3D.12 Finishes
Comments about this discussion:
Started
The current section 3D.12 concerns timing only. From the discussions we had in the WR committee, it seems that the use of the photofinish mentioned in this section is not necessary for the validation of a WR rounded up to the next second. The photofinish may be important for the classification of a race, especially at UNICON, but it should not be mandatory. I would be more of the opinion that if 2 competitors are less than 0.1 seconds apart and there is no photofinish, the competitors should be ranked equally.
Another important element seems to me to be the introduction of safety rules to define the finish area of a confrontation race (where there may be a sprint: wave start and mass start). I suggest adding that the finish should not be downhill, and that turns should be avoided in the last few hundred meters.
Downhill finishes more often result in falls (the finish of the UNICON20 marathon was on a downhill false flat and resulted in several falls, 1 of which was quite serious). As far as bends are concerned, there is often 1 bend close to the finish line, and when this is the case, the order of finish is usually decided before this bend.
Comment
I agree with you that a photofinish should not be mandatory - especially as it has rarely or never been used in practice.
I also think the other aspects mentioned are important. Definitely a good discussion.
We should also consider whether it would make sense to merge sections 3B.5.8 Finishes and 3D.12 Finishes to have everything in one place and not fragment the rules too much.
Comment
0.1 sec apart is equal to 111cm with 40km/h speed.
Acurracy of time measurement with passive chips is 0.1 sec, and with active chips 0.003 sec.
0.1 sec would not be a good separation for ranking. Riders could draft all the way and just get little closer to be ranked the same position.
Having few riders ranked equally may pose also a problem with awards.
I would keep fotofinish required at Unicon.
For the ease of judging I may consider fotofonish only for awarded competitors.
Fotofinish accuracy should be specified. 120 fps is not difficult to reach with amateur cameras and some smartphones, and with 40km/h gives 9 cm resolution.
Regarding the design of the finish area, I would include it in the rule for the design and safety of the racing course.
Comment
I'm convinced that the photofinish is useful for UNICON, but for other competitions, if there's no photofinish, I don't understand what you're proposing. Are you suggesting filming the finish with a smartphone?
Active chips are very accurate, but if the transponder is attached to the ankle or shoe, there can be a 30cm difference depending on the position of the feet on the finish line. In this case, the accuracy is more like 0.03 sec than 0.003 sec.
I find it hard to imagine competitors sabotaging the race by deliberately finishing within the same tenth of a second. It's not something that happens often. We could also say that if the gap is less than 0.1 seconds and there is no photo finish, the decision rests with a finish judge. And if he is unable to break the tie, then it's a tie...
Comment
Since we requiere photo finish, we should require minimum requirements for the camera. It is useless to have photo finish with 30 fps.
Human eye can perceive the image with 30-60 fps.
Suggested by me, 120 fps at least doubles the ability of human eye and is not difficult to reach without professional photo finish cameras (3000-20000 fps). Means, it is easy and cheap for the organizer.
I agree that neither active chips are good solution, because the rule states the front of the wheel crossing the finish line.
Photo finish must be required at Unicons and at least visual judging on other events.
Yes, if the judge cannot break the tie, its the tie. I imagine it being very unlikely.
3D.16 Accuracy of results already covers situations with ties, with or without foto finish.
Comment
I think that yes, a video of the finish line is a very appropriate way to determine finish order without professional photo finish. It would have to be set up in line with the finish to clearly capture the tire breaking the plane. I don't think issuing ties as official results make sense, and participants will not be satisfied with a tie being issued because we don't want to try to be accurate at the finish line. In practice we have had line judges to make the call and video for review of difficult calls.
I wouldn't say that human eyes "resolve" at any sort of reasonable framerate, their performance in tracking moving objects is much higher than a framerate measure would suggest simply because they work differently. I think a tie could be awarded if it truly cannot be resolved otherwise, but should be avoided when it is practical to capture video of the top finishers and review results for accuracy.
Comment
I'm no expert on photofinish devices. We discussed this subject in the WR committee and concluded that it was necessary for the classification of road races at UNICON.
But there are local or national races where there is no photofinish, and I don't think it's necessary to impose the use of a photofinish or video device for all races (potential IUF endorsed event?).
For example, at the last French nationals, the timekeeper was on the start line to validate the start times of the different waves, but he couldn't be at the finish to monitor a photofinish device. A 2nd timekeeper would have had to be paid, and it would have cost around twice as much to use a photofinish system.
