More Statistical Evidence Orbs Are Paranormal

 Photo date: April 10, 2015. Photographer: Mark Mahin.

Orbs are unusual circular anomalies that appear in photos like the photo below. In previous posts I have done experiments debunking the "orb zone theory" -- the claim that orbs can be explained as flash reflections of dust particles near the camera lens. My tests showed that even when you raise heavy dust, you merely get small, dull, faint orbs unlike any of the more dramatic orbs shown on this site (which are so often big, bright, colorful, fast-moving or having what looks likes faces).

In this post I will discuss an entirely different type of experiment: a statistical experiment. The experiment will be done purely by analyzing a series of photographs taken on a particular day. In the experiment I will be looking for a certain type of “location bias” that we would not at all expect to see if orbs are just dust.

To explain this idea, let me first show a photo of the main terminal of Grand Central Station in New York as observed from one side of the East Balcony (the side opposite the US flag). The view looks like the one shown below. The strange blue items at the top are orbs.

You will notice that there is a row of lights that stretches roughly through the middle of the photo We can consider this row of lights as a divider. We can consider the area above the row of lights as the “upper area” of a photograph of the terminal made from this angle, and the area below the row of lights as the “lower area” of such a photograph of the terminal.

Now under the hypothesis that orbs are just dust, should we expect to see orbs more often in the upper area of photos like the one above, or in the lower area of such photos? Under such a hypothesis, there should be little or no difference between the frequency of orbs seen in the upper area and in the lower area. The theory that orbs are dust holds that the photographer is photographing little particles of dust floating around very close to the camera. If that theory is true, we should not at all expect to see orbs appearing much more frequently in the upper area of photographs like the photo shown above.

Consequently I had a nice opportunity for a statistical test. My procedure was as follows:

1. Using the photographs that I had taken on April 10, 2015, I selected all that were taken from an angle similar to the angle in the photograph shown at the beginning of this post. I threw out a few of these photographs which did not show a roughly equal amount of “lower area” and “upper area.” The row of lights in the terminal was used as the dividing line between the upper area and the lower area. I was then left with a set of 183 photos.
2. I examined each such photograph, and judged whether the total surface area of orbs shown in the photo was greater in the upper area or the lower area. If it seemed hard to tell whether the surface area of the orbs in the lower area was more or less than the surface area of the orbs in the upper area, or if no orbs were shown in the photo, I made no judgment regarding that photo.

The results were as follows: 154 of the photographs were judged to have a greater amount of “orb surface area” in either the upper area or the lower area (as defined above).  Of these 154 photos, 130 had a greater amount of “orb surface area” in the upper area, the area above the row of lights. Only 24 of the photos had a greater amount of "orb surface area" in the lower area, the area below the row of lights.

The table below summarizes these results.

Photos showing no orbs or no clear difference between the total surface area of orbs in the upper part of the photo and the total surface area of orbs in the lower part of the photo	29
Photos showing orbs in the upper part of the photo having a higher total surface area than the total surface area of orbs in the lower part of the photo	130
Photos showing orbs in the lower part of the photo having a higher total surface area than the total surface area of orbs in the upper part of the photo.	24

To put it succinctly, on this day (April 10, 2015) the abundant orbs in my photos seemed to appear much, much more commonly in the upper area of my photos.

Now how unlikely is it to have got a result such as I have just described, purely by chance? It's about the same as the chance of you flipping a coin 154 times, and getting 130 heads. If only chance was involved, and if orbs are just dust, then each of the photograph examinations I made (listed in the last two rows of the table above) should have been like a coin flip, with about a 50% chance of observing more orbs above the “row of lights” dividing line, and a 50% chance of observing more orbs below this dividing line.

But is there any way to calculate the chance of such a thing happening? Yes, there is a convenient online calculator we can use. The calculator is on this page at the stattrek.com website. The calculator uses a binomial probability calculation to estimate the odds.

The screen shot below shows the result I got when I typed the relevant numbers into the calculator.

The answer given by the calculator is that there is a 0% likelihood of getting these results by chance. The binomial probability (stated above in exponential notation) is calculated as about 1 chance in  3 quintillion (1 chance in 3,000,000,000,000,000,000).

So we have our answer to the question about odds. There is essentially zero chance that you might flip a coin 154 times and get 130 heads. There is also essentially zero chance that I would have got the results I got, under the theory that orbs are produced by flash reflections of dust floating in front of the camera. The statistical evidence I have presented here is extremely strong evidence against such a theory. The orbs in my photos at this location on this day were showing a very strong tendency to “preferentially” appear in the upper half of the photo area. Such a fact is powerful evidence against the theory that these orbs are produced by random particles of dust floating in the air. 

The video below is one that I presented in a previous post entitled "300 Mysterious Orbs Seen In One Hour."  The photos were all taken on the same day discussed in this post,  April 10, 2015. At the time I compiled this video last month, I was only interested in showing the photos that showed the most orbs, and was not paying any attention to the location of the orbs. You can see from watching this video what a very strong tendency there was on this day for the orbs to appear in the upper half of the photos.  The video is best watched in full-screen mode, by clicking on the full-screen icon on the lower right.

 The post here describes another day in which I observed orbs with a pronounced "vertical bias" that we would not expect from random dust particles.  The statistical evidence from that day (March 26, 2015) is even more powerful than the evidence on this day (April 10, 2015).

The following is a summary of the results on these 2 days:

Date	Likelihood of result occurring by chance
March 26, 2015	1 in 1,000,000,000,000,000,000,000,000
April 10, 2015	1 in 3,000,000,000,000,000,000