Find fields & stores near you!
Find fields and stores
Zipcode
PbNation News
PbNation News
Community Focus
Community Focus

Reply
 
Thread Tools
Old 07-17-2008, 09:28 AM #85
brycelarson
 
 
brycelarson's Avatar
 
Join Date: Apr 2007
Location: Minnesota
so, I've now requested three times on this thread for a list of shot velocities. 20 or more shots over the chono - write them down, when you get home, wipe off your paint-y little fingers and come here and type them in. CP and I will even do the math for you.

Show me your numbers!

I think we've had one response so far that's accurate - which was paintballer9876 claiming +/- 8. I belive that. the rest of the +/- 2 or 3 guys - no way, you're simply not getting those numbers.

If you really think you are - test it. 20 or more shots, post all of them.
__________________
-Bryce Larson
b_teeth@yahoo.com
brycelarson is offline   Reply With Quote
Old Sponsored Links Remove Advertisement
Advertisement
Old 07-17-2008, 01:57 PM #86
Khanhfucious
w0rd
 
Khanhfucious's Avatar
 
Join Date: Apr 2006
Location: Boston
Khanhfucious owns a Planet Eclipse Ego
Khanhfucious owns a Planet Eclipse Ego
Khanhfucious supports Team VICIOUS
Quote:
Originally Posted by Agent6 View Post
the most consistent gun i've seen was an invert mini, straight out of the box, shooting +/-2
Really..all 4 of my Minis came straight out of the box spiking between 200 and 300 FPS averaging around 260..thats +/-40FPS!
__________________
Khanhfucious is offline   Reply With Quote
Old 07-17-2008, 04:39 PM #87
Stupified
WotWor
 
Stupified's Avatar
 
Join Date: Mar 2005
Location: Canada
Quote:
Originally Posted by brycelarson View Post
so, I've now requested three times on this thread for a list of shot velocities. 20 or more shots over the chono - write them down, when you get home, wipe off your paint-y little fingers and come here and type them in. CP and I will even do the math for you.

Show me your numbers!

I think we've had one response so far that's accurate - which was paintballer9876 claiming +/- 8. I belive that. the rest of the +/- 2 or 3 guys - no way, you're simply not getting those numbers.

If you really think you are - test it. 20 or more shots, post all of them.

I'm not sure I understand how this works.

Bear with me now:

Let's say I shoot my gun 150 times over the chrono. Of those 150 shots, 146 are within +/-5 fps of each other. However four of them go above this +/-5 by let's say 1-3 fps.

For all intents and purposes for the sport of paintball, shouldn't this still be considered +/- 5? Aren't you guys just being a tad bit anal about a technicality?

Please correct me if I'm way of mark here.
__________________
BrUp

DBD

"Originally posted by 0102030405: Most people were younger in the past, Im pretty sure."
Stupified is offline   Reply With Quote
Old 07-17-2008, 07:08 PM #88
cockerpunk (Banned)
Show Me the Data
 
cockerpunk's Avatar
 
Join Date: Nov 2003
Location: Nowhere near Sparta
 has been a member for 10 years
Quote:
Originally Posted by Stupified View Post
I'm not sure I understand how this works.

Bear with me now:

Let's say I shoot my gun 150 times over the chrono. Of those 150 shots, 146 are within +/-5 fps of each other. However four of them go above this +/-5 by let's say 1-3 fps.

For all intents and purposes for the sport of paintball, shouldn't this still be considered +/- 5? Aren't you guys just being a tad bit anal about a technicality?

