Published as a letter in Magonia Supplement 30 August 2000

I would like to say some words on a certain often repeated argument that goes: ‘The results of the Battelle Memorial Institute study showed that the better the sighting, the more likely it is to be unexplainable in terms of known phenomena, hence true UFOs do exist’. This study, commissioned by the USAF in the fifties, found that ‘excellent’ reports contained a higher percentage of ‘unknowns’ than ‘poor’ reports (besides ‘knowns’ and ‘unknowns’, there was a separate category for ‘insufficient information’ reports, so that couldn’t be counted). But can we really draw any conclusion from such heterogeneous data, such disputable criteria and so many factors playing their roles?

The usual assumption underlying the argument is: “If there were no true UFOs, the most reliable cases would have the lowest percentage of unexplained” I’ll try to show that this is dubious, at best.

The Battelle analysts divided the sightings into reliability groups, based on the quality, completeness and self-consistency of the report and upon the quality and experience of the witness. We can argue about how can this evaluation be accomplished in practice and about its true relevance, but this is not the point that I want to make here.

If we focus on the report side of this concept of reliability, it seems reasonable that the cases considered most reliable are the least likely to have erroneous data or to be incomplete in their descriptions and hence should have the least percentage of unknowns if there were no true UFOs. But what if we focus on the witness side?

1. We’ll assume that the most reliable witnesses are the least prone to experience misperceptions/misinterpretations that could lead them to report UFOs.
2. Let’s suppose that most, if not all, UFO cases are explainable, most of them as misperceptions/misinterpretations (for simplicity’s sake, we’ll set aside delusions, hoaxes, etc. ).
3. The key point is that the most reliable witnesses will report the least number of cases resulting from to misperceptions/misinterpretations, but that those reported will be in the class of those most difficult to explain, since they are not easily fooled by most stimuli, at least under normal conditions.
4. Therefore, after the analysis, the group of cases with the more reliable witnesses will show the larger percentage of unknowns. Note that in any group of less reliable witnesses, besides these difficult cases we’ll find many cases of misperceptions/misinterpretations more easily resolvable after analysis, that will count as knowns, so lowering the percentage of unknowns.
5. Finally, this trend will also appear in the overall analysis of cases vs. reliability, since witness reliability is one of the pillars of the general concept of reliability handled in the study. Hence we conclude that the cases considered most reliable should have the higher percentage of unknowns if there were no true UFOs!

Obviously, this is not to say that the Battelle results prove that there are no true UFOs. What I intended to show is that they don’t admit a straightforward interpretation as many ufologists think. I hope the examples below will help to clarify all this. Let’s start with a group of so-called “reliable witnesses” and another of average people. Now imagine that individuals from both groups experience the following situations:

• a) At night, in a secluded place, members of a sect perform a silent procession, holding torches and wearing black clothes. Casual observers are surprised by strange lights moving in circle near the ground for some minutes.
• b) An unusual red light (in fact, Venus) seems to approach the witness’s plane and keep pace with it for a while before disappearing at a fantastic speed. Later, the witness will report a wrong date for the event.
• c) Witnesses observe a landed “flying saucer” and, afterwards, an ascending green light, all arranged by sophisticated pranksters. Observers in both groups report the three sightings above as UFOs. Subsequent analysis fails to solve them and they remain as unknowns.
• d) A plane brightly illuminated by the sun makes an odd display in the twilight sky.
• e) A cloud in the upper atmosphere resulting from a ballistic missile secret test is taken for a mysterious nearby phenomenon by some observers. Again, witnesses in both groups report these sightings as UFOs. But this time, subsequent analysis finds the right explanations.
• f) One night, an observer from the ‘average’ group discovers the hovering lights of a phantom airship. A nearby reliable observer recognizes Venus and Jupiter, very close in conjunction.
• g) A yellowish disc follows the car of a witness of the “average” group for many miles and, finally, it seems to land behind some trees. Shortly after, a ‘reliable’ observer experiences the same, but when he stops the car to better observe he quickly realises he has been watching the moon. In these two last examples, only observers from the second group (less reliable observers) report seeing UFOs. Analysis comes up later with the correct identifications.

To sum up, the resulting proportion of unknowns in the ‘reliable witnesses’ group is 60% (3 out of 5), while in the ‘average witnesses’ group it is about 43% (3 out of 7)! Hence, even if there were no true UFOs, the cases with most reliable witnesses would have the highest percentage unexplained.

Manuel Borraz, Barcelona, Spain