[drg]

I'm not complaining about your methodology, just pointing out it doesn't prove what you say it does. The calculated drop at 10 yards with a 30 fps fluctuation is very small, less than the width of a paintball, i.e. your test actually cannot show the difference. The differences magnify further downrange and do end up being significant. You will notice that cockerpunk also noted the same issue in the thread.

EDIT: Rntlee's test also occurred at a distance of fairly minimal calculated drop (about 1.5" @ 30 fps change) and the majority of his values were within about 12 fps so ... the drop is not expected to be detectable there either.

This post has been edited by drg: 06 August 2012 - 02:14 PM

[Troy]

drg, on 06 August 2012 - 01:44 PM, said:

I'm not complaining about your methodology, just pointing out it doesn't prove what you say it does. The calculated drop at 10 yards with a 30 fps fluctuation is very small, less than the width of a paintball, i.e. your test actually cannot show the difference. The differences magnify further downrange and do end up being significant. You will notice that cockerpunk also noted the same issue in the thread.

EDIT: Rntlee's test also occurred at a distance of fairly minimal calculated drop (about 1.5" @ 30 fps change) and the majority of his values were within about 12 fps so ... the drop is not expected to be detectable there either.

By George, he's starting to get it.... Keep in mind that when you increase the range, the background "noise" is increased as well. You can note the rate of expansion by viewing my results at different distances. I didn't put a whole lot of effort into analyzing the different distances (I'll go over it again when I get home), but I suspect that the noise increases exponentially (which makes sense), whereas the possible changes due to velocity fluctuations expand at a much lower rate. Any effects on the Y axis due to speed fluctuations are totally lost in the noise that is produced by vortex shedding and thus all significance is lost. Further testing isn't going to find any changes in velocity being correlated to velocity changes, simply because, the background noise will be so high. So, again, no significant correlation. This doesn't mean that Newtonian physics has flown out the window, but if you are using simple models to predict the ball's endpoint, you are neglecting some of the most major forces that effect the ball's accuracy. If your model didn't even account for some, fairly, major velocity fluctuations, it wouldn't narrow the degree of error in a detectable manner.

Furthermore, to reiterate the point that I made initially, if you can't detect the differences in shot velocities at ranges that people ACTUALLY shoot at, then what the hell is the point of arguing that a marker that shoots +/-5fps over the crono is worse then shooting +/-1fps over the crono from an accuracy standpoint?

I would much rather talk about important things that effect accuracy, and not get caught up in the undetectable minutia.

This post has been edited by Troy: 06 August 2012 - 03:07 PM

[drg]

Troy, on 06 August 2012 - 03:04 PM, said:

Furthermore, to reiterate the point that I made initially, if you can't detect the differences in shot velocities at ranges that people ACTUALLY shoot at, then what the hell is the point of arguing that a marker that shoots +/-5fps over the crono is worse then shooting +/-1fps over the crono from an accuracy standpoint?

I would much rather talk about important things that effect accuracy, and not get caught up in the undetectable minutia.

I don't find a 30 fps velocity difference undetectable by a long stretch, nor would any honest paintball player. You seem to have your nose a little too close to the data and, I suspect, are searching for a little too much ... specialness ... in your analyses. This is not an uncommon event in this forum, unfortunately.

I was hoping you'd bring up the distance "people actually shoot at" ... I regularly shoot to the maximum range of my marker and many if not most players do the same. In your defense of the 10-yard test as relevant you mentioned that you noticed a lot of your kills come at that range, so it was not necessarily relevant to study longer ranges. I think that's a classic case of the hypothesis adversely affecting the study, a kind of self-fulfillment. I venture that it's the other way around, the consistent trajectory at that range contributes to the increase in kills. In a way the math tells us it is a foregone conclusion that this would be so.

[Troy]

drg, on 06 August 2012 - 06:01 PM, said:

I don't find a 30 fps velocity difference undetectable by a long stretch, nor would any honest paintball player. You seem to have your nose a little too close to the data and, I suspect, are searching for a little too much ... specialness ... in your analyses. This is not an uncommon event in this forum, unfortunately.

