Measuring Breakaway Gaps

[fujitourer]

Moto 1 follows breakaway. Every 5 seconds the GPS pings the location and time.
Moto 2 follows peloton, Measure time difference between the two groups on every measuring point created by moto 1.

Everyone should ignore everything written in this thread, except for this. It is really quite simple and I am astounded by the lengths some have gone to over (or maybe under) think this issue, only to confuse themselves entirely.

 

[El Pistolero]

Everyone should ignore everything written in this thread, except for this. It is really quite simple and I am astounded by the lengths some have gone to overt (or maybe under) think this issue, only to confuse themselves entirely.

I've been saying pretty much the same all along

 

[Waterloo Sunrise]

I've been saying pretty much the same all along

You made 1 previous post, which was straight forwardly incorrect.

The time gap measured is inversely proportional to the speed of the peloton and can lead to perverse results when interpretted with insufficient care. Some people have grasped that, and others have not, but i'm really past caring now Cav has been CRUELLY ROBBED by Purito.

 

[sagard]

Apologies, I should have said peloton speed.

The reason I brought this up is the frequent site of someone saying the gap has tumbled in a sprint stage when the peloton wakes up and shifts from 35kph to 45kph. You get instantaneous drops in gap and many people don't seem to have realised that it is largely a figment of the measurement method rather than a genuine phenomena.

No. If the break is 10k up the road with 10k remaining, having raced the previous 10k at 30kph. For our example they are 20 minutes up. If the speed increases for both groups to 60kph, the chase group will make up a bunch of time because the will have doubled the speed of the break over that same 10k.

Speed = distance / time

For the break on the second to last 10k. Time = distance / speed = 10k / 30kph = 20 min
For the chase on the second to last 10k. Time = distance / speed = 10k / 60kph = 10 min

Each group then takes 10 minutes for the final 10k due to riding the same speed. Break ends up winning, but by only 10 minutes.

 

[Waterloo Sunrise]

No. If the break is 10k up the road with 10k remaining, having raced the previous 10k at 30kph. For our example they are 20 minutes up. If the speed increases for both groups to 60kph, the chase group will make up a bunch of time because the will have doubled the speed of the break over that same 10k.

Speed = distance / time

For the break on the second to last 10k. Time = distance / speed = 10k / 30kph = 20 min
For the chase on the second to last 10k. Time = distance / speed = 10k / 60kph = 10 min

Each group then takes 10 minutes for the final 10k due to riding the same speed. Break ends up winning, but by only 10 minutes.

Yes, I am aware that 2+2=4 of course when you multiply it out the numbers net out.

The point is that in such a situation where both groups double their pace, the time gap halves, and lots of people watching get excited, but in reality they are no more likely to catch the group.

I was simply demonstrating how measuring by distance (which requires you to take parcours in to account) and measuring by time (where it is subject to significant swings if the peloton varies its pace) both have their weak-points.

 

[Libertine Seguros]

Yes, I am aware that 2+2=4 of course when you multiply it out the numbers net out.

The point is that in such a situation where both groups double their pace, the time gap halves, and lots of people watching get excited, but in reality they are no more likely to catch the group.

I was simply demonstrating how measuring by distance (which requires you to take parcours in to account) and measuring by time (where it is subject to significant swings if the peloton varies its pace) both have their weak-points.

And given both have their weak points, and that the race winner will be judged according to the time taken, it makes no sense to change the system to an equally flawed one that makes working out the state of the race in a stage race more complicated and confusing.

 

[Caruut]

In real racing situations, the swings are not significant, though.

Take a situation where you have the two groups, group 1 (g1) ahead of group 2 (g2), riding at a constant speed u. At a distance D1, the time gap is of size T1. When g2 passes D1, both groups increase their speed to a constant speed v. Let D2 be the position of g1 when g2 is at D1. Define r=D2-D1.

Then the second split is going to be equal to the first split minus the difference in times, that is;

T2=T1-(time for g1 to travel between D1 and D2)+(time for g2 to travel between D1 and D2)

Because g1 rode at a speed u between the points and g2 rode at speed v, their times to ride between D1 and D2 are r/u and r/v respectively.

