# Finding water with coat hangers **URL:** \ **Category:** Cecil's Columns/Staff Reports\ **Created:** [October 21, 2008, 9:55pm UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954 "2008-10-21T21:55:24Z")\ **Posts on this page:** 20\ **Page:** 9
**Author:** ![Princhester](https://avatars.discourse-cdn.com/v4/letter/p/3e96dc/32.png) [@Princhester](https://boards.straightdope.com/u/Princhester)\ **Post date:** [November 6, 2008, 4:30am UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/161 "2008-11-06T04:30:42Z")
What if someone says they can find water lines, septic tanks, etc. with a coat hanger? ---
**Author:** ![Czarcasm](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/czarcasm/32/4050_2.png) [@Czarcasm](https://boards.straightdope.com/u/Czarcasm)\ **Post date:** [November 6, 2008, 4:32am UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/162 "2008-11-06T04:32:47Z")
> [@Peter\_Morris](#): > > Depends on the actual claim made. How about designing a protocol for the OP? ---
**Author:** ![Princhester](https://avatars.discourse-cdn.com/v4/letter/p/3e96dc/32.png) [@Princhester](https://boards.straightdope.com/u/Princhester)\ **Post date:** [November 6, 2008, 4:33am UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/163 "2008-11-06T04:33:50Z")
Beatcha. ---
**Author:** ![Mangetout](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/mangetout/32/19_2.png) [@Mangetout](https://boards.straightdope.com/u/Mangetout)\ **Post date:** [November 6, 2008, 8:41am UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/164 "2008-11-06T08:41:00Z")
> [@Peter\_Morris](#): > > > [@Mangetout](#): > > > > OK, but is there any way to verify that the baseline test is presented in anything like the threatening manner you describe? > > Just for the moment, can we restrict the discussion to Musicat’s proposed test. I think that the way HE wants to do it is fundamentally dishonest. > > Whether or not anyone else’s test has the same problem is another discussion. > > > [@](#): > > > > It’s not enough to point at the procedure - > > Yes it is. It’s perfectly valid to look at the proposed test and spot what’s wrong with it before it happens. > > I am unaware of Musicat actually conducting any tests in the way he proposes. If he ever actually does so, we can analyse them. OK, I’m happy enough to leave it for now - my concern was not so much over the procedure, but the violent application of it that you were portraying. If you do not wish people like me to pick up on things like that, it might serve you better to insert less emotive hypothetical dialogue in your descriptions of the process. ---
**Author:** ![MarcusF](https://avatars.discourse-cdn.com/v4/letter/m/67e7ee/32.png) [@MarcusF](https://boards.straightdope.com/u/MarcusF)\ **Post date:** [November 6, 2008, 9:46am UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/165 "2008-11-06T09:46:20Z")
> [@Peter\_Morris](#): > > A few guidelines. > > [ol] > > [li]The test MUST be conducted by a qualified scientist, who actually understands proper protocol design, and how to analyse the results.[/li][li]The results of the test must be given a statistical analysis by a qualified person to determine their validity. Success or failure will be determined by statistical significance of the results. It is not sufficient to set an arbitary target. [/li] > [li] The test has to be subjected to peer review to determine its validity. The results will only be considered valid only if it survives the peer review process. [/li][li] Requiring all participants to sign a statement that the test is fair is NOT an acceptable substitute for peer review. Nor shall the test be self published.[/li][li] The test conducted MUST closely resemble the actual claim made. If, for example, the applicant claims the ability to find natural water in the ground, then he will not be tested on finding a jug of water in a box. Discussion of the test design shall be conducted in a calm manner. Anyone who is notoriously short tempered, rude and aggressive, or who habitually uses insulting or abusive language shall not participate in such discussion. [/li][li] Let there be an experienced magician giving advice to the scientist. The job of the magician should be to prevent fraud, that is to spot anyone trying to cheat. It is not the job of the magician to give advice on protocol design.[/li][li] The magician shall act as advisor to the scientist who conducts the test, not vice versa. The scientist shall remain in charge. No tail shall wag the dog.