I think the rulebook should provide tiebreaker solutions or not in the absence of a video system.
Official race and WR times are rounded up to the nearest second, so the use of a photofinish or other video device is only useful for tie-breaking.
> Regarding the design of the finish area, I would include it in the rule for the design and safety of the racing course.
It can be integrated into this other section, but in that case it must be made clear that these elements only concern races with waves start or mass start. For a race with an individual start or for a Criterium, these elements do not apply.
Comment
I think it is important that we differentiate in this discussion whether a photofinish system should be mandatory or simply a video recording of the finish. A photofinish system is not a normal video system!
Especially considering the inaccuracy of distance measurement in road races, I see no need for a photofinish system in a road race, its purpose would only be tie-breaking, as Simon has already written. Such a system is of course perfect for this - but in practice I don't think it's necessary for the vast majority of races.
If we want to prescribe a tie-breaking system in the rules, then this should be made clear that it is a tie-breaking system only and, in my opinion, we should definitely list alternatives to a photofinish system.
Comment
So i think in principle we need to have some form of video at the finish line. There are many parts of the course that may be out of our control as far as having resources to review, but a video of the finish to be retained until the protest period has ended seems the least we ought to do for anything that is IUF sanctioned.
A photo finish system at Unicon where the world champion title is on the line doesn’t seem overly onerous to me either but there is a clear increase in cost present there.
Comment
I think we should perhaps first answer the question of exactly what purpose we want to prescribe a photofinish system for?
As I said, I personally only see the purpose as a tie-breaker and I wonder whether such a system is really necessary. How much effort is it really worth to break a tie? At what point are two performances even worthy of differentiation (for example, one could take into account slightly different starting positions, a slightly different route resulting in different distances covered, external influences, etc.)?
If we look at other sports, e.g. athletics, where there are also fixed distance races in the form of road races, in which world records are also recognized and which, in my view, are very similar to our fixed distance races, then there is no need for a photofinish system there, for example.
Comment
Hmm, I see breaking a tie in exactly the opposite way. The only truly comparable performance between two athletes is their performance on a given day across the same course. The conditions are as similar as possible for everyone, and so a win or loss comes down to fitness and racecraft. Maybe top athletes see it differently, but I’d personally be very disappointed if an effort wasn’t made to say who finished ahead if I were involved in a very close finish for a podium. That’s the very nature of shoulder-to-shoulder confrontational races.
Comment
> The only truly comparable performance between two athletes is their performance on a given day across the same course.
I would agree with this in principle - however, I personally believe that this is not feasible up to any accuracy. For example, if someone would start two meters behind me at the start, but I would only be 1/1000 second ahead of this athlete at the finish, then the tie breaking would have a very unpleasant flavor for me. I would definitely prefer to be scored equally in this case, because I don't believe that the performances in this case can really be differentiated with such a high degree of accuracy.
But I can also understand if there are people who prefer to separate the rankings within a race with the highest possible accuracy. In this case, of course, a correspondingly high-resolution tie-breaking system is required.
Comment
How do the others see this issue? Should the rules continue to insist that a photo finish system must be used for tie-breaking purposes?
Comment
trying to follow the conversation quick, I'm not sure what we're discussing anymore.
I think photo finish for Unicon should still be required - plus almost all professional timing systems that are hired for Unicons have them.
But for small events, if it comes at a substantial cost to the organizer, could be encouraged by the rulebook, but not required to be IUF compliant.
Comment
I can try to briefly summarize the core of the discussion:
Since a performance in road races consists not only of the time component, but also of the distance covered and the distance measurement always gives the greater uncertainty, it makes no sense (especially in fixed distance races, where the focus is on the comparability of performances between different events) to indicate the times of the riders on the official results lists with an accuracy greater than one second. For this reason, the rules currently stipulate that measured times must be rounded to full seconds and world records are also kept with this accuracy.
A photo finish system of course offers a significantly higher time resolution than the one second that is relevant for the results. This means that a photo finish system is currently only really useful/neccessary for tie breaking. The question is therefore whether such a system is really required for the purpose of tie breaking in the rules.
Comment
>>The question is therefore whether such a system is really required for the purpose of tie breaking in the rules.