I was hoping you'd bring up the distance "people actually shoot at" ... I regularly shoot to the maximum range of my marker and many if not most players do the same. In your defense of the 10-yard test as relevant you mentioned that you noticed a lot of your kills come at that range, so it was not necessarily relevant to study longer ranges. I think that's a classic case of the hypothesis adversely affecting the study, a kind of self-fulfillment. I venture that it's the other way around, the consistent trajectory at that range contributes to the increase in kills. In a way the math tells us it is a foregone conclusion that this would be so.

Here I thought you were actually getting somewhere... *sigh*

In my very first post I said "Exaggerations aside..." meaning that I thought that Bryce was exaggerating a bit with the 30fps speed range, using hyperbole to make his point. You can continue to beat on your strawman, if that makes you feel better about yourself.

I guess you totally missed what I was saying about vortex shedding increasing the degree of error at a faster rate than speed fluctuations... Whether or not you make kills at 5 feet or 500 feet, the significance of initial speed fluctuations effects on the ball's flight decreases exponentially in relationship to how much vortex shedding effects the flight path. I'm glad you had that pre-programmed response for distance, but I don't give a shit what range you shoot at. The data is there, you can either look at it, or ignore it... your choice.

[drg]

Troy, on 06 August 2012 - 07:14 PM, said:

In my very first post I said "Exaggerations aside..." meaning that I thought that Bryce was exaggerating a bit with the 30fps speed range, using hyperbole to make his point. You can continue to beat on your strawman, if that makes you feel better about yourself.

Actually those fps figures are accurate to their testing, as far as I know. He is, unfortunately, like you, overstating the meaning of the data, though you are doing it to a much greater degree.

Troy, on 06 August 2012 - 07:14 PM, said:

I guess you totally missed what I was saying about vortex shedding increasing the degree of error at a faster rate than speed fluctuations... Whether or not you make kills at 5 feet or 500 feet, the significance of initial speed fluctuations effects on the ball's flight decreases exponentially in relationship to how much vortex shedding effects the flight path. I'm glad you had that pre-programmed response for distance, but I don't give a shit what range you shoot at. The data is there, you can either look at it, or ignore it... your choice.

Unfortunately no, there has yet to be data presented on these forums of the ranges where differences in initial velocity make the most difference, which is well within playing ranges (though on the long side).

This post has been edited by drg: 06 August 2012 - 07:42 PM

[Troy]

drg, on 06 August 2012 - 07:39 PM, said:

Actually those fps figures are accurate to their testing, as far as I know. He is, unfortunately, like you, overstating the meaning of the data, though you are doing it to a much greater degree.

Clearly, you know more about what I meant when I made that reply then I did.

drg, on 06 August 2012 - 07:39 PM, said:

Unfortunately no, there has yet to be data presented on these forums of the ranges where differences in initial velocity make the most difference, which is well within playing ranges (though on the long side).

It's high time for you to put up or shut up. I've submitted a perfectly reasonable hypothesis, prove me wrong.

[cockerpunk]

we have 100, 125, and 150 foot data available ... from out testing. clearly shows the same lack of a trend.

[Bauzer]

jimmies are being rustled in this thread

[drg]

cockerpunk, on 07 August 2012 - 10:31 AM, said:

we have 100, 125, and 150 foot data available ... from out testing. clearly shows the same lack of a trend.

These tests are not testing what we are discussing here and slicing up the data therein to comment on something they didn't test is IMO misleading.

Part of the issue is the way "+/- 15 fps" has been expressed. It is actually a very vague statement. +/- 15 says nothing about the velocity SD, so it could mean anything from a few outliers to fairly heavy inconsistency. The practical result, as we all agree, is that a more consistent gun can chrono higher than another.

None of these tests are comparing the results of one gun chronoed at a different speed than another because of a difference in velocity SD. I'm not 100% sure how you would test this.

This post has been edited by drg: 07 August 2012 - 03:50 PM

[cockerpunk]

drg, on 07 August 2012 - 03:47 PM, said:

cockerpunk, on 07 August 2012 - 10:31 AM, said:

we have 100, 125, and 150 foot data available ... from out testing. clearly shows the same lack of a trend.

These tests are not testing what we are discussing here and slicing up the data therein to comment on something they didn't test is IMO misleading.

i call it using data effectively. we took it, its accurate, and it has clear conclusions, just because we didn't explicitly state we were testing it doesn't mean the data is suddenly less conclusive

drg, on 07 August 2012 - 03:47 PM, said:

Part of the issue is the way "+/- 15 fps" has been expressed. It is actually a very vague statement. +/- 15 says nothing about the velocity SD, so it could mean anything from a few outliers to fairly heavy inconsistency. The practical result, as we all agree, is that a more consistent gun can chrono higher than another.

