ARRS - Association of Road Racing Statisticians

The IAAF Takes a Giant Step Backwards

The IAAF has made their decision on keeping world road records. Two of the four criteria that were detailed in the IAAF press release (of a total of eight criteria) are good or at least reasonable. These are the requirements that the course have been measured by an "A" or "B" IAAF/AIMS approved measurer and that the decrease in elevation shall not exceed 1 m/km.

The doping control criterion has a caveat that "no new world bests will be ratified from 1 January unless the athlete has had a doping control on the date of the race." The 1 January refers to the 01 Jan 2003 starting date for the recognition of world records. Clearly, this is intended to "grandfather" in a bunch of "records" that do not meet this standard. Eventually, one would assume that all of the grandfathered "records" would be surpassed by performances that adhere to this criterion but drug-aided marks have a tendency to persist for decades.

Performances that lie six standard deviations out on the tail of the distribution, i.e., much faster than any other peformances at that distance (and for that sex) will tend to remain on such lists and cast doubt on the overall veracity of the process. In addition, drug-aided performances may still go undetected despite this criterion, and not be uncovered until the performances have been "institutionalized", i.e., not to be removed. If an anti-doping requirement is to be effected, then it should apply to performances prior to 01 Jan 2003. Otherwise, there is a double standard. Score this as a step backwards.

The definition of a "record-quality" course includes, "The start and finish points on a course, measured along a straight line between them, shall not be further apart than 50% of the race distance." This has been discussed at some length in the previous issue of ADR and again in this issue. The real question is "Why has the IAAF adopted this 50% S/F rule?"

Apparently one member of the committee (council) that made this decision insisted that courses with up to a 50% separation between the start and finish be allowed. It turns out that two of the performances on the list proposed for acceptance as "world records" were set in this person's country by nationals of his country. If a 30% rule were adopted, neither of these marks would have qualified. This serves to demonstrate that the basis for the IAAF road record criteria is political rather than technical. Score this as a big step backwards.

A glaring omission is that there is no provision for a validation. The USA has 20 years of experience with its validation process and we have learned that a number of things can produce a short course, no matter how expert the measurer may be. Altho the course itself may be accurate, the validation serves to determine whether or not the race was run as the course was certified. Were the start and finish lines (and any turn-arounds) in the correct locations? Were all coned restrictions adhered to?

Announcing a record and then having to make a retraction puts the race administration and to some extent, the NGB, in an embarrassing situation. One can see why a politician would oppose such a validation process. However, had the USA followed such a "no validations needed" policy, our records list would be full of garbage "records" as witnessed by the (minority) number of courses that have failed their validation. The validation of a course should be conducted by a person not connected with the race organization and, in the case of world records, should not be a citizen of the country whose course is being validated nor of the nation the would-be record-holder is from. Validations should also be done by measurers who maintain high technical standards. Score this as a giant step backwards.

The IAAF has decreed that women's records can be set in mixed races. While this is the present policy in the USA, this is not the direction that we should be going in. It seems evident that one motivating factor in having the elite women start before the elite men/masses start was in anticipation that the IAAF would restrict world records to races in which one could be certain that the elite women were not paced by men,as evidenced by a number of previous statements issued by the IAAF That this is a good idea for several reasons has been clearly demonstrated by the London Marathon and this year, by the New York City Marathon. It definitely enhances one's appreciation of the (women's) race when it is not buried in the men's race.

A positive step in this direction by the IAAF would have induced many more marathons (and other races) to take this step. This in turn, would have made it feasible for NGB's to rule likewise. This combination would have all but assured that the elite women would commonly have their own race (and glory). As it stands, marathons in particular are discouraged from having an early start for the elite women since a mixed race with male pacing would be more likely to produce a record. The IAAF really dropped the ball on this one, score it as a fumble, loss of yardage, and a turn-over.

Examination of the list of performances meeting all of the criteria reveals that intermediate times may be used for IAAF record purposes. This is a bad idea for a number of reasons. First, the care taken to lay out intermediate distances is rarely at the same level as that taken to lay out the start => finish distance. Often, intermediate distances are "laid in" with a single measurement. For an intermediate distance to qualify for "record quality" it should be subject to ALL of the same requirements that the overall course is subject to, including two measurements (within 0.08%), complete documentation, and location marks (PK-nails, etc) in the same way the finish line is marked.

Second, a special validation is needed to assure that the intermediate times were indeed taken at the proper location. This is much more difficult than with a finish line which is generally well photographed and well documented. As a minimum, photographs showing the runner passing the timer with sufficient ground detail to ascertain the exact timing location (to within a meter or two) should be required. All too often, intermediate timers are not positioned properly due to time limitations during the conduct of the race. In addition, there is rarely anything more than a sign and a person timing to indicate the intermediate "finish", quite unlike a "real" finish line, making it much more difficult to validate the "finish" for an intermediate distance.