Having been the awards director at 4 Unicons (16, 17, 19, 20) and many NAUCCs, I can tell you it is the most irritating thing to hear from the race director that theres a tie. Awards are not purchased with ties in mind. Yes there is a buffer in the number of medals/awards ordered, but without rules to help break tie breakers, ordering awards becomes and even more complicated process than it already is.
For this reason - I think it should still be required for breaking ties at Unicon. Smaller events, like NAUCC and the like is should not be required.
Comment
Yes, from that perspective it makes perfect sense to have as few ties as possible, I agree.
But for me it sounds like a sensible compromise to only really require this technique for Unicons in the rulebook and leave it up to the organizers of other competitions to decide whether they want to use a photo finish system as a tiebreaker.
Comment
To move this discussion forward, I am making an unofficial suggestion to revise section 3D.12 Finishes. In my opinion, the section deals less with the actual finish and more with the timing, which is why I would rename it:
3D.12 Timing, Photo Finish
1. The use of Transponder Timing Systems or Fully Automatic Timing and Photo Finish Systems that fulfill the specified criteria are permitted for timekeeping in road races.
2. The use of Transponder Timing Systems is permitted, provided that:
i) None of the equipment used at the start, on the course or at the finish line is a significant obstacle to the progress of a rider.
ii) The weight of the transponder and its housing carried or worn by the rider is not significant.
iii) The System is started by the start signal.
iv) The System requires no action by a rider during the competition, at the finish or at any stage in the result processing.
3. The use of Fully Automatic Timing and Photo Finish Systems, provided they meet the criteria set out in Rule 2D.8 and are operated accordingly.
4. For all road races at Unicon, a Photo Finish System according to paragraph 3. must be used to determine the placings. The placing order must be determined at the finishing moment in accordance with rule 3B.5.8.
I would not copy the requirements for a photo finish system from the track chapter, but simply refer to it. If the majority is in favor of copying and pasting the rules here, I would be fine with that.
What do the rest of you think about the permissibility of hand-stopped times in road races? This is certainly not an option for the Unicon, but for very small events with few participants it could be an alternative under certain circumstances... Hand-stopped times are also permitted for world records, as long as at least two timekeepers have stopped the rider.
Comment
Any comments on that?
Comment
It would be nice if there would be some feedback on this - especially on whether we should include something on hand-stopped times in the rule.
Comment
I also think adding something regard the nature of the course finish would be useful.
For example, USA cycling rule:
"3A4. The finishing area should be at least eight meters wide and be adequately
protected so as to prevent spectators from entering onto the course. The last 200
meters should be free of turns and curves."
I also believe downhill finishes should be discouraged. But, this may be hard to create an objective rule, since I think it would depend on the rest of the course. If the majority of course is very flat, then the finish should be similarly flat. If there are plenty of uphill/downhills, then a similar downhill at the finish probably would be not as much of a concern.
I think this should be a separate section from "3D.12 Timing, Photo Finish"
For tie breakers at smaller events:
I think a video, even at 30 fps (33 milliseconds per frame), would be useful for tie-breakers. This is only 37 cm resolution at 40 km/h, but likely it is more useful than just a 37 cm gap, since you can estimate the speed of the rider and who was in front using the few frames before the riders cross the finish line.
I'd be hesitant to say two riders who are 100 milliseconds apart should get the same result. But less 33 milliseconds apart, with video evidence, would satisfy me if I were a competitor.
Chip timing:
Regardless of the accuracy of chip timing, I personally dislike chip timing because of the timing mat. A 1-3cm bump might not be a big deal for most riders, but I almost always slow down beforehand because I get scared.
I might be interpreting this incorrectly, but there is something similarly noted in a UCI document:
"g) Detection loop
This must comprise cables traversing the road. Mats or other
devices over 5 mm thick are prohibited for safety reasons."
Hand stopped timing:
I think hand-stopped times in road races is fine given precautions/redundancies, like having two timers. Like mentioned, if the course distance inaccuracy/resolution can easily be more than 10 meters, then half a second or two second error from a hand-stopped timing should not be a big issue. I personally prefer hand stopped timing over chip timing given my dislike of chip timing.
Comment
> I also think adding something regard the nature of the course finish would be useful.
In my opinion, this would be something that belongs in the “3D.1 Venue” section - there are already discussions about this here in the Committee, so you are very welcome to contribute your ideas and thoughts there.
> I think a video, even at 30 fps (33 milliseconds per frame), would be useful for tie-breakers.
