CFEPP 1996
1066 sportsmen

This page deals with the third and last test done by a group of skeptics, the CFEPP (Comité Français pour l'Etude des Phénomènes Paranormaux - French Commitee for the study of paranormal phenomena). Results were published in 1996 in a booklet ; the official published list contains 1066 sport champions.
Two sources were used to include this test in g5 database: the CFEPP booklet and final3, a file sent by Jan Willem Nienhuys, who kindly authorized its use for inclusion in g5 database.
Among the 1120 records contained in file final3, 975 are already in Gauquelin or Ertel data. This file then brings 128 new records.
A complete restoration of this group would be quite complex, because it contains several subgroups:
  • The original selection: 1439 records.
  • The official list published: 1066 records.
  • Additional correct records found by Nienhuys but forgotten by CFEPP: 54 records (see below).
  • Additions suggested by Gauquelin: 79 records.
  • Deletions suggested by Gauquelin: 16 records.
  • Cases for which neither the CFEPP nor MG has been able to obtain information: 3 records.
For the moment, inclusion to g5 database is limited to 2 groups: 1066, 1120 (= 1066 + 54, see below).



This test gave rise to two documents:
  • Preliminary report, a CFEPP internal document, containing a list of 1439 selected sports champions.
  • The "Mars Effect", A French Test of Over 1,000 Sports Champions, paper booklet (157 pages), published in 1996 by Prometheus Books.
    By Claude Benski, Dominique Caudron, Yves Galifret, Jean-Paul Krivine, Jean-Claude Pecker, Michel Rouzé, Evry Schatzman, with a commentary of Jan Willem Nienhuys.
    This book can be bought online.

Information sources

  • Two files related to this test can be downloaded from : 4b_cfp1120.csv and 4c_cepp1066z.csv
  • Another file can be obtained upon request from Jan Willem Nienhuys ; you can find his email on page "The Mars Effect in Retrospect".
    This file was retained to integrate CFEPP test in g5 database because it is more complete than the files from It contains more fields, in particular both legal time and universal time ; its format is almost identical to the official list published by CFEPP.


This test has been mainly analyzed by two persons : Suitbert Ertel and Jan Willem Nienhuys, and references can be found in several articles from these authors.
TODO List the precise articles

The input file

The file sent by Nienhuys is called final3 and is versioned with the program, in data/raw/cfepp/ ; it contains 1120 persons.

Different groups related to this test

Several groups are related to this test, listed in their order of publication.

1439, by CFEPP

This list was publihed in 1990 CFEPP's preliminary report. It corresponds to the raw selection done by CFEPP members, extracted from the books they used as sources:
  • L'athlege - Biographies des plus grands champions français de tous les sports, Kléber, 1951.
  • Le Roy - Dictionnaire Encyclopédique des Sports, des Sportifs et des Performances, by Bernard Le Roy, Denoël, 1973.
Among all the athletes listed in these books, the CFEPP kept only the sportspersons corresponding to their selection criteria.

Personal communication from Nienhuys permits to have more information about this group:
  • Jan Willem Nienhuys thinks that the selection ended in 1987.
  • The original list contained 1442 persons, and 3 were removed because they didn't fit the criteria.

1066, by CFEPP

This is the official list published in CFEPP booklet, which is a subset of the 1439 list. Finally, after eliminating athletes for whom complete bith data could not be found, 1 066 athletes were compiled. (CFEPP booklet, p 14)

1076, from Ertel

A group of 1076 persons is used by Suitbert Ertel in the analysis exposed in the book "The Tenacious Mars Effect". This number of 1076 is given p SE-13 of this book: CFEPP's birth data were scanned from an appendix of the French Commitee's report (Benski 1990, N = 1,439, complete information for N = 1,076).

1120, by Nienhuys

final3 contains 1120 records. This is explained in an article by Nienhuys, Ertel’s ‘Mars Effect’: Anatomy of a pseudo-science, in paragraph "The CFEPP memorandum of 1990": « The CFEPP had also neglected to sort out a small pile of town hall answers that had been kept separate from the main bulk because of some need for follow-up actions. In this pile there were 11 more names with good data . Altogether the number of data to be added to the ones in CFEPP’s memorandum was 11 + 43 = 54, thereby raising CFEPP’s 1066 to 1120. » JW Nienhuys sent supplementary explanations (personal communication): « CFEPP had received answers from city halls that hadn't been entered into their database, for no clear reason. I have taken the position that any form received from town halls should be included, unless the protocol clearly forbade that. But the protocol said nothing about a cutoff date, and the forms weren't marked with a date of reception.
I suspect that at a certain point in time a list was prepared with all data received and afterwards, when the printouts were ready and one of them sent to Gauquelin, some more forms were received and put together in a folder that carried the inscription 'what should we do with this?' or words of similar nature, and then forgotten. By the time I did my checking there was nobody left who could provide any information. »