So, we see that;
T2=T1-r/u+r/v
T2=T1-(rv/uv-ru/uv) (note that r/v=ru/uv and r/u=rv/uv)
T2=T1-r*(v-u)/uv

Now if we say that a "dramatic" reduction is the time gap halving, then T2=0.5*T1, so;

0.5*T1=T1-r*(v-u)/uv
2r*(v-u)/uv=T1

Now since this is a theoretically flat stage, let's go into the real world and say that u=40km/h, and the sprinters' teams are looking for the stage, so have kept the break in check all day, with the gap at about 6 minutes (or 0.1h). Then the break is 4km up the road, to r=4km. So, plugging it all in,

2*4km*(v-40km/h)/(v*40km/h)=0.1h
8km*(v-40km/h)=v*40km/h*0.1h
8km*v-320kmkm/h=4km*v (cancel one km term)
8v-320km/h=4v
4v=320km/h
v=80km/h

There you have it - in a real world situation, the half the time gap that much that quickly, the speed has to hit 80km/h.

Sorry about all the maths.

 

[Libertine Seguros]

Forget the mathematics.

Just think of it this way.

If the TV was reporting that Thomas de Gendt was 1200m ahead of the maglia rosa group earlier today, we wouldn't know how to interpret it. We would need to know the relative speeds of him and the group, then calculate the gap in time, because ultimately a cycling race is scored on time. By the time we've done that, the gap might have changed.

As the TV said that Thomas de Gendt was 5'20" ahead of the maglia rosa group, we knew instantly that meant he was 19" off the maglia rosa.

Cycling races aren't scored on distance, because everybody completes the same distance if they want to be a classified finisher. They're scored on time, which is why we are told in time.

In a one-day race, perhaps distance between groups may have some value as a metric. In a stage race? Not so much.

 

[Waterloo Sunrise]

In real racing situations, the swings are not significant, though.

Take a situation where you have the two groups, group 1 (g1) ahead of group 2 (g2), riding at a constant speed u. At a distance D1, the time gap is of size T1. When g2 passes D1, both groups increase their speed to a constant speed v. Let D2 be the position of g1 when g2 is at D1. Define r=D2-D1.

Then the second split is going to be equal to the first split minus the difference in times, that is;

T2=T1-(time for g1 to travel between D1 and D2)+(time for g2 to travel between D1 and D2)

Because g1 rode at a speed u between the points and g2 rode at speed v, their times to ride between D1 and D2 are r/u and r/v respectively.

So, we see that;
T2=T1-r/u+r/v
T2=T1-(rv/uv-ru/uv) (note that r/v=ru/uv and r/u=rv/uv)
T2=T1-r*(v-u)/uv

Now if we say that a "dramatic" reduction is the time gap halving, then T2=0.5*T1, so;

0.5*T1=T1-r*(v-u)/uv
2r*(v-u)/uv=T1

Now since this is a theoretically flat stage, let's go into the real world and say that u=40km/h, and the sprinters' teams are looking for the stage, so have kept the break in check all day, with the gap at about 6 minutes (or 0.1h). Then the break is 4km up the road, to r=4km. So, plugging it all in,

2*4km*(v-40km/h)/(v*40km/h)=0.1h
8km*(v-40km/h)=v*40km/h*0.1h
8km*v-320kmkm/h=4km*v (cancel one km term)
8v-320km/h=4v
4v=320km/h
v=80km/h

There you have it - in a real world situation, the half the time gap that much that quickly, the speed has to hit 80km/h.

Sorry about all the maths.

I appreciate the effort, but you have simply just demonstrated that to halve the gap you need to double the speed, like I said all along.

My basic point is just trying to get people to recognise that both distance and time as a way of measuring gaps can be perverse.

It is quite easy to create scenarios in both where the situation gets worse for the breakaway, but the gap increases, or vice versa.

In the time measurement scenario, if both groups ride at 20kph, then the peloton increases to 40kph, but the break increases to 45kph, the breakaway is then pulling away and has a better chance, but the time gap will fall.

In the distance scenario, clearly if a break has already peaked a climb it will distort the measurement.

Time gaps are clearly vastly more practical - I just wanted to get a discussion going so people could see there are issues with both. Using time gaps, the perverse situations tend to be subtle, and the information required to spot them is typically not available, so I would suggest it is more insideous than using distance, which wears its flaws on its sleave for all to see. But time is more practical and will always be used - no debate or complaint on that score.

 

[Buffalo Soldier]

Time gaps can be perverse only because you interpret them the wrong way, waterloo

 

[Waterloo Sunrise]

Time gaps can be perverse only because you interpret them the wrong way, waterloo

They are put on the screen to give a measurement of how the breakaway is doing.

If the gap goes down, most people quite reasonably assume that means they are getting caught.

