Column by Steven V. Selthoffer, Chief European Columnist
DEBRECEN, Hungary, June 6. THE 31st LEN European Long Course Championships concluded with controversy once again regarding the timing systems used in the sport. Initially, this time, the timing system problem was regarding a relay exchange and no-show data streams that failed, however, the problems may go much deeper than originally suspected.
The 31st LEN European Championships were held as another chance for swimmers to showcase their talents in an attempt to achieve FINA A and B Olympic qualifying times prior to the upcoming London 2012 Olympic Games. However, the swimmers' efforts and their times and places may be in jeopardy.
Preliminary Analysis
The 31st LEN European Championships appear to have a high number of ties (meaning two or more swimmers with the same time, to the hundredth of a second, in the same heat, or event) in relation to the number of swimmers in the event and/or in the meet compared with other championship meets.
The high numeric quantity of ties apparently occur almost predictably at certain rank/places, 6th, 8th, 13th and 26th place. The total number of ties appear to be in two separate clusters/groups. And there appears to be an unusually high number of ties in the semifinals and finals in relation to other championship meets.
This is cause for concern regarding the integrity of the system, the actual times and places of the swimmers, and if the same system and/or engineering, design and components will be used in the upcoming London 2012 Olympic Games.
Process and Examination of the Charts
The entire 31st LEN European Championships were examined and analyzed by every heat, in every event, in every day, for the full seven-day period. While the analysis is only preliminary and not definitive (there may be errors) there are some startling revelations.
All of the ties that occurred were recorded in each day, by place and the number of individual swimmers in the tie were recorded also. Charts were made from Day 1 through Day 7, that were color coded, assigning a color for the ties occurring in each particular day, and then assigning one square to represent one swimmer in each tie. So, if there was one tie for 8th place, on Day 1 in an event, two red squares would be colored in for 8th place on the Day 1 chart. If there was another tie for 8th place on Day 2, between two swimmers, then two orange squares would be colored, in the 8th place tie column, and so on.
Then, there was a Total Chart made of all the ties and places for the entire LEN championship meet so that any repetition and total accumulation could be seen in the data.
Results Confirm Clustered Repetition
The preliminary analysis of the number of ties in the championships and the clustering and frequency of certain specific ties repeating in certain specific numerical places in initial observations is confirmed from the preliminary data charts.
Please see FULL ANALYSIS CHARTS.
The ties in rank/places 6, 8, 13 and 26 beginning on Day 1 appear to be compounded and repeat four times the following days.
Ties for 3rd, 11th, 12th, 21st, 22nd, 24th, 35th, and 36th, happen three times.
On Day 1, of the 18 separate places that contains ties that day, 72.22%, of the places repeat over the next six days.
On Day 3, 14 out of 20 of the ties, 70% are exactly on rank/place with other ties from Day 1 and/or Day 2, including 5th, 6th, 8th, 13th, and 26th places. That does not demonstrate random distribution.
Two Clusters of Ties
The ties appear to be clustered in two distinct groups, one large cluster from 3rd to 26th places and a second minor cluster from 34th to 38th places. The distribution of ties does not appear to be random.
After a tie in Day 1 for 25th place, on Day 3 there were three more ties in three separate events also for 25th place. That is a very unusual anomaly.
On Day 1 in the men's 100m Backstroke there were six different ties involving 12 swimmers. Then the ties in those specific places, 6th, 10th, 19th, 21st, and 36th continue to be the base of the ties and continue in other events, on other days through the championships.
Semifinals and Finals
There appears to be a high number of swimmers in semifinals and finals involved in ties, given the total number of swimmers in the meet. Of the 11 places that contained ties in the semi-finals and finals, 72.72% were the same and repeated from the heats on Day 1.
In the USA Swimming Nationals there are sometimes 150 – 190 swimmers per event, with A, B, and C finals. This meet was much smaller.
However, in the 31st LEN European Championships there were 42 swimmers involved in ties in the semi-finals and finals. And there were a total of 170 swimmers involved in ties for the 31st LEN European Championships.
2CY Chemometry Consultancy
The top European consulting agency on chemometrics and statistics, 2CY Chemometry Consultancy, Beek-Ubbergen, NED, has been given the charts and has been notified of the meet results and ties. Klaas Faber, PhD, and CEO, 2CY Chemometry Consultancy, briefly discussed the results today, expressing his concerns and is awaiting more data and comments from Omega and FINA. 2CY Chemometry Consultancy provides consultancy for statistics conducting research with a number of universities, corporations and non-profit institutions in Europe and around the world.
Omega Timing
Laurence Germiquet, Swiss Timing Ltd., the developers of the Omega Timing System used at the meet stated in an email, “The statistics (number of ties) what we can tell you about are the following Championships. Rome World Championships 2009, Budapest European Championships 2010, Shanghai World Championships 2011, Canada Olympic Trials 2012…
According to internal research ties are between 3% – 4%. (The) Debrecen European Championships 2012 were also in this percentage. We controlled the ties in our company in 1,000 (0.000) of a second and we had only one tie, and this (in this particular) one in 10,000 (0.0000) of a second was a difference.
For Rome, Budapest, Shanghai and Canada we used the ARES system. In Debrecen we used the QUANTUM system. For this reason we could go to the 10,000 of a second.
We are open to discussion with FINA and/or LEN for any further questions.”
Omega's response was timely and appreciated.
FINA and LEN
FINA and LEN have been contacted. LEN has no further comment and there has been no reply from FINA as of this date.
More Data Needed
Analyzing the data alone so far can be risky without agreed upon comparisons to other European championships with the same timing system and a different one, in order to give knowledge of the context.
That will be accomplished in the next phase. We will continue to report on these developments as they happen.
Note:
1.) We wish to thank Laurence Germiquet, on behalf of Peter Hurzeler, Swiss Timing Ltd., for their contributions to the article. Swiss Timing Ltd. www.SwissTiming.com.
2.) Errors. We have reviewed thousands of swimmer/events over the past 10 days for the three Olympics, U.S. National Championships, the 31st LEN European Championships, and other meets. These are only preliminary figures for discussion purposes only. They are not definitive and may contain errors. Thank you for your consideration.