[/li][li] The test must be conducted by someone independent, with no vested interest in the results. It must not make any personal difference to them what the results are. It shall not be conducted by someone who expresses hatered towards dowsers, and seeks to humiliate them at every opportunity. Neither shall it be conducted by someone who profits from sales of stories describing “how I beat the woowoos” [/li][li] Anyone who on a regular basis uses terms such as “woowoo” or other figures of hate speech has demonstrated a lack of professionalism, and must not involved with the test in any capacity whatsoever.[/li] > [/ol] > There, that should do for a start. OK, based on these ground rules, what was wrong with the German test described on **Musicat’s** post #89? (Please don’t say Randi’s involvement - your list allows for advice from a magician.) As far as I can see from the original [write up](http://http://www.phact.org/e/z/kassel.htm) it was set up in an appropriate manner and tested the specific ability the OP claims - "I can locate water lines, septic tanks, etc. " - but unfortunately there was no positive result. Perhaps the OP can do better? ---
**Author:** ![Musicat](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/musicat/32/20189_2.png) [@Musicat](https://boards.straightdope.com/u/Musicat)\ **Post date:** [November 6, 2008, 2:08pm UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/166 "2008-11-06T14:08:18Z")
> [@Peter\_Morris](#): > > Depends on the actual claim made. How about " I can find water in plastic jugs when covered by a cardboard box"? Let’s see your protocol for that, keeping in mind that the golden standard would be results that are objective and do not require subjective judging. ---
**Author:** ![Contrapuntal](https://avatars.discourse-cdn.com/v4/letter/c/e274bd/32.png) [@Contrapuntal](https://boards.straightdope.com/u/Contrapuntal)\ **Post date:** [November 6, 2008, 5:37pm UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/167 "2008-11-06T17:37:45Z")
> [@Peter\_Morris](#): > > If someone says “I can find natural water in the ground” then you test his ability to find natural water in the ground. You don’t offer to test his ability to find a jug of water in a box. What if the claim is “I can jump 3 feet straight up?” Would it be fair to test whether he can jump 2 feet straight up? ---
**Author:** ![BunnyTVS](https://avatars.discourse-cdn.com/v4/letter/b/977dab/32.png) [@BunnyTVS](https://boards.straightdope.com/u/BunnyTVS)\ **Post date:** [November 6, 2008, 9:04pm UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/168 "2008-11-06T21:04:39Z")
Before I go any further I apologise in advance for mentioning Randi. Imagine a dowsing test set up the same as those described by Randi and **Musicat** with a couple of differences. i) instead of testing one person you tested 100 people split into four groups of 25\*. Group 1 professed dowsers. Group 2 those who had never dowsed but believed in it Group 3 those who had no opinion one way or the other about dowsing Group 4 disbelievers in dowsing Each victim, I mean participant, takes part in both the non-blind and blind parts of the trials and their scores recorded. Hell you could even randomize which test the take first. The results are then put through whatever types of statistical analyses are appropriate. You would then have a breakdown by individual, group and combined. \*Or whatever initial pool size is deemed neccessary I was wondering if **Peter Morris** and **Musicat** would agree to the validity of that type of test? ---
**Author:** ![Irishman](https://avatars.discourse-cdn.com/v4/letter/i/b487fb/32.png) [@Irishman](https://boards.straightdope.com/u/Irishman)\ **Post date:** [November 6, 2008, 9:26pm UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/169 "2008-11-06T21:26:29Z")
**Peter Morris** said: > [@](#): > > Just for a start it’s right there in Musicat’s description of the test. Read what he said. > > > [@](#): > > > > And if they pass the baseline test where it is NOT blinded … then, and only then, will they be invited to continue with a propery blinded one. > > Right there, you have the clear threat: Declare a success in the baseline test, or be disqualified. And the pass mark is fixed at 100%, whatever the dowser claims as his success rate. It’s an explicit threat. That’s the most preposterous reading of that sentence I can imagine. The whole point is to establish a test that the participant can pass. There is no point in continuing to a test when the participant has not demonstrated that the test method itself is acceptable. There is no statement “If you say anything other than you can detect this, you are disqualified.” What they say is, “Does this work for you? Can you pass this? Is there any reason this will not work for you?” **Musicat** ’s hypothetical example demostrates the method and the intent. Eliminate all the things that prevent them from succeeding. > [@](#): > > So, if they don’t pass the baseline, then they are disqualified. Your