Yes, a normal video can certainly also be used for tie-breaking, the resolution is of course significantly reduced compared to a photo-finish system. That was also my question above, whether we really want to prescribe a photofinish system only for the purpose of tie-breaking, or whether alternatives might also be an option.
I would definitely be open to giving organizers the option of using a normal video at the finish line for tie-breaking purposes. However, I would not make this mandatory for all events, as it makes the evaluation much more time-consuming, for example, as the video has to be evaluated manually and the placings have to be entered manually into the results system.
So I would be open to it as an option, but I would be against it being mandatory for all competitions.
> I might be interpreting this incorrectly, but there is something similarly noted in a UCI document:
The proposed rule also contains a sentence on this:
i) None of the equipment used at the start, on the course or at the finish line is a significant obstacle to the progress of a rider.
Comment
I agree with Jan that a separate Timing section should be created. I suggest integrating the current section "3D.16 Accuracy of Results" as a subsection. The official time is closely linked to the timing method.
I'm wondering about adding a specific case for the official time of a timed event, as a tiebreaker can't be achieved with a photo finish; only the time would break a tie. I'm not sure that rounding to the nearest second is always desirable as an official time if 1/we want to break potential ties and 2/we have the means to obtain a time with greater accuracy.
Comment
The current rule “3D.16 Accuracy of Results” is more about how to publish results - but you are right that this is linked to the timing itself.
> I'm wondering about adding a specific case for the official time of a timed event, as a tiebreaker can't be achieved with a photo finish; only the time would break a tie.
Basically, this is exactly what is done in a tie-breaking with a photo-finish system - you look at the unrounded time and use the time difference as a tie-breaker. However, a photo-finish system is the only system that actually records the time difference between the two times at which two different competitors finished the race (as this is the time at which the front of the wheel crosses the finish line). This time cannot be determined with the times from a transponder timing system and therefore time differences from the transponder measurement can never represent the actual difference between two finishes. Therefore, unrounded transponder times do not necessarily have to correspond to the actual order of finishing.
I think it would be very difficult for outsiders to understand why the times are rounded on the one hand, but the same times are used for the ranking on the other. I think it would therefore be more expedient to either use a system explicitly designed for tie-breaking or to leave ties as they are.
Comment
I'm not sure I understand your answer, especially since I made a mistake in my sentence. I meant the tiebreaker in a time trial. Official times can be rounded to the next second, but I suspect that depending on the timing method, the ranking can be determined more accurately or not.
Comment
If a photo finish system is used, then the official order can of course be determined very precisely, as the exact time of the finish can also be determined very accurately. This is difficult with other methods, as the exact time of the finish (front edge of the tire) cannot be determined when measuring with timing chips for example.
Comment
I don't mind having ties in the official time rounded to the second in a time trial, as it's necessarily rarer than in a race with a mass start.
If, during a UNICON, we have the means to install a photofinish connected to a central clock, we could break the tie using this method, but it seems complex to implement.
In cycling time trials, active ships and mats are often used at the start and finish. I believe this method is relatively simple and is suitable for most cases.
Comment
I'm not sure what the discussion at the end means for the proposed rule - should we adjust the tie-breaking paragraph again, or can we leave it at what's already covered in the proposal?
If we want to include something about hand-stopped times, then I would suggest the following:
Add to paragraph 1:
1. The use of Transponder Timing Systems, Fully Automatic Timing and Photo Finish Systems or Hand Timing that fulfill the specified criteria are permitted for timekeeping in road races.
Insert a paragraph for hand timing:
2. The use of Hand Timing is permitted, provided that:
i) Timekeepers are in line with the finish and all have a good view of the finish line.
ii) Timekeepers use manually operated electronic timers with digital readouts. All such timing devices are termed “watches” for the purpose of the Rules.
iii) The time is taken from the start signal.
iv) Two official Timekeepers time the winner of every event and any performance for record purposes.
v) Each Timekeeper acts independently and without showing their watch to, or discussing their time with, any other person.
vi) The time measurement must be carried out with a resolution of at least 1/10 of a second, unless the measurement system ensures that the times are always given to at least the next longer full second. Unless the time is an exact full second, the time will be converted to the next longer full second. If the measuring system only displays full seconds and it cannot be ensured that this is the next longer full second, one second must be added to the displayed time. If two watches are prescribed and, after converting as indicated above and two watches are prescribed, the two watches disagree, the longer time will be official.