From       (1600 words 2 graphics)       Home       Fast-Find Index

The $110,000 Dowsing Challenge
Were they making ESP history? Absolutely not

By James Randi

Abridged from an original article in Skeptical Inquirer 8(4) Summer 1984, reprinted in the Skeptic 4(3) September 1984. A British dowser claimed he could determine the on-off status of an electric circuit, and tried for the Australian Skeptics $100,000 prize and James Randi's (then) $10,000 prize. But tests showed the dowser could hear and see the on-off status. Dowsing was not needed. His challenge failed. James Randi is a professional magician, author, and world-famous investigator of unusual claims. He has logged over 100,000 miles a year in his checks on pseudoscience. In 1976 he was a co-founder of CSICOP. In 1996 he founded JREF, the James Randi Educational Foundation for promoting rational and critical thinking in paranormal areas, see


In August 1983, Australia's Dick Smith received a letter from Robert Homer of Worcester, England, who was the organizer of the Worcester Dowsers. The letter said that John Rainbow, a gifted young dowser, had performed a startlingly positive test on videotape and that a copy was on its way to Smith for his viewing. Said Homer, "I am sure that we are making ESP history." If his claims were correct, that would certainly be true.

The dowsing claim
Homer had constructed a simple randomizing device mounted in a box. An electronic circuit delivered a small current to a 500-ohm resistor in pulses occurring a few times every second. Its on-off status could be viewed via a sensitive meter on the face of the box by depressing a "read" button, which in its off position was supposed to short out the meter so it read nothing. After the circuit had been pulsing for a while, Homer would activate a switch to freeze the circuit's status (on or off), which the dowser then tested with his pendulum. Finally Homer would depress the "read" button to determine the actual status of the circuit.

In each trial, Rainbow had waved a small jeweled pendulum near the resistor, and apparently (by watching the pendulum's movement) was able to determine whether the current was on or off. Homer invited Dick Smith to view this process in person, and to surrender his offer of $100,000 upon seeing a successful demonstration.

Dick Smith is amazed
The videotape arrived in due course, and Dick Smith was amazed to see that Rainbow had beaten odds of over a million-to-one, calling twenty out of twenty trials correctly! If necessary, Rainbow could opt out of any trial by declaring it "void". Smith immediately called me and asked if I would go to the UK on his behalf to accept the challenge. He offered to pay my expenses for the trip, and after seeing the videotape, I agreed. He forwarded to me his check for $100,000, to which I added my own $10,000. In the UK I would have the help of others to make the arrangements, record the tests, and be impartial witnesses. The date of the test would be 8 January 1984.

My viewing of the videotape had aroused my suspicions. But the box differed somewhat from that on the videotape -- it had been painted and a carry handle had been added. Furthermore, the possibility of trickery was offset by Homer's offering to give me the box as soon as the test was completed. As it turned out, this was not needed.

Dowser scores 100% before witnesses
Homer had intended to hold the test in the living-room of Rainbow's home. I objected to this, since neutral ground would allow better control of possible trickery, so the test was made in a local hotel. Homer and Rainbow signed a document agreeing that all was suitable for the actual test. With my helpers present, I then asked Rainbow to perform a preliminary test run of 10 trials. These trials were to be followed by 20 formal test trials, where at least 16 correct was required to win the $110,000. The probability of success purely by chance was about one in two hundred. As in his videotaped test, Rainbow was allowed to declare any trial "void," which would then be ignored in the scoring. I was the scorer, and my helpers carefully watched what I wrote down.

Homer switched on the randomizer and depressed the "read" button temporarily to show that everything was working. Then, after a pause, he threw the "freeze" switch and asked Rainbow to make his guess. The dowser sat concentrating and waving his pendulum. Due to a recording error (my fault), only nine trials were recorded in this preliminary run, yet Rainbow scored 100%! However, all was not well. My helpers (with some puzzlement) had noticed that, after the first three trials, I began writing down my own guesses -- all of them correct -- before Rainbow called out his. What could be happening?

