff]

171.5PF consistently at home, before and after Area 6. While at Area 6 I went 164.1, minor. I just want to add that I feel that being 7PF higher is just not enough of a cushion in my opinion. I am now convinced, after all my planning, spending, and driving that I will never go to another match and not have a cushion of 10PF. The variables are too many to list between different chronographs, better safe than sorry, as I was with my 6.5PF cushion.