I'm fairly certain the previous poll didn't get up to n=15, but I could be wrong. The reason I think this, is I remember fretting that "11 people isn't very many to answer the poll."
If your light fails the test, it is of course a lot more probable that you tell about it in discussion forum.
Very true. This is where problems will pop up for sure. For example, I've noticed several long-running discussions of problems with the SF Saint. It's generally regarded as a very high quality light, but you might not get that impression if you read all the failure reports here on CPF. I'm suspicious it's as you say, a quality light with vocal critics.
Now, how many of those people did buy this special reflectorles full flood headlamp? My guess is, only a little fraction of those people.
Perhaps; on the other hand, with the main H502 thread logging 140
thousand views, we can say there was a lot of interest in the H502, and that might translate into high sales. Only ZL knows.
Another thing that I would like to ask is, how common it is that once a new light is released (regardless of brand), users report water leaking into those new and newly released lights? A single report? Or a couple reports? Is that normal? Common?
The frequency is totally and completely biased, of course, by the general interest in the light and how many people purchase it. Nobody's reporting the Dosun headlamps leak--hardly anybody owns them. Whereas ZL is far and away the most popular, most owned, and most discussed light on this forum, so inevitably there would be more discussions of failures of ZL,
even if the base rates of all lights were exactly the same.
That's why the denominator (that is, the question, "out of how many?") is so important. You can't draw conclusions by examining the numerator alone. The denominator corrects for frequency of ownership (popularity).
It's remarkable, but you'll have people look at this poll like this, that has a numerator and a denominator, and say, "you can't determine anything by that" (with some justification, since it's low-n) and then they'll jump to a conclusion based on examining the numerator all by itself (in other words, just the number of times they see a problem mentioned, without factoring ownership size). I believe that's called the "baserate fallacy," that people are prone not to consider base rates before coming to conclusions about incidents.