Problems with pinging and ponging
At the outset I had noticed a few problems that were not apparent from the original videotape. First, the rate of oscillation was about 3 to 5 times per second, which was insufficiently rapid to prevent direct guessing of on and off. Second, there had been no tests to make sure that the durations of on and off were the same. Third, the oscillator in the box was quite audible to all present, even though Homer denied being able to hear anything. It was making pinging and ponging noises as it oscillated between off and on. By listening carefully, I could tell the circuit's status (off or on, ping or pong) when the "freeze" switch was thrown. To get around this, we decided to drown out the oscillator noise by rapping on the table during each trial.

To score 100% just read the meter
Following that amazingly successful trial run by the dowser, I sat in Rainbow's chair and used his pendulum for trial runs of my own. Despite the table rapping, my guesses were always correct! I had noticed that whenever the "freeze" switch was thrown, the meter needle ended up just below the zero point when the circuit was off, and just above it when the circuit was on. Evidently the "read" button was not the perfect electrical short it was supposed to be, which then allowed the sensitive meter to detect the difference between on and off. So by simply glancing at the meter needle I was able to score 100%. One of my helpers tried his skill, and he too scored 100%.

For zero success try dowsing
Now, had Rainbow been using this method to make his guesses, or had he been using his dowsing powers? To find out, I suggested that he now perform the run of 20 formal tests, but with the meter covered over when it was not being read. Rainbow said he was uncertain, tired, and rather rushed. Homer objected that the change was not in accord with the original test procedure. But I was in charge of $100,000 of Dick Smith's money, plus $10,000 of my own, and I was not about to be put off.

Eventually Homer and Rainbow agreed to the run. I now surrendered the checks for $110,000 to a helper, either to be awarded by him to Rainbow if the test was a success, or to be returned to me if not. We then ran the 20 teats with table rapping to hide oscillator noise, and a slip of paper covering the meter. Rainbow scored 13 out of 20, which was within chance expectation. The two checks were returned to me, and the test was over.


Did the dowser cheat?
The question remains: Did John Rainbow cheat by reading the meter needle? Certainly he could have. One fact I have not revealed: On the third trial of the preliminary run (which Rainbow had called "void"), with the meter fully exposed, I had entered in full view of my helpers a "V" for "void" just before Rainbow called it out. I had noticed that he leaned forward after the "freeze" switch was thrown, ostensibly to see if it was in the correct position. He then fiddled with his pendulum before making his announcement. But I believe that he was actually getting a glimpse of the needle, which I could clearly see from where I was, but which he could not easily see unless he leaned forward. On that third trial he did not lean forward, so I suspected he would announce a "void" -- and he did.

It seems significant that Homer and Rainbow could use the box for more than two years without becoming aware of its faults. It also seems significant that when the table was rapped to guard against auditory cues, and the meter was covered to avoid visual cues, Rainbow's performance on both tests dropped from 100 percent to chance level. There is of course one way to resolve the matter -- repeat the tests using a simple coin-toss in another room to switch a simple battery-resistor circuit on or off. There would be no conceivable way, short of clever electronics or collusion, that Rainbow could obtain his claimed success rate unless he actually had the claimed dowsing ability.

The usual excuses
Shortly after the failed test, Homer provided a long list of excuses for Rainbow's failure. For example there were tape-recorders running, the pattern of the table cloth was confusing, the pendulum tended to get tangled, and there were many delays and distractions. Well, one thing has been forgotten -- John Rainbow scored 100% with the tape-recorders running, and with the confusing table cloth, the stubborn pendulum, and the delays and distractions. All until the auditory and visual cues were removed. Then he failed. I look forward to a definitive test of John Rainbow's dowsing ability. Both prizes -- Dick Smith's and mine -- are still available.

[No such test has been reported in the two decades since]

From       (1600 words 2 graphics)       Home       Fast-Find Index