"closed bolt"

Easy.

First, proper statistics. WARPIG found a 0.5 inch variation in spread. With a 20-shot sample, this may or may not be statistically significant. We won't know until we do the math. And to do the math meaningfully, we will need a lot more than 20 shots. It is almost impossible to show statistical significance with 20 shots.

Second, while the difference between 11 inches and 11.5 inches may be "not much," perhaps if we used an Angel we would get shot spreads of 2 inches and 2.5 inches. Still 0.5 inches difference, but now suddenly it looks important. This is why we must use different equipment. Stingrays aren't exactly representative of all other paintball guns out there.

Third, why just measure spread? A single outlier skews the entire variability calculation. Which is more accurate - a gun that fires an 11-inch scattershot spread, or a gun that fires 19 balls on top of each other with a single 11-inch outlier? I say we have to measure every shot. Unless, of course, we are just testing whether closed bolt operation leads to fewer outliers.

Fourth, barrels - if the theory is that the forward-moving bolt affects the ball, then it could be important whether there is a tight barrel fit for the paint, because a tight fit might counteract whatever the bolt does. Maybe open/closed doesn't matter with a good fit, but does matter with a small paint/large barrel. We won't know until we check.

Same for wind - maybe closed bolt operation is more or less affected by wind? This argument has been brought against Flatlines for years. Same for distance - maybe closed bolt is only more accurate at long ranges? This discussion has been had about barrel length for years.

Same goes for other confounding variables. Proper science consists of eliminating confounds, not just saying "eh - that probably doesn't matter."

NOTHING can be concluded from the WARPIG study. Had they fired an extra few hundred shots, measured every shot, and done some actual stats, then we could have made a conclusion. But that conclusion would have been: Using this equipment under these conditions, we found no difference between open and closed bolt operation. Additional studies will have to be done to eliminating confounding variables and establish external validity.

We can either be scientific about it or not. Or, of course, we can just be unscientific and pretend that we are being scientific. That is also an option.

I repeat - the WARPIG study was a noble but amateurish attempt, and ultimately proves nothing.

(Although as noted by a previous poster, "prove" is a tricky scientific concept)

Author	Message Topic Search Topic Options Post Reply Create New Topic Printable Version Translate Topic
ImpyKing2 Members Profile Send Private Message Find Members Posts Add to Buddy List Member Guested - Cannot follow any rules Joined: 14 July 2004 Location: San Marino Status: Offline Points: 500	Post Options Post Reply Quote ImpyKing2 Report Post Thanks(0) Quote Reply Posted: 20 August 2004 at 6:49pm
	Thank you Enos.
	(I'm a nob with no clue about almost anything...)

Clark Kent Members Profile Send Private Message Find Members Posts Visit Members Homepage Add to Buddy List Platinum Member Joined: 02 July 2004 Location: United States Status: Offline Points: 8716	Post Options Post Reply Quote Clark Kent Report Post Thanks(0) Quote Reply Posted: 21 August 2004 at 7:48am
	Easy. First, proper statistics. WARPIG found a 0.5 inch variation in spread. With a 20-shot sample, this may or may not be statistically significant. We won't know until we do the math. And to do the math meaningfully, we will need a lot more than 20 shots. It is almost impossible to show statistical significance with 20 shots. Second, while the difference between 11 inches and 11.5 inches may be "not much," perhaps if we used an Angel we would get shot spreads of 2 inches and 2.5 inches. Still 0.5 inches difference, but now suddenly it looks important. This is why we must use different equipment. Stingrays aren't exactly representative of all other paintball guns out there. Third, why just measure spread? A single outlier skews the entire variability calculation. Which is more accurate - a gun that fires an 11-inch scattershot spread, or a gun that fires 19 balls on top of each other with a single 11-inch outlier? I say we have to measure every shot. Unless, of course, we are just testing whether closed bolt operation leads to fewer outliers. Fourth, barrels - if the theory is that the forward-moving bolt affects the ball, then it could be important whether there is a tight barrel fit for the paint, because a tight fit might counteract whatever the bolt does. Maybe open/closed doesn't matter with a good fit, but does matter with a small paint/large barrel. We won't know until we check. Same for wind - maybe closed bolt operation is more or less affected by wind? This argument has been brought against Flatlines for years. Same for distance - maybe closed bolt is only more accurate at long ranges? This discussion has been had about barrel length for years. Same goes for other confounding variables. Proper science consists of eliminating confounds, not just saying "eh - that probably doesn't matter." NOTHING can be concluded from the WARPIG study. Had they fired an extra few hundred shots, measured every shot, and done some actual stats, then we could have made a conclusion. But that conclusion would have been: Using this equipment under these conditions, we found no difference between open and closed bolt operation. Additional studies will have to be done to eliminating confounding variables and establish external validity. We can either be scientific about it or not. Or, of course, we can just be unscientific and pretend that we are being scientific. That is also an option. I repeat - the WARPIG study was a noble but amateurish attempt, and ultimately proves nothing. (Although as noted by a previous poster, "prove" is a tricky scientific concept)