Fields of the raw file

As Nienhuys' file roughly contains the same fields as the published list, the meanings of the fields can be easily understood because they are precisely described in CFEPP booklet (p 48).
CFEPP explanations are reproduced in the following table.
CFEPP name G5 name
(in tmp csv file)
SRC SRC Information source
  • A: L'athlege - Biographies des plus grands champions français de tous les sports, Kléber, 1951.
  • L: Le Roy - Dictionnaire Encyclopédique des Sports, des Sportifs et des Performances, by Bernard Le Roy, Denoël, 1973.
  • G: Gauquelin's study of 568 champions.
LV LV Level
  • M: World record or championship
  • O: Olympic medal
  • E: European medal or championship
  • F: French record or championship
  • I: Selected for international competition
LOC DATE DATE Legal day = day written in the birth certificate.
Not present
in CFEPP booklet
TR Time regime
This field is not present in the original list, and was added by Nienhuys, who sent explanations permitting to understand its meaning:
Note that following the date there is a column with a letter, indicating which 'time regime' applies:
  • f for France,
  • o for 'occupied zone during WWI'
  • a for Alsace-Lorraine between 1880 and 1918.
LT DATE Legal time = time written in the birth certificate.
POSTAL CODE C2, C3 This field is used to compute département (French administrative code level 2) and arrondissement (French administrative code level 3) for Paris.
LONG LG Longitude, in decimal degrees.
LAT LAT Latitude, in decimal degrees.
UNIV-DATE DATE-UT Day of date expressed in UTC.
UT DATE-UT Time of date expressed in UTC.
S M12 Mars sector (from 1 to 12)

G5 integration

Inclusion of this file in g5 is done using Ertel's file of 4384 sportsmen, which contains the CFEPP ids.
Import is done issuing the following commands:
php run-g5.php cfepp final3 raw2tmp
php run-g5.php cfepp final3 ids
php run-g5.php cfepp final3 tmp2db small

Looking at data

Step raw2tmp converts file final3 to data/tmp/cfepp/cfepp-1120-nienhuys.csv.
Step ids adds Ertel, Gauquelin and Comité Para ids to this tmp file. The code permits to identify 129 wrong CFEPP ids in Ertel file, and 54 missing associations. One record (Marcel Domergue, A1-2089) is not present in Ertel file but already in Gauquelin data.

Other commands permit to display additional informations :
php run-g5.php cfepp final3 look ertel
N invalid = 3           # aberrant values in Ertel file
N different dates = 62  # different birth days, between Ertel's file and final3
Once included in database, some queries permit to provide more informations:
  • Number of persons in CFEPP group (of 1120) already present in Gauquelin data:
    select count(*) from person where partial_ids->>'cfepp'::text != 'null' and partial_ids->>'lerrcp'::text != 'null';
    => 710
  • Number of persons in CFEPP group (of 1120) already present in Ertel's file:
    select count(*) from person where partial_ids->>'cfepp'::text != 'null' and partial_ids->>'ertel'::text != 'null';
    => 975
  • Number of persons in CFEPP group (of 1120) already present in Comité Para test:
    select count(*) from person where partial_ids->>'cfepp'::text != 'null' and partial_ids->>'cpara'::text != 'null';
    => 361

Miscalaneous remarks

Paris arrondissement

One of the interesting feature of Nienhuys file is that most births in Paris come with the arrondissement (administrative code level 3).
This information is missing for 6 persons:
126  SCHOEBEL Pierre 1942-12-24
215  PERRIER Jacques 1924-11-22
231  BRETONNEL Fred 1905-06-11
638  DUBESSAY Jean-Francois 1921-04-01
645  MANOUKIAN Diran 1919-03-22
1048 COLAS Paul 1880-05-06
1 personin the file has no given name: MARTIN-LEGEAY

CFEPP unique id

An id, stored in column CFID of tmp files, is assigned to each record of the file.
Equals to the string "CF-" followed by the position of a record in file final3.
Goes from CF-1 to CF-1120.
Records CF-1 to CF-1066 correspond to the list published in CFEPP booklet.

CFEPP unique id is one of the partial ids defined in g5 database. See complete list of partial ids.