Skip to main content

A Validation Question from Jim Gerweck

In a message dated 6/2/2006 7:18:44 A.M. Eastern Standard Time, jim@runningtimes.com writes:

Dear Pete,

I'm heading up to Albany to pre-validate the revised Freihofer's 5K
course at the request of George Regan, the RD. Kevin Lucas did the
measurement and will accompany me on the ride. He said when they
measured the course there were cars parked on some of the park roads
(they clear them out early Saturday for the race, maybe the only time
that happens all year) so they did some offsets. Given that, I'm
anticipating the course might come out slightly short.

Here's the validation drill as I understand it:

Pre- and post-cal as normal

One ride of the course.

Use the average constant and DO NOT use the 1.001 SCPF.

Here's my question: Having done all that, what figure am I looking
for - 5,000m, or 5,005? I assume the former, and that if the course
comes up short of that it must be lengthened. George is obviously
looking for records, and said he doesn't want the course
unnecessarily long, either (although I doubt this will happen). So
basically, I just want to know what final figure I should come up
with, after adjustment.

Thanks,

Jim

Dear Jim,

Use average constant, and if the course comes out to 4997.5 meters or more it passes. When you are done, the course, if less than 5005, should be adjusted to 5005. If more, leave it alone.

Do NOT use the 1.001 SCPF

If Kevin rides with you his ride can be used as the second measurement if the course comes out to less than 4997.5 and has to be recertified.

That's how I understand it.

Neville Wood, RRTC Validations Chairman, has the final word on this.

I am going to put this on the BB. More opinions may be forthcoming.

Best, Pete
Last edited {1}
Original Post

Replies sorted oldest to newest

Based on Pete's comment that "if the course comes out to 4997.5 meters or more it passes", this suggests that measurement accuracy has been determined to be +/- 5%. Was there a study/analysis done at some point to determine that this is indeed the case? This question has been asked of me by race directors and I've kinda avoided giving a direct answer....
The negative "mercy allowance" of 5 m in 10 km came into being when it was realized that to shoot down a marathon course because it validates out to 42194 meters would be ridiculous. We can't measure that well. After all, if a 10 km course measures out to exactly 10,000 meters there is a 50 percent change that it is short. We don't know the truth and never will.

Some opted for 10 m in 10 km (me included). Others said zero. 5 m in 10 km was adopted. No study was made except for examination of validation records. It was an arbitrary decision that most thought was fair enough.

Our rule on validations requires the course to be "shown" to be short to fail, which implies that some level of proof should be present.
Surely, does not the SCPF take care of the fact that "we can't measure that well"??? If the measurer has done his job correctly then a 5K course is 5005 meters. If the course is validated at less than 5000 m based on the method Pete described, then that sounds to me that the course is definitely short without any question.

By the way, I'm sure that someone will notice that in my earlier posting I intimated that 2.5m in 5000 m is 5%. This was obviously in error. It is 0.05%.
I was not suggesting that measurers need to be punished for making mistakes. Mistakes happen. I'm in agreement that validation should never be a finger-pointing exercise, rather a learning experience for the measurer (or confirmation by an expert measurer that the original course measurer did a good job). It just seems to me that the 5 m SCPF for 5000 m is already plenty of so-called mercy factor.
Maybe I should wait until a course that I have measured undergoes validation before I reply about how big the mercy factor should be. Anyway, if the expert opinion is that the validation measurement for a 5000 m course needs to be more than 4997.5 m and not merely greater than 5000 m to invalidate a measurement due to the degree of mesurement uncertainty, should not the SCPF be 1.0015 and not 1.001? What am I missing here?
You are not missing much. There are a lot of things to be addressed, and what stands in the way of a perfect solution is our admittedly imperfect measurement technique.

The arbitrary things we use have stood the test of time, and that's a positive thing. No substantial proof of the validity of our methods yet exists.

Last time I looked at the validation figures I saw that about 90 percent of the courses passed, when using all measurements. When using only those measurements done by certifiers or highly experienced people, the figure is closer to 99 percent. Those are the numbers. What people choose to make of them will differ.
I was not aware of the proposed validation. In postvalidation the Validations Chairman appoints the validator, and the same should be true for prevalidation even though in this second case the race director pays for the work.

If validation indicates that the course is 0.05% short of the actual distance, the race director has an option not mentioned above in that he can leave the course unaltered and forego prevalidation.

It would probably be more logical to certify courses without the SCPF and accept a certain degree of shortness in validation.
Neville, I couldn’t agree with you more on your final statement. From a purely engineering viewpoint, I abhor the confusion between “SCPF meters” and “real meters.” I believe it would make more sense to lay out courses right at the nominal length, without SCPF, and accept a certain amount of shortness upon validation.

I would consider such courses as “reasonably accurate.”

However, the present grows from the past. In Measurement News #84 (July 1997) a letter from Ken Young describes how things got the way they are. Around 1980 Ted Corbitt decreed that an extra 0.1 percent must be added to each course if certification was to be granted. Ken did not mention this, but I suspect that Ted was influenced by Ken to do this, as Ken was the US’s first record keeper, and he has always adhered to the idea that “a course shall not be short.”

In those days it was not possible to effectively argue with Ted, partly because of his already iconic stature and partly because there was no forum in which to do so. Also, there had been few validations and no set procedure was in place, and the confusion between the two types of meters had not yet become apparent.

The years rolled by and now the SCPF is accepted worldwide. It may be an abomination in the minds of some, but it does make sense from the record-keeper’s perch.

If measurement existed in a vacuum it would make things easier and more logical, but, as in the rest of life, there are various viewpoints to be considered, and various compromises to be made.
Pete:
I was interested to learn something of the history of the SCPF. My remark about it does not mean that I advocate change anymore than I would for the layout of the computer keyboard. If the SCPF were not used, then the degree of shortness acceptable for validation should be that which falls within the normal statistical variation of the measurement. Therefore this would be accurate rather than reasonably so.

Add Reply

Post
×
×
×
×
Link copied to your clipboard.
×