Please correct me if I'm way of mark here.
well, range is really a bad why to describe it.

range is simply the difference between the high and low. so when you say 146 shots were +/- 5 i assume you mean, all shots were within 10 fps of each other.

well thats fine and dandy, however, the distribution of those shots within that range are important. this is becuase when we look at probabilistic situations like this, we find that things do not fall under nice neat ranges. they fall under what you might know as a bell curve.

what im saying is this - when you shoot your gun and get a +/-5 range you can have this -

270, 280, 271, 280, 280

or

275, 277, 276, 270, 280

and those to models are VERY different.

so to get a true confidence interval (which is not a range) the standard divation, or the normalized distribution around the mean in important.

when constructing the CI for those two stings of shots we will get very different SD, and thus very different CI.
cockerpunk is offline   Reply With Quote
Old 07-17-2008, 07:27 PM #89
brycelarson
 
 
brycelarson's Avatar
 
Join Date: Apr 2007
Location: Minnesota
Quote:
Originally Posted by Stupified View Post
Let's say I shoot my gun 150 times over the chrono. Of those 150 shots, 146 are within +/-5 fps of each other. However four of them go above this +/-5 by let's say 1-3 fps.

For all intents and purposes for the sport of paintball, shouldn't this still be considered +/- 5? Aren't you guys just being a tad bit anal about a technicality?
Yup, we're being a pain in the butt about a technicality. However, that's the roll that CP and I have decided is needed right now in Paintball. There are tons of claims being made all over the place - and we want to apply real math, real science and real statistics to things and see if the claims are real or marketing.

We're suggesting that the confidence level of most velocity tests is simply too low to mean anything. Yes, in your test with 150 samples you could say with an extremely high level of confidence that your gun shoots a particular consistency.

However, if you read back along this post - most people are claiming +/- XX fps based on a sample of 3 or 4 shots. Using that few samples you simply cannot claim what they're claiming - the confidence interval is far to low.

Basically the confidence interval says how likely you are going to be with a given sample to be within a certain set of values. In your test where 146 out of 150 were within your range - you're great. The same isn't true of a smaller test.
__________________
-Bryce Larson
b_teeth@yahoo.com
brycelarson is offline   Reply With Quote
Old 07-17-2008, 07:28 PM #90
brycelarson
 
 
brycelarson's Avatar
 
Join Date: Apr 2007
Location: Minnesota
started my reply, walked away from the computer and my evil twin has already posted.
__________________
-Bryce Larson
b_teeth@yahoo.com
brycelarson is offline   Reply With Quote
Old 07-17-2008, 07:31 PM #91
cockerpunk (Banned)
Show Me the Data
 
cockerpunk's Avatar
 
Join Date: Nov 2003
Location: Nowhere near Sparta
 has been a member for 10 years
Quote:
Originally Posted by brycelarson View Post
started my reply, walked away from the computer and my evil twin has already posted.
wait ... since when am i the evil one?

i always though you look more evil. i just look like a nerd.
cockerpunk is offline   Reply With Quote
Old 07-17-2008, 07:48 PM #92
brycelarson
 
 
brycelarson's Avatar
 
Join Date: Apr 2007
Location: Minnesota
Quote:
Originally Posted by cockerpunk View Post
wait ... since when am i the evil one?

i always though you look more evil. i just look like a nerd.
fair enough.
__________________
-Bryce Larson
b_teeth@yahoo.com
brycelarson is offline   Reply With Quote
Old 07-17-2008, 07:50 PM #93
cockerpunk (Banned)
Show Me the Data
 
cockerpunk's Avatar
 
Join Date: Nov 2003
Location: Nowhere near Sparta
 has been a member for 10 years
Quote:
Originally Posted by brycelarson View Post
fair enough.
damn ... well i walked right into that one.
cockerpunk is offline   Reply With Quote
Old 07-17-2008, 08:26 PM #94
Stupified
WotWor
 
Stupified's Avatar
 
Join Date: Mar 2005
Location: Canada
Quote:
Originally Posted by brycelarson View Post
Yup, we're being a pain in the butt about a technicality. However, that's the roll that CP and I have decided is needed right now in Paintball. There are tons of claims being made all over the place - and we want to apply real math, real science and real statistics to things and see if the claims are real or marketing.

We're suggesting that the confidence level of most velocity tests is simply too low to mean anything. Yes, in your test with 150 samples you could say with an extremely high level of confidence that your gun shoots a particular consistency.