None of these tests are comparing the results of one gun chronoed at a different speed than another because of a difference in velocity SD. I'm not 100% sure how you would test this.

im fine with that, still doesn't make our data less conclusive on the topic.

[drg]

cockerpunk, on 07 August 2012 - 06:15 PM, said:

i call it using data effectively. we took it, its accurate, and it has clear conclusions, just because we didn't explicitly state we were testing it doesn't mean the data is suddenly less conclusive

Actually it is less conclusive due to the fewer data points. Your data is heavily skewed by outliers.

drg, on 07 August 2012 - 03:47 PM, said:

im fine with that, still doesn't make our data less conclusive on the topic.

Your data is mostly inconclusive on the topic, which is not surprising because it doesn't test the topic. The elephant in the room is the source of the fluctuation -- shots which amount to outliers on an otherwise consistent marker could be paint-based fluctuations, i.e. an oversized, undersized or out-of-round ball, situations which your and other peoples' testing has shown can induce spin. Whereas a marker-based fluctuation, say, from an inconsistent pulse, could see less spin-induced trajectory change. So no, your data is not conclusive by a long shot.

FWIW Troy's data actually comes closer and actually does show a correlation, however I'm not sure how valid that is at 30 feet. In the end though, it's mathematically conclusive and patently obvious to any player that velocity has a correlation to trajectory.

This post has been edited by drg: 08 August 2012 - 01:50 AM

[cockerpunk]

drg, on 08 August 2012 - 01:29 AM, said:

cockerpunk, on 07 August 2012 - 06:15 PM, said:

i call it using data effectively. we took it, its accurate, and it has clear conclusions, just because we didn't explicitly state we were testing it doesn't mean the data is suddenly less conclusive

Actually it is less conclusive due to the fewer data points. Your data is heavily skewed by outliers.

drg, on 07 August 2012 - 03:47 PM, said:

im fine with that, still doesn't make our data less conclusive on the topic.

Your data is mostly inconclusive on the topic, which is not surprising because it doesn't test the topic. The elephant in the room is the source of the fluctuation -- shots which amount to outliers on an otherwise consistent marker could be paint-based fluctuations, i.e. an oversized, undersized or out-of-round ball, situations which your and other peoples' testing has shown can induce spin. Whereas a marker-based fluctuation, say, from an inconsistent pulse, could see less spin-induced trajectory change. So no, your data is not conclusive by a long shot.

FWIW Troy's data actually comes closer and actually does show a correlation, however I'm not sure how valid that is at 30 feet. In the end though, it's mathematically conclusive and patently obvious to any player that velocity has a correlation to trajectory.

we have the data on the topic, we don;t need to test it because we have nearly 100 patterns shot to grab data from. all with different averages, all with different SD .... its all there. again, just because we did not take the data with the express purpose of looking at the topic, doesn't make the data less applicable, or less conclusive.

[drg]

cockerpunk, on 08 August 2012 - 09:06 AM, said:

we have the data on the topic, we don;t need to test it because we have nearly 100 patterns shot to grab data from. all with different averages, all with different SD .... its all there. again, just because we did not take the data with the express purpose of looking at the topic, doesn't make the data less applicable, or less conclusive.

I disagree, for the reasons stated above, but I suspect we aren't going to come any closer to agreeing on this one.

I'll still be trying to get the best consistency I can, personally.

This post has been edited by drg: 08 August 2012 - 08:04 PM

[cockerpunk]

drg, on 08 August 2012 - 08:03 PM, said:

I disagree, for the reasons stated above, but I suspect we aren't going to come any closer to agreeing on this one.

and your stated reason makes no sense at all

I'll still be trying to get the best consistency I can, personally.

as will we all ... some of us for real reasons, and others for not real reasons

^^^^

[drg]

cockerpunk, on 08 August 2012 - 09:08 PM, said:

and your stated reason makes no sense at all

It makes perfect sense as long as one is thinking rather than being mindlessly defensive. You can't compare different patterns to each other absolutely without an aimpoint control, something your tests never have. You have simply never tested the claim, yet you are calling your data conclusive on it. Shame on you.