rules. No, they are asked how to modify the test conditions so they can pass. If the requested modification does not invalidate the test, then it is accommodated. Similarly, I will agree with you that the claimant may not claim 100% success rate. Maybe the only claim an 80% successrate, which is still above expected chance. Okay, so the testing protocol needs to be evaluated based upon an 80% rate not a 100% rate. Fine. But when running the preliminary tests to establish that the test conditions are adequate, one should strive to eliminate _all identifiable_ test interferences. It does no good to try to evaluate your performance for 80% when there’s an identified obstacle that will throw your performance off. If that obstacle interferes 20% of the time, then your 80% success rate just became 60%. So of course the test operators want the claimant to succeed 100% in the open pretest trials. The purpose of the trials is not to measure the performance rate, it is to establish that the conditions are valid and eliminate all identifiable interferences. Anything less than 100% success on the trials means there is another variable to the test than just the person’s ability. The point of the declaration is not to force the person to claim ability they don’t have, the point of the declaration is to acknowledge that all attempts have been made to eliminate interferences and provide the best conditions for the test. It is as much a requirement on the test administrator as the test claimant. It is a contractual bind that the administrator is giving fair conditions, not forcing the claimant to do something he doesn’t think he can. ---
**Author:** ![Irishman](https://avatars.discourse-cdn.com/v4/letter/i/b487fb/32.png) [@Irishman](https://boards.straightdope.com/u/Irishman)\ **Post date:** [November 6, 2008, 9:30pm UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/170 "2008-11-06T21:30:59Z")
**BunnyTVS** said: > [@](#): > > Hell you could even randomize which test the take first. I’m not sure what point that would serve. The point of the non-blinded part is to establish that the participant can function under the test conditions given. It doesn’t make sense to determine afterwards that “the guy in the corner with the red shirt is putting out bad vibes throwing my stick off.” I can’t tell the guy in the red shirt to leave for the non-blinded test if it’s already been performed. ---
**Author:** ![Musicat](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/musicat/32/20189_2.png) [@Musicat](https://boards.straightdope.com/u/Musicat)\ **Post date:** [November 6, 2008, 9:37pm UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/171 "2008-11-06T21:37:42Z")
> [@BunnyTVS](#): > > Before I go any further I apologise in advance for mentioning Randi. No need to for me. 🙂 Your protocol is simple in some ways, and that’s good. Here are the problems I see. It is imperative that you define what constitutes a success or failure _in advance._ Since there can be only one right answer per jug (it has water or not), figure out how many must be guessed correctly to constitute success. Since a 50% rate, on the average, would be expected solely by chance, you will need a much greater figure than that if you want to prove guessing isn’t the method used. If the claimant can get it 100% right in the baseline test, and says that cardboard boxes over the jugs make no difference, it is entirely logical to expect a 100% rate when blinded. Another reason success must be pre-defined is if you perform “data mining” on numbers _after_ the event, you can always find something “anomolous”, which is often used as proof of a paranormal force. Or in a large group, some will have a higher success rate than others (cf. “bell curve”), and if you point to those as proving what is claimed, you have just cherry-picked the data by ignoring those who had a lower rate. To be fair, you must include ALL the results, not just the ones you like. Sure, if it seems one person is luckier than others, do a re-test. Unless cheating is going on, or there really is some paranormal force at work here, the re-test results will likely converge on the 50% mark the more times you do them. It seems like one of the things you are looking for is a difference in performance between “experienced” and “amateur” dowsers, or even “disbelievers”. That would be indeed an interesting statistical analysis, but, again, _before_ you run the test, define what level of performance difference you will accept as confirmatory to your theory. As an example, what if your experienced group got 57% of the guess right, and your disbelievers got 49%. Aha, you exclaim, there must be something to it, as the dowsers were right 8% more of the time! Nonsense, as any statistician will tell you. What you actually proved is there is practically no difference. Is 8% significant or not? Keep your test simple. 