However, if you read back along this post - most people are claiming +/- XX fps based on a sample of 3 or 4 shots. Using that few samples you simply cannot claim what they're claiming - the confidence interval is far to low.

Basically the confidence interval says how likely you are going to be with a given sample to be within a certain set of values. In your test where 146 out of 150 were within your range - you're great. The same isn't true of a smaller test.
Ahh, I see. Thanks for taking the time to clear it up.
__________________
BrUp

DBD

"Originally posted by 0102030405: Most people were younger in the past, Im pretty sure."
Stupified is offline   Reply With Quote
Old 09-02-2008, 10:08 PM #95
yuik
Mr Diet Coke on the Rocks
 
yuik's Avatar
 
Join Date: Dec 2004
Annual Supporting Member
yuik is playing at Living Legends V
yuik is playing at Living Legends VI
yuik is an NCPA player
yuik plays in the APPA D4 division
yuik is one of the top 250 posters on PbNation
yuik is Legendary
just checking up on this thread to see if anyone put my shens claim to rest.
Quote:
Originally Posted by brycelarson View Post
This is the reality that CP and I have come to. His emag got ridiculously good consistency - but only with extreme underboring - too much to be useful in a game situation. Especially with overboring as people tend to do - you're really never going to get to the mythical +/- 2-4 fps that people commonly claim.
i was trying to do this for a while but i never pulled it off. i also tried overboaring like a mother****er just to see how ****ty the paint shot. (i had some jt red paint that you buy in a long 3000 round case) and it rolled through a .679.
how how extreme was it like would you get barrel breaks on every other shot? i was thinking maybe alittle water filled balls (like ones left outside in a humid place) might help because they would mold to fit the barrel, mind you as long as your shooting at like 220 so they dont break in the barrel. (i think a closed bolt gun would work better for this).
__________________
if you have an ebay account and want to make 3$ in paypal for 5 min of your time pm me!
Closed Bolt Ballers!

Huge gearbag sale. Everything you can imagine including bob saget!

RED JMJ Smooth Mill viking RobTown Valve Ultralight Frame!
yuik is offline   Reply With Quote
Old 09-02-2008, 10:14 PM #96
ipodpeter35
Jizzed in My Pants
 
ipodpeter35's Avatar
 
Join Date: Sep 2005
Location: Washington (the state)
ipodpeter35 helped look for balloons
Quote:
Originally Posted by Adrenaline-Junkie View Post
I dont need to prove or verify my claims. Just because you havnt ever seen it doesnt mean its not true.
This reminds me of something I heard in The Santa Clause lol.
ipodpeter35 is offline   Reply With Quote
Old 09-02-2008, 11:05 PM #97
Ri0T
Mad Scientist At Work...
 
Ri0T's Avatar
 
Join Date: Jun 2001
Location: Los Angeles, CA
Ri0T is a founding member
 has been a member for 10 years
Bryce/CP I think that while your math and theory are sound, the way you're interpreting the question is off. Don't get me wrong, I wish I had the time to do the kind of testing you guys are doing. I've wanted to for a while, due to the horse **** marketing in PB feeds the masses. I think the question at hand is aimed at more simple statistics, however.

The +/- simply means that the deviation from the average shot doesn't exceed that value. So, if I claim +/- 5, and shoot 100 shots, and all of them fall between 270 and 280, that is an accurate figure. I've NEVER seen a gun chrono at +/- 30 that wasn't running CO2 with crap paint. Again, as soon as I can get my Banshee to a chrono, I'll record the readings for 50 shots. Who knows what result I'll get, but I'm willing to do it.
Ri0T is offline   Reply With Quote
Old 09-03-2008, 12:37 AM #98
brycelarson
 
 
brycelarson's Avatar
 
Join Date: Apr 2007
Location: Minnesota
post up a 20 shot string and I'll do the math for you.
__________________
-Bryce Larson
b_teeth@yahoo.com
brycelarson is offline   Reply With Quote
Old 09-03-2008, 06:18 PM #99
SpankyCocker14
Keep On Trollin' On
 
