5.1.1 add explicitly that the measurement can be made after for 1h/24h TT WR

---

Comments about this discussion:

Started 12 Sep 06:33

Organizing a time trial record attempt isn't easy. If the competitor wants to put all the chances on his side, must have the course measured as precisely as possible (in particular with the method recommended with the Jones counter).

This measurement must be carried out before a record attempt for a fixed distance (100km, 100 miles, but it's also valid for race records: 10km, marathon or 100km). This measurement is essential for positioning the start and finish.

It's different for an hour or 24-hour record attempt. For these attempts, it's important to identify the shortest route actually taken by the competitor, and that there are witnesses and good timing. Make sure that the competitor has crossed the line (start = finish) one last time after the hour or 24h. A precise measurement of the course can be made after the attempt, in order to calculate the distance covered as accurately as possible in the allotted time.

I think that explicitly adding a reference to this in 5.1.1 for 1-hour and 24-hour records could take some of the burden off a competitor who wishes to make an attempt. In the event of failure or suspicion of failure, it would not be necessary to take this measurement, which is a major difficulty in the preparation of a time trial record.

Comment 13 Sep 02:24

I think it's a very valid point that you raise Simon. The current guidelines do not specify when the course for the 1-hour and 24-hour time trial records is to be measured - an explicit indication that this measurement can happen retrospectively is not a bad idea from my point of view.

Comment 13 Sep 02:45

For any race to be valid, I agree that a course must be measured and verified first.  Although does a fixed distance record attempt really have to be measured first in every instance? Surely that is the ideal and of course highly recommended, but shouldn't the criteria be that it must be measured and verified before submission as part of the evidence pack?

Comment 13 Sep 03:03

In races over a certain distance, however, the course must necessarily be measured beforehand so that you know where the finish is, i.e. between which points you take the time... otherwise it will be difficult to set a record over this specific distance.

You could of course measure “roughly” beforehand and add a safety margin. Then you might ride a little further (which is allowed and not a problem) and only measure exactly afterwards when you reach a WR. That would also be okay for me in principle, but of course there is the risk that you have made a mistake during the pre-measurement and in the end find out during the necessary exact measurement that the distance ridden was too short and the performance is therefore invalid. But basically you are right from the pure point of view of whether a WR can be recognized or not, in the end there must be proof that the distance ridden corresponds at least to the desired distance.

Comment 13 Sep 11:19

I assumed that the measurement was necessarily taken beforehand for fixed distances, whether in the case of a race or a time trial (TT) record attempt. But perhaps this needs to be clarified.

Here's what I think about making an "rough" measurement and adding a safety margin for a fixed distance (race or TT attempt):
1/ for the UNICON20 marathon, a rough measure was done. In the end, the course was almost 500m short, even though they thought they had measured more than 42.195 km.
2/ if you run a 43km marathon instead of 42.195m, it takes almost 2 minutes longer to finish the course, so you can forget about any record.
3/ it's not possible to calculate a 42.195 km time from a 43 km time. If this were possible, it would greatly simplify things, but it's not very qualitative. It might be a decent solution, but it would be a step backwards.

In all cases (race or TT attempt / fixed distances or 1h/24h), I think it's a good idea to recommend a precise measurement before the event. This has the advantage of being able to communicate quickly on the results. If the press would be invited at TT attempt, it's a shame not to know the result at the end of the attempt. We could add that this is less important in the case of a 1-hour or 24-hour attempt, but strongly recommended.

It's also important to bear in mind that one measurement of a route is not the truth. If the same measurer measures the same route on the same day 2 times, he will necessarily have 2 different measurements, normally with a small deviation (less than 1 per 1000 in the case of a measurement with the Jones counter).

When measuring a marathon (with a Jones counter, for example, but the principle applies to all methods):
- we don't position the start and finish with a distance of 42.195 km 
- we don't position the start and finish with a distance of 42.195 x 1.001 (safety factor) km
- we position the start and finish at a distance strictly greater than 42.195 km (with a measurement error of minus 1 per 1000 and a safety factor of 1.001).

Comment 15 Sep 07:29

I totally agree that it's a big risk to measure afterwards in a race with a fixed distance and I would recommend everyone to measure beforehand. I think the point Roger and I are making is that in the end it doesn't matter for the recognition of a record whether the measurement is taken before or after. As long as the course actually ridden meets the requirements as officially measured, it would not be logical for me not to recognize the performance as a world record just because the measurement of the course was done afterwards.

If we now explicitly state that the course for time trail records can be measured in retrospect, this of course implicitly creates the impression that this is not permitted for other records. Technically, however, as I said, I see no reason why the course should not be officially measured afterwards for other records as well.

Comment 16 Sep 11:14

I agree with the theory.

But in practice, measuring a course to make sure at it's not too short with a "rough" measurement is almost as complicated as making an official measurement. Unless you make a course that's far too long, where it's impossible to beat a WR.