This post has been edited by drg: 10 August 2012 - 02:46 AM

[Troy]

drg, on 10 August 2012 - 02:26 AM, said:

cockerpunk, on 08 August 2012 - 09:08 PM, said:

and your stated reason makes no sense at all

It makes perfect sense as long as one is thinking rather than being mindlessly defensive. You can't compare different patterns to each other absolutely without an aimpoint control, something your tests never have. You have simply never tested the claim, yet you are calling your data conclusive on it. Shame on you.

Their aim point is controlled. It's the statistical center of the data. Where the marker is ACTUALLY pointing is irrelevant (assuming it's static) as long as there is deviation in the velocity, they have the ability to test for correlation between Y positioning and velocity. No significant correlation has been found... ever, at any distance.

Who among us is being mindlessly defensive? The people that are unconvinced of your theory and can point to data that backs it up (even if the data isn't conclusive by your standards), or the person that is 100% sure of his theory, but can't find a shred of observational data to bolster his position? You are more then welcome to do your own test, but this armchair scientist approach that you are taking isn't working for you. Personally, I would LOVE to know how much velocity fluctuation it takes to influence the final shot position... that would be a cool experiment that you could tackle, or you could just sit back and continue to make unsupported claims.

[drg]

Troy, on 10 August 2012 - 08:43 AM, said:

Their aim point is controlled. It's the statistical center of the data. Where the marker is ACTUALLY pointing is irrelevant (assuming it's static) as long as there is deviation in the velocity, they have the ability to test for correlation between Y positioning and velocity. No significant correlation has been found... ever, at any distance.

No, the real aimpoint is significant if you want to compare absolute y between two different patterns. Tests may find similar levels of deviation at each chrono point but that does not measure the trajectory difference between chrono points, in fact the trajectory is factored out deliberately.

Your test data actually comes the closest to testing this, and even at 30 feet your data shows a correlation between chrono velocity and y position.

But of course anyone who has done much shooting with a paintball gun, particularly with a 12g powering it, knows that velocity absolutely has a correlation to trajectory.

This post has been edited by drg: 10 August 2012 - 06:19 PM

[rntlee]

It's not as obvious as you seem to think, drg.
The reason being that in order to hit a level target 100' away with a paintball gun, you need to tilt the gun on an angle somewhere around 3 dgrees. The balls traverse a ballistic arc, so accelleration from gravity begins at the top of the arc (or over halfway to the target). This makes the difference in drop betwwen a 300 fps shot and a 270 fps drop quite small at 100' actually. Using a free ballistic app on my ipad (not great, I know), I calculated the difference to be less than a paintball diameter at 100'. Sure, the calculated difference is greater at 150' (22"), but it's still far smaller than the shot spread and hitting something at that range is shit luck anyway.
If you go out and record some shot locations vs. velocity with your marker, it will very quickly become obvious that paintballs have many large forces affecting their shot spread. You'll notice that the lower velocity shots will not "follow the rules" and are just as often higher than the faster balls as they are lower on the target.

This post has been edited by rntlee: 10 August 2012 - 08:26 PM

[drg]

I'm coming up with a much larger number (between 5 and 10 inches), even assuming completely flat terrain and a target at equal height as the muzzle (big assumptions indeed). As I mentioned earlier, it's not simply a question of firing a bunch of shots and looking at the velocities and impact location of each. There is a difference between measuring the performance of outliers and normal shots (i.e. chrono speed shots). You can't just take the low shots and high shots from a given string of shots, with no marker tuning change, and compare them as if they were normal shots at the target velocity.

Regardless, the presence of the calculated difference indicates that there does exist a force differential between balls traveling at different velocities, and I just can't think of a reason it would not be a good idea to reduce or eliminate that differential. That force HAS to be affecting the flight of the balls, if we accept the physics as valid.

This post has been edited by drg: 10 August 2012 - 09:24 PM

[cockerpunk]

drg, on 10 August 2012 - 02:26 AM, said:

cockerpunk, on 08 August 2012 - 09:08 PM, said:

and your stated reason makes no sense at all

It makes perfect sense as long as one is thinking rather than being mindlessly defensive. You can't compare different patterns to each other absolutely without an aimpoint control, something your tests never have. You have simply never tested the claim, yet you are calling your data conclusive on it. Shame on you.

aim point is irrelevant in this analysis.

This post has been edited by cockerpunk: 10 August 2012 - 11:23 PM