SpankyCocker14's Avatar
 
Join Date: Dec 2001
Location: Balls Deep In Panty Land
SpankyCocker14 is a founding member
 has been a member for 10 years
Someone needs to dig up the old AKA videos of the Viking and Excal shooting solid +/- 2 and 3 over the chrono to shut the OP up...
__________________
Just Trollin' For The LuLz.
SpankyCocker14 is offline   Reply With Quote
Old 09-03-2008, 08:24 PM #100
GoNexus1991
 
 
GoNexus1991's Avatar
 
Join Date: Aug 2006
Location: Ottawa, Canada
Quote:
Originally Posted by cockerpunk View Post
well, range is really a bad why to describe it.

range is simply the difference between the high and low. so when you say 146 shots were +/- 5 i assume you mean, all shots were within 10 fps of each other.

well thats fine and dandy, however, the distribution of those shots within that range are important. this is becuase when we look at probabilistic situations like this, we find that things do not fall under nice neat ranges. they fall under what you might know as a bell curve.

what im saying is this - when you shoot your gun and get a +/-5 range you can have this -

270, 280, 271, 280, 280

or

275, 277, 276, 270, 280

and those to models are VERY different.

so to get a true confidence interval (which is not a range) the standard divation, or the normalized distribution around the mean in important.

when constructing the CI for those two stings of shots we will get very different SD, and thus very different CI.
I never fully understood CI properly
I am going to ask one question though.When we deal with statistics, a lot of the stuff that we do are very arbitrary. For example, evaluating a normal distribution because 68% or more fall under the first SD, 95% in the second Etc is cool and all, but aren't we merely stating this so we can compare some stats we might get here to some others? If we were to all take 3 shots over the chrono and "test" the consistency like that, as long as we all did that, would it not be somewhat representative of what we are trying to achieve?
I guess what you guys are getting at is that it's hard to get a "true" +/- of 5 fps, in a technical way. However, what I don't see is if the OP had in his head that the person were to take a few measured shots to show that their gun shot +/- x, why are we going through all this noz?
Unless the entire idea was to piss everyone off with all this statistical talk which, as interesting as it can be, can be ****ing confusing at times too.
Someone who is knowledgeble, can you please explain the entire referencing CI of 99 or 95 upon the results? I was always told that a CI was that, over time, x % of the shots should fall between this range, (given a normal distribution ?)
although I might have it all wrong
__________________
Ashton Expendable Heroes Eastern MXL
Hopperz Paintball
Underground Productz
Draxxus Paintballs
Exalt Paintball
GoNexus1991 is offline   Reply With Quote
Old 09-03-2008, 09:54 PM #101
cockerpunk (Banned)
Show Me the Data
 
cockerpunk's Avatar
 
Join Date: Nov 2003
Location: Nowhere near Sparta
 has been a member for 10 years
Quote:
Originally Posted by GoNexus1991 View Post
I never fully understood CI properly
I am going to ask one question though.When we deal with statistics, a lot of the stuff that we do are very arbitrary. For example, evaluating a normal distribution because 68% or more fall under the first SD, 95% in the second Etc is cool and all, but aren't we merely stating this so we can compare some stats we might get here to some others? If we were to all take 3 shots over the chrono and "test" the consistency like that, as long as we all did that, would it not be somewhat representative of what we are trying to achieve?
I guess what you guys are getting at is that it's hard to get a "true" +/- of 5 fps, in a technical way. However, what I don't see is if the OP had in his head that the person were to take a few measured shots to show that their gun shot +/- x, why are we going through all this noz?
Unless the entire idea was to piss everyone off with all this statistical talk which, as interesting as it can be, can be ****ing confusing at times too.
Someone who is knowledgeble, can you please explain the entire referencing CI of 99 or 95 upon the results? I was always told that a CI was that, over time, x % of the shots should fall between this range, (given a normal distribution ?)
although I might have it all wrong
the issue is finding the "true mean" vs the "sample mean"

if i want to find the average age of all poeple in the entire world, what do i do?