According to part 1A.5 of the IUF 2019 rulebook : "the host is strongly encouraged to ensure that the competitions are conducted and recorded according to the current IUF World Record Guidelines." [...] "If this is not realistically possible, then the competition results cannot be used for new
world records. In this case, every effort must be made that competitors are made aware of this at least two months in advance."

I interpret this to mean that the official measurement must be taken before the race. If the organizers of a marathon measured 43 km for the marathon, I don't think we can say that they did their best to ensure that the "results were suitable for WR". It's true that the course is "homologable", but the extra distance has an impact on the "quality of the results for WR".

I know that the official measurement of a fixed distance course represents a significant burden (or cost). I'm not sure it's always necessary to make an official measurement, but in this case you'd “only” have to give 2 months' notice that a record can't be beaten.

As for a time trial WR attempt over a fixed distance (100 km or 100 miles), I can't imagine anyone would go for it without measuring first.

Comment 17 Sep 02:14

It may be that the Rulebook requires that the course must be officially measured before the competition - and I have no objection to that, because otherwise an organizer cannot really ensure that the course corresponds to the required distance. But we are not discussing in the Rulebook Committee here!

Perhaps we have a slightly different understanding of what the World Record Guidelines actually represent... For me, they represent the requirements that must be met for a performance to be recognized as a world record. And for the pure recognition of a world record, it doesn't matter in my opinion whether a measurement of the course takes place before or after a race. As long as the measurement confirms that the performance was achieved over the required distance, this performance should be recognized as a world record in my opinion.
Or to put it the other way round, should the World Record Committee really refuse to recognize a performance as a world record if it has been proven by an official measurement after the race that the required distance was covered, simply because the measurement was not carried out beforehand?

Comment 17 Sep 03:30

I agree with Jan on both points.  

Comment 17 Sep 04:04

We all agree on the substance.

Sorry to argue about methods, but for me it's the most important thing. In science, the method always comes before the results. The results mean nothing if the method is wrong.

My discussion/proposal aims to provide a framework to avoid nasty surprises like “Yeah! I beat a WR, but nobody measured the course. What do we do now?”.
A measurement afterwards is in 99% of cases a waste of time or money. If by chance the rough measurement was accurate, there's a risk of not being able to conclude because of the measurement error. Today's course measurement is not an exact science. This type of method is not meant to be used afterwards.

Perhaps a strong recommendation to take the measurement first would be a start. This would be clearer for competitors and for the committee that validates the records.

I think there needs to be consistency between the rulebook and the WR guidelines. WRs are a transversal subject that is dealt with in both documents.

Comment 17 Sep 06:01

Of course, world records are an interdisciplinary topic that is dealt with both in the Rulebook and in the WRG - but for this very reason, I see the WRG as the document that only specifies what the WRC needs to recognize the records and the Rulebook as the document that specifies the rules for competitions and ensures that all conditions of the WRG are met at the competitions. Therefore, in my opinion, the Rulebook is the document that must ensure that no participant in a competition event has any nasty surprises and not the WRG - in my opinion, the proposal is therefore technically something for the Rulebook Committee. The fact that there is currently no section for Time Trails in the Rulebook is a different matter - which in my opinion should definitely be changed.

From my point of view, it is therefore sufficient for the WRG to stipulate that the course must be officially measured and that this measurement protocol must be made available for the recognition of a world record. The fact that this measurement usually has to be carried out in advance of the competition for purely practical reasons is a different issue in my opinion.

Just as an example: at the German Championships Marathon this year, the official course measurement for the world record was only done after the competition - but that doesn't necessarily mean that there was no (accurate) measurement of the course beforehand. There was just no measurement by an official body. To be honest, I see no reason why this procedure should be a problem.

Comment 17 Sep 08:16

I agree with Jan, it's a rulebook committee issue rather than a WR committee issue.  The course should be measured accurately before the race, but a world record is still a world record if a post-race measurement confirms the correct distance.  A post-race measurement is more likely to occur if there is controversy about the accuracy of the pre-race measurement, rather than organisers intentionally doing a rough measurement.

You can do a tweak of Rulebook 3D.1 and 3d.15.2 when the committee is up and running.

Regarding time trial records (eg Hour record and 24hr record)- we specify the inner circumference of a circuit, it is not a shortest possible route measurement.  The actual distance ridden will be greater than the official world record, so it is in the riders best interest to ride as close to the inner circumference as possible. 

 

Comment 18 Oct 03:31

After there were no further comments, I think I would summarize the issue as follows and suggest the next steps:

- The topic of Time Trials should be addressed in the Rulebook Committee and basic rules should be established there
- For the recognition of a world record, it remains crucial that the distance is officially measured, regardless of whether this measurement took place before or after the race

Comment 18 Oct 04:26

I get the impression that you don't know that measuring a course is not an exact science. If it were, I'd agree with you, regardless of whether it's done before or after the race. If the measurement is taken after the race, it's possible that you won't be able to reach a conclusion about a sufficient distance, and that makes the homologation decision more difficult, if not impossible. This has an impact on the WR committee.