10 jugs, one tester, expect 100% correctness. Anything else is failure. No judging needed, no statistical analysis needed, everything objective. And for those who feel 100% is too stringent, let’s make a bet for $100. Give me 10 pencils, some red, some blue – and we’ll agree in advance that they are indeed red or blue with no ambiguities – and I’ll bet you that I can tell you, 100% of the time, which ones are which when spread out on the table, in full view, in bright room light. Every time, 100%. Wanna see it again? Put your money on the table and let’s try it one more time. I can do this for as long as your money holds out. Why shouldn’t a dowser be able to do the same with water? On preview, I see that Irishman is suggesting loosening the success percentage requirement a bit down from 100%. I’m not insisting on 100%, although I feel it is not an impossible goal, but we would need someone better versed in statistics than I to tell us what number would be good enough to rely upon. I doubt that 60% makes any sense, but 90% sounds good enough to me. I’ll leave the exact number to the numbers experts. ---
**Author:** ![Musicat](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/musicat/32/20189_2.png) [@Musicat](https://boards.straightdope.com/u/Musicat)\ **Post date:** [November 6, 2008, 9:45pm UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/172 "2008-11-06T21:45:13Z")
> [@Irishman](#): > > Musicat’s hypothetical example demostrates the method and the intent. Eliminate all the things that prevent them from succeeding. > [Quote=Peter Morris] > So, if they don’t pass the baseline, then they are disqualified. Your rules. No, they are asked how to modify the test conditions so they can pass. If the requested modification does not invalidate the test, then it is accommodated. [/quote] Not only that, but if they can’t find water with a stick when it is out in the open for all to see, _then their claim to find water has been proved false even before the test has begun!_ Why keep going when the conditions only get tougher? Sure, they could lie and say everything is hunky-dory, but what purpose would that serve? If they can’t perform under ideal conditions, are their skills going to improve under harsher conditions? Just to avoid confusion, this post is primarily directed to Peter Morris. I’m taking Irishman’s post as a starting point for elaboration. ---
**Author:** ![Peter\_Morris](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/peter_morris/32/359_2.png) [@Peter\_Morris](https://boards.straightdope.com/u/Peter_Morris)\ **Post date:** [November 6, 2008, 11:06pm UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/173 "2008-11-06T23:06:24Z")
> [@Princhester](#): > > What if someone says they can find water lines, septic tanks, etc. with a coat hanger? Then you test their claim. See if they actually can find septic tanks underground. Go to some property with an underground septic tank, with the owner’s permission, have the dowser mark where he thinks the water lines or septic tank lies, then check the architect’s plans to see if he got it right. > [@MarcusF](#): > > OK, based on these ground rules, what was wrong with the German test described on Musicat’s post #89? (Please don’t say Randi’s involvement - your list allows for advice from a magician.) As far as I can see from the original write up it was set up in an appropriate manner and tested the specific ability the OP claims - "I can locate water lines, septic tanks, etc. " - but unfortunately there was no positive result. Perhaps the OP can do better? Your link doesn’t work. I don’t know the test you mean. But why would you assume that there’s anything wrong with it? Tests run by persons other than Randi are sometimes valid. Perhaps you are under the impression that I’m advocating dowsing. I’m not. I never have. I make no claim that dowsers would succeed in a fair test. To make this clear, I have several reasons to object to Randi, specifically, which don’t apply to other people running tests. Two reasons among many for objecting to him: 1. He is frequently seen to cheat. This only makes dowsers seem more plausible to the public, and damages the reputation of decent sceptics. 2. People are sometimes persuaded by his lies to give him money. These people are victims of a scam. ---
