Hi Dave, well done, I really like the fact that you do actual statistical testing on the results and thus try to remove bias. Since you, or someone on your team, apparently has sound knowledge of that, I hope you will take into account the following feedback for an update of your results.davedej wrote: ↑Mon May 21, 2018 9:49 pmMore of the same here:
http://www.wetestkites.com/2018/05/21/w ... game-over/
(but maybe some more details involved)
We Test Kites
There are a number of possible biases: number of riders, number of sessions, location on the board of the WOOs, that you'd need to account for in order to make a statistically sound statement about Woo's accuracy between versions. And then there's different WOOs...
First, I'd like to know how many riders participated. It seems 1? In a follow-up, a larger number of riders would account for possible rider-related bias. You may set up a board and lend it to as many riders as possible to eliminate this factor. If you'd organize something like that on a day at Wijk or even Schellinkhout, you will get a huge session with a lot of data points. Best would be to do this at both places so you will have both a flat water spot and a wave spot. Second, of course, the more sessions the better. One session results can be skewed for whatever reason so more sessions will eliminate bias there.Third, the location on the board of the WOOs. It seems that this can impact the results so you would need to alternate the WOOs between locations in order to eliminate that. Fourth, and you rightly point that out already in your article, more WOOs are needed of all versions in order to eliminate their influence. Let alone the different software versions.
I know, this is much harder to achieve than a simple one-day test with some statistical analysis. But your statement "WOO: game over" is quite serious and with the current data you cannot claim that with any trustworthy backup. Some hotheads above are talking about class action suits against the company, which I find stupid, but in that light statements like this might entice others to stake claims against you.