Third, what about runners who are deprived of a "record" simply because the race organization did not provide timing at the corresponding intermediate point? For example, Naoko Takahashi is listed with "WR's" for 25 km and 30 km for intermediate times from the Berlin Marathon. The pace for those two "WR's" is 3:18 per km (for both). However, Paula Radcliffe averaged 3:15 per km for the entire marathon at Chicago, making it a virtual certainty that, had those intermediate times been taken, she would be credited with the 25K and 30K records.

Fourth, what about runners who run negative splits (like Khalid Khannouchi)? Perhaps race organizers could document an intermediate time at 12.195 km and use the time differential from that point to the finish for a 30 km record. That has just as much validity as using a time from the start to an intermediate point.

Fifth, allowing say 30 km records to be set during a marathon, greatly discourages runners from racing the 30 km distance. Likewise, allowing 20 km records to be set during a half marathon discourages race promoters from holding 20 km races. In case you hadn't noticed, the road racing scene is evolving to a future in which there are 5K, 10K, half marathon, and marathon races and little else. Look at the variety of different distances that runners had to choose from in the 1960's and 1970's. You will find the 25km distance commonly contested by the top runners. Today, only two significant 25 km races are left of the dozens that were held in the 1970's. Today, 30 km races are as scarce as hen's teeth (well, not quite but almost).

Why have records at distances such as 20 km and 30 km if they are dominated by intermediate times made enroute to longer distances and there are virtually no races being held at these distances? Wouldn't it make better sense to restrict records to actual race distances?

Sixth, what if a pacer goes through 30 km and "claims" a 30 km record but drops out shortly thereafter? There is a precedent that maintains a runner must finish the stated distance of the race in order for intermediate times to count. Is the race a 30 km race or is it a marathon? You can't have it both ways. Score a step backwards for this one.

Ekiden records??? They have to be kidding. This is another blatant political decision. How many countries actually hold ekiden races? I'll warrant that there are more countries holding 10 mile races than those holding ekidens and there are MANY more ten mile races than ekidens. What standards are there for ekidens? Are the lengths of each leg specified as well as the number of legs? The very few ekidens that have been held outside the Orient have, for the most part, been one-time affairs and have not proved popular. Give us a break. Score a big step backwards for this one.

This would have been a perfect opportunity to institute the use of fractional seconds in road record keeping, following the ADR practice of using tenths. However, this would be too much to expect from an organization that seems intent on trashing the credibility of road racing as a legitimate sport, on par with track and field.

The standards for record-keeping must be determined in a method as free from political considerations as possible. Back in the very early 1980's, I was given the task of setting up the rules whereby road records would be kept for the USA. Bob Hersh, chairman of the TAC Records Committee, called me on the phone and out-lined what he wanted me to do in an hour long conversation. I had no connection with a major race and was not actively involved with TAC except on matters related to course certification. I had started an organization known as the National Running Data Center with the purpose of gathering, organizing, and publishing records and rankings lists for the USA.

In order to provide meaningful lists, I had to consider course accuracies and the effects of drop and wind on performance. Having a sizeable database of race results, I was able to come up with what I felt to be reasonable requirements for recognizing road records. These included four main components: (1) a certified course, (2) post-race validation, (3) drop not exceeding 2 m/km, and (4) S/F separation not exceeding 10%. These were adopted by the USA in 1983 and were modified in 1988, relaxing the S/F separation to 30% and tightening the drop requirement to 1 m/km. The 1988 modifications were based on a more rigorous examination than the original rules were based on.

The choice of record distances was based on commonly run distances (in the USA). The choice of the 8 km distance rather than 5 miles was made so that either distance could qualify for a record. Until the 8 km distance was made a record distance, there were relatively few 8 km races in the USA. Now there are roughly equal numbers of 8K and 5M races.

These standards have stood the test of time quite well and the present USA road record list (now maintained by the RRIC) is generally recognized as very reliable. Now the IAAF comes along, 20 years later, and tries to re-invent the wheel. Rather than adopt what has already been shown to work, the process is thrown to the politicians and you see what devolves.

This latest decision by the IAAF does not promote the sport of road racing and if anything, degrades the record-keeping process to the level of political in-fighting. Overall, the IAAF has taken a giant step backwards.

Note - ADR will continue to maintain its present standards, without regard to what the IAAF proclaims as world records. The objective is to compile meaningful and reliable statistical lists.

This article appeared as an editorial in the "Analytical Distance Runner" for week 940, and was written by Ken Young.