**Author:** ![Peter\_Morris](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/peter_morris/32/359_2.png) [@Peter\_Morris](https://boards.straightdope.com/u/Peter_Morris)\ **Post date:** [November 6, 2008, 11:20pm UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/174 "2008-11-06T23:20:37Z")
> [@Irishman](#): > > That’s the most preposterous reading of that sentence I can imagine. No, it’s accurate. > [@](#): > > The whole point is to establish a test that the participant can pass. There is no point in continuing to a test when the participant has not demonstrated that the test method itself is acceptable. There is no statement “If you say anything other than you can detect this, you are disqualified.” Wrong in every respect. The baseline test serves precisely one purpose. It is there to give Randi an excuse when his misbehaviour in the test is exposed. If the participant scores 100% on the baseline test, then Randi can dismiss every complaint made against him. So the rules contain a direct and explicit rule: score 100% or be disqualified. Every single person taking the test is doing so under that threat. And very few agree to do so. ---
**Author:** ![Peter\_Morris](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/peter_morris/32/359_2.png) [@Peter\_Morris](https://boards.straightdope.com/u/Peter_Morris)\ **Post date:** [November 6, 2008, 11:27pm UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/175 "2008-11-06T23:27:54Z")
> [@BunnyTVS](#): > > Imagine a dowsing test set up the same as those described by Randi and **Musicat** with a couple of differences. … > > I was wondering if **Peter Morris** and **Musicat** would agree to the validity of that type of test? I can’t see any obvious flaw in it. But I don’t understand the point of testing people who don’t believe in it. ---
**Author:** ![Telemark](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/telemark/32/372_2.png) [@Telemark](https://boards.straightdope.com/u/Telemark)\ **Post date:** [November 6, 2008, 11:30pm UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/176 "2008-11-06T23:30:34Z")
> [@Peter\_Morris](#): > > Go to some property with an underground septic tank, with the owner’s permission, have the dowser mark where he thinks the water lines or septic tank lies, then check the architect’s plans to see if he got it right. On a typical suburban house where I grew up, I can do that without any pretense to using dousing. I know where they are generally laid out, there are easy to see clues based on the way the grass grows and subtle changes in the terrain. Since I can do that without dousing, this test (and this claim) mean nothing. Anyone who installs and pumps septic tanks can do this to at least 90% accuracy. Unless people are claiming that the only thing they can find is a tank full of shit; but they never claim that. The claims are that they can find water – one example they give is finding septic tanks. ---
**Author:** ![geniios](https://avatars.discourse-cdn.com/v4/letter/g/9d8465/32.png) [@geniios](https://boards.straightdope.com/u/geniios)\ **Post date:** [November 6, 2008, 11:49pm UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/177 "2008-11-06T23:49:08Z")
My dad doused his family well when he was a boy. He doused several others over the years. In the 7os my cousin did not believe him so they dug… a salt water well. After a few years of living with that they decided to let him try. Took him about thirty minutes and the well they got was pure and clear. Yeah maybe just luck, over and over, just luck. He’s still alive if you want him to, I am sure he would oblige, he’s a gentleman. btw he does not charge for it ---
**Author:** ![Jackmannii](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/jackmannii/32/311_2.png) [@Jackmannii](https://boards.straightdope.com/u/Jackmannii)\ **Post date:** [November 7, 2008, 12:07am UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/178 "2008-11-07T00:07:44Z")
I sometimes douse my Labrador retriever, but only when she gets really rank. ---
**Author:** ![Princhester](https://avatars.discourse-cdn.com/v4/letter/p/3e96dc/32.png) [@Princhester](https://boards.straightdope.com/u/Princhester)\ **Post date:** [November 7, 2008, 12:32am UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/179 "2008-11-07T00:32:16Z")
What **Telemark** said. What precise controlled test protocol would you use **Peter Morris** to test the OP’s claims? A couple of sentences isn’t good enough. Nor is coming up with a protocol that is obviously flawed. ---
**Author:** ![Peter\_Morris](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/peter_morris/32/359_2.png) [@Peter\_Morris](https://boards.straightdope.com/u/Peter_Morris)\ **Post date:** [November 7, 2008, 12:39am UTC](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954/180 "2008-11-07T00:39:59Z")
I’ve already said that protocol design should be left up to a scientist who is experienced in designing tests. [Previous page](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954.md?page=8) [Next page](https://boards.straightdope.com/t/finding-water-with-coat-hangers/468954.md?page=10)