Regarding the measurement of the German championship marathon, Jan, you've put me in doubt. Sorry about the delay, but I've just checked and it was indeed measured 2 days before the race (on 9 May for a race on 11 May).

Comment 18 Oct 06:20

To be honest, I don't understand what this has to do with an exact science.

Let's assume a race is to run from the start line to the finish line. The distance for this race must be measured using method X.

Now it is the case that a route can be measured at time t with method X and with this measurement from a starting point A a distance to an end point B is recorded. The result is a route between A and B measured using method X with a corresponding recorded distance (which of course will never correspond to the true distance).
If, in a race, there is at least the distance run between the documented points A and B, it can be assumed on the basis of the measurement with method X that the race has at least the documented distance, right?

So what difference does it make if I
- with known points A and B, organize the race so that it runs at least the distance between the two points, i.e. set the start line at A and the finish line at a point C that is at least at B or further than B away from A
or
- measure a distance from point A that is at least the required distance and thus determine point B (which corresponds exactly to the measurement with method X to obtain points A and B in the first case) and then check whether a given distance AC contains at least the distance AB?

In the second case, I can use the measurement to prove that the distance AC has the required distance in exactly the same way as when defining the distance AC on the basis of known points A and B.
Of course, I will achieve the minimum difference between the distances AB and AC if AB is known and I can therefore determine AC. Nevertheless, I can always check whether a distance AC contains the distance AB and thus a required minimum length, can't I?

Measuring distances for races is not about determining the true value of a length between two points, it is only about making a statement as to whether a distance has at least a certain length, right?

Comment 24 Oct 10:37

> Measuring distances for races is not about determining the true value of a length between two points, it is only about making a statement as to whether a distance has at least a certain length, right?

Exactly! We measure a distance that can't be less than the official distance. We still have additional information. For example, when you measure a marathon with a Jones counter, the distance actually measured is included in the 100% confidence interval between 42.195 and 42.279 km.

This means that the official/precise measurement taken can position point A and B at 42.279 km (maximum).

This method has a measurement error of strictly less than 1 in 1000 (the true measurement error is certainly more in the range of 1 in 2000), hence the use of a correction factor of 1 in 1000, which gives us 100% certainty that we'll obtain a distance that's greater than the official distance.

In itself, it's not a big deal to cover an extra 84m during the race and records are rarely set by a couple of seconds.

However, I have more doubts about the rough measurement used. 

As far as I'm concerned, there are 3 situations relating to this rough measurement:

1/ the rough measurement method is similar to the official measurement => no problem, instead of a correction factor of 1.001, you should use a correction factor of 1.003 to be 100% sure that the confidence intervals do not overlap. But the principle of being able to make a rough measurement is not to make 2 precise measurements.

2/ the rough measurement method has an unknown measurement error => what distance should be measured so that the confidence interval does not overlap with the confidence interval of the official measurement?

3/ the rough measurement method has a very high margin of error: for example <1% with a correction factor of 1% (a method which would be 5 times more precise than the current method described in the rulebook) => it is necessary measure a distance of approximately 42.7 km to be 100% sure that the confidence intervals do not overlap (CI100% 42.273;43,127)

Comment 28 Oct 06:42

> If 2 measurements are taken, to be 100% sure of being able to conclude, the 100% confidence intervals must not overlap.

I see your point - if the result of the measurement is that the distance AB determined with measurement method X is longer than the distance of the race AC, this is of course no proof that the distance AC was really too short, as the distance AB determined could also be at the upper end of the confidence interval.
In the other direction, however, if the distance AB determined using measurement method X is shorter than the distance of the race AC, it can be concluded with certainty that the distance AC was long enough.

 

Overall, I would say that a measurement after the race increases the probability that a course that is actually long enough will be found to be insufficient - nevertheless, this fact does not change my basic opinion that it should not be decisive for the recognition of a WR when the course was measured. As long as the measurement proves that the distance ridden was long enough in any case, the performance should be recognized as a WR in my opinion.
As I said, I do not deny that a retrospective measurement represents a significantly increased risk that the distance may not be deemed sufficient. But in my opinion, this is a different issue than the question “Can we accept a retrospective measurement or not?”
I don't want to encourage organizers not to do the official measurement in advance - a race over a fixed distance is not possible without a measurement. I am only concerned with the question of whether a performance can be recognized as a WR if the course was only officially measured afterwards.

 

Once again: I don't want organizers to measure their course only roughly! I am more concerned with the aspect of “official measurement”. Of your examples, only the first case would really be relevant for me, as I think it is conceivable that a course could be measured in advance, but not by an official measurer. In this case, the increased corrector factor could ensure that an official measurement can take place afterwards.