i decide on what would be a representative sample of the world, poll those poeple, and calculate the mean right?

when we do that we are finding a sample mean.

the question is - how close to the true mean (if you acutally polled everyone on the entire planet) age?

that's where the CI comes in.

so yes, we can do a comparative study by agreeing on a common sample size and say "well that guns three shots are more consistent than that guns three shots"

however, what we really want to know is how consistent EVERY shot from the gun is.

to do this, we are comparing a sample mean (the shots fired over the chrono) to the true mean (the velocity of ALL shots the gun ever fires).

a CI simple compares the variation in the data (standard deviation), with the number of samples taken, and then shows you how accurate to the true mean our sample mean is.
cockerpunk is offline   Reply With Quote
Old 09-03-2008, 10:29 PM #102
Waxball88
Ego♥
 
Waxball88's Avatar
 
Join Date: Sep 2005
My friends dm7 shot true +/- 1 over the chrono (big red) at CFP i was astounded i remember the exact reading 297 297 297 297 with occasional +/- 2 jump
__________________
NINJA Geo+
Ego Mafia
Waxball88 is offline   Reply With Quote
Old 09-04-2008, 08:50 AM #103
brycelarson
 
 
brycelarson's Avatar
 
Join Date: Apr 2007
Location: Minnesota
Quote:
Originally Posted by Waxball88 View Post
My friends dm7 shot true +/- 1 over the chrono (big red) at CFP i was astounded i remember the exact reading 297 297 297 297 with occasional +/- 2 jump
have him shoot 20 shots, record them and post them up. sounds impressive - but giving 4 numbers and then claiming +/- 1 fps is bogus - you need a bigger sample size to claim something like that.
__________________
-Bryce Larson
b_teeth@yahoo.com
brycelarson is offline   Reply With Quote
Old 09-04-2008, 01:51 PM #104
GoNexus1991
 
 
GoNexus1991's Avatar
 
Join Date: Aug 2006
Location: Ottawa, Canada
Quote:
Originally Posted by cockerpunk View Post
the issue is finding the "true mean" vs the "sample mean"

if i want to find the average age of all poeple in the entire world, what do i do?

i decide on what would be a representative sample of the world, poll those poeple, and calculate the mean right?

when we do that we are finding a sample mean.

the question is - how close to the true mean (if you acutally polled everyone on the entire planet) age?

that's where the CI comes in.

so yes, we can do a comparative study by agreeing on a common sample size and say "well that guns three shots are more consistent than that guns three shots"

however, what we really want to know is how consistent EVERY shot from the gun is.

to do this, we are comparing a sample mean (the shots fired over the chrono) to the true mean (the velocity of ALL shots the gun ever fires).

a CI simple compares the variation in the data (standard deviation), with the number of samples taken, and then shows you how accurate to the true mean our sample mean is.
alright, I do see what you're saying now
__________________
Ashton Expendable Heroes Eastern MXL
Hopperz Paintball
Underground Productz
Draxxus Paintballs
Exalt Paintball
GoNexus1991 is offline   Reply With Quote
Old 09-04-2008, 02:00 PM #105
a_sock
 
 
Join Date: Aug 2007
Location: The Great White North
lmao hes shooting an etek and wonders why people claim amazing chrono consistency...btw etek is a mid end and in know way up to other marker standards, if you get a macdev product or a dye product you shouldnt see a shot varrying over 5+/-.
anyone notice that certain markers like to chrono at a certain spot? I find my droid loves 287, It will do a string of 10 and all be on 287, Ill shoot 30 and get little variance, but if I put it up to say 295 it will be less consistent. noticed same thing with some other markers, I call it the consistency grove, I should patent it lol.
__________________
_____________________________________________
a_sock is offline   Reply With Quote
Reply


Thread Tools

Posting Rules
Forum Jump