Originally Posted by msolson
he brings a bushnell verified speed gun. feel free to debate the margin of error on that gun, but if you think its error is on the order of 10% or more, then please provide links about how a bushnell verified speed gun is that far off.
Well, I just ran a quick search on radar gun accuracy in scholarly journals. Nothing came up. Lots of radar guns used in research on sports, conditioning, and biomechanics, but reliability is always assumed, couldn't even find references to another article. Then I Googled Bushnell Radar Gun Accuracy, and came up with dozens of typically bullsh*t discussions, chiefly related to baseball. One forum was solid that the Bushnell consistently underestimated the real speed by 7-8 mph, using a $600 unit as a "gold standard." Another forum was equally assured that the Bushnell overestimated by 3-5 mph. All forums agreed that higher priced units like the Stalker were more accurate, and plastic units like the Mattel the least, although I could not find a single instance of anyone knowing this through more than anecdotal information. The only "research" I can find was offered by Bushnell in-house sources, so not precisely unbiased. Not surprisingly, they claim plus or minus 1 mph. But in all honesty, I'd believe an experienced baseball scout over you guys.
Apparently the problems with all guns include the "Cosine Effect," where the placement of the gun relative to the axis of motion will affect the estimate, as well as where in the objects' path the actual measurement is made (these are not average speeds but "snapshots."), and whether or not the gun itself is in motion, however slight. Some guns apparently measure balls as they are midway to the plate, the Bushnell was said to measure at the plate, and so on. Moreover, no gun can be positioned so it's exactly on the axis of the ball, even if the catcher wants to hold it, or there's no catcher, because the ball describes a curvilinear trajectory. Much disagreement about what a 90 mph fastball even means, in fact, as in when it leaves the pitcher's hand, when it crosses the plate, other? And finally, all guns need to be "tuned" with a special device everytime they're used, I discover. Apparently they have a lot of instrumental "wander" and can lose precision within a few usages.
I did not find any particular awe for law enforcement guns, and apparently courts routinely throw their findings out if the defendant bothers to contest it and get an attorney. So more of the "that's how it's done" than any reason why it's consider a standard. Lot of funny stories about sprinting palm trees and so on. A recent court case, in fact, was brought to argue that GPS is more accurate than police radar. Couldn't find outcome, although what courts know about science could be put in your back pocket with room to spare.
The takeway, for me, is that a) none of these devices approach anything like scientific reliability except under experimental lab conditions. And b) pointing a garden variety gun at a moving skier will be affected by the pitch and angles of the slope, when the reading is actually taken, whether the skier is accelerating or decelerating, how steady the hand of the user is, and where the gun is relative to the axis of skier motion when the actual measurement is taken. The further off the axis of motion, or the more the various motions change trajectories in space, the less accurate the reading. I don't find a lot here to suggest that this is a "gold standard" that can calibrate GPS.
So sadly, I still don't believe all these speeds I read about on Epic. Or more precisely, I don't know which ones to believe, since by chance some will be more accurate than others, and a couple will be very accurate. But the differences between them look fairly random, so impossible to predict which ones to believe. Except for my own speed estimates, which are dead on. Always.