TangoWhiskey
Touchdown! Greaser!
Another way to do the trial is to have your container of water hidden, but not buried. At 6Y9 I might hide the water in an aircraft or tent, just out of sight. You could have a series of 20 overturned buckets, with only three covering containers of water, for example. Just so long as it wasn't visible.
Thanks for the suggestions. What EdFred said is accurate; the rods deflect when you pass OVER water. They don't "point to" where the water is (like "walk southeast, it's OVER THERE!").
I do like the suggestion of covered (hidden) water. For initial baselines, and to see if it's worth burying samples, I'll try this first. Paper sacks (like lawn/leaf recycling bags) or plastic buckets over quantities of water that "seem" to deflect when not covered. If I find that works with any sort of consistency, we'll try with different people dowsing. Another idea would be to lay out garden hoses, capped on both ends, where SOME of them have water in them and others do not. Can't see the water, but should only indicate over the ones that are filled.
Both of these methods have the disadvantage that any "false hits" might be due to true pipes under the ground that you're not aware of... how do you 'cancel out' that negative indication without calling it 'an excuse'? You'd have to do the test in an area where you KNOW there are no pipes or underground water. How do you find such an area?