As we have now all agreed, that actually is not the case. The gap can go down even when the break is pulling away.

I'm sure you were aware of this all along, it shines through in your posts.

 

[Buffalo Soldier]

No we do not all agree. If time gaps go down, the chasing group is getting closer to the leaders. Not in distance (which is irrelevant), but in time.

 

[Waterloo Sunrise]

No we do not all agree. If time gaps go down, the chasing group is getting closer to the leaders. Not in distance (which is irrelevant), but in time.

So when everyone speeds up, but the breakaway speeds up more, and is going faster than the peloton, they are getting caught.

Got it.

 

[Buffalo Soldier]

They will be caught when the time gap is 0.

If the chasers speed up, they will get closer to the leaders. If the leaders speed up to, they can avoid most of the damage, yes, because they started gaining speed earlier on the course.

 

[Waterloo Sunrise]

I'm just saying that most people assume a falling time gap means people are getting caught.

We all now agree (you disagree on how to interpret it) that it is possible to have a falling time gap even whilst the breakaway are riding away from the peloton on the flat.

If you think everyone already understands the factors at play, fine, but I see enough posts from people getting overexcited about moving time gaps because they don't get these points that I thought it worth discussing.

 

[Lanark]

I'm just saying that most people assume a falling time gap means people are getting caught.

We all now agree (you disagree on how to interpret it) that it is possible to have a falling time gap even whilst the breakaway are riding away from the peloton on the flat.

If you think everyone already understands the factors at play, fine, but I see enough posts from people getting overexcited about moving time gaps because they don't get these points that I thought it worth discussing.

Or they get them, but the understand that in real race situations a time gap getting closer does mean the breakaway is more likely getting caught in 99 out of a 100 times.

 

[Waterloo Sunrise]

Or they get them, but the understand that in real race situations a time gap getting closer does mean the breakaway is more likely getting caught in 99 out of a 100 times.

I was primarily thinking of the moment every breakaway stage when the gap falls from 9min to 6min inside 30 seconds because the peloton decides to wake up, and someone gets shocked about how it is possible.

Or earlier today, when the gap went from 3mins to 5mins for Zaugg inside 2 minutes because the favourite group started freewheeling.


And frankly, several perfectly bright posters tried to claim that the time gap would stay the same when the speed increased in this thread, so it's pretty clear lots of people don't get it.

 

[Buffalo Soldier]

But still, the time gaps ARE a realistic measurement on these (on all) situations!

(And of course the gap couldn't go from 3 to 5 mins inside 2 minutes, unless the peloton just stopped riding , but you know that)

 

[Waterloo Sunrise]

But still, the time gaps ARE a realistic measurement on these (on all) situations!

Yes.

The time gap is a realistic measurement of the time gap in all situations.

The distance between them is also a realistic measurement of the distance between them in all situations.

Both of them are incomplete if you are relying on them to tell if the breakaway is getting caught with certainty. And telling that is what most people use the time gap for.

 

[roundabout]

I don't know. Seems some kind of issue with GPS in the Giro. Can't remember any other races where the real time gaps have been fluctuating so wildly. Don't think it's an issue with gaps being measured by time in general.

 

[icefire]

The problem with the GPS system in the giro seems to be a concept problem for measuring time gaps.

From what I heard the commentators say, they are taking the position of the race leaders and the chasers, compute the distance and estimate how much time will take the chasers to get to the current position of the race leaders. This way of estimating time gaps is very sensitive to the instantaneous velocity of the chasers (read crap).

It would be much simpler to just measure how much time ago the race leaders were at the point the chasers are now. GPS engineering can't be any simpler and more accurate than that.

 

[cineteq]

As the TV said that Thomas de Gendt was 5'20" ahead of the maglia rosa group, we knew instantly that meant he was 19" off the maglia rosa.

Cycling races aren't scored on distance, because everybody completes the same distance if they want to be a classified finisher. They're scored on time, which is why we are told in time.

Very good, this should be enough argument kill further talks.

 

[Master50]

First
Time is what determines place. We use a stopwatch as the tool to measure gaps. To get distance is not so hard but it means nothing and it requires accurate odometers.
Gps Units are not on the rider but in a car or on a motorcycle because they are transmitters and heavy unlike the gps receivers on your bicycle. They also suck a lot of power too. So wildly fluctuating gaps usually indicate a vehicle has changed its place in the race. Maybe the gps was in the chief comm car (who is generally behing the main peloton) and he moves up to the next group suddenly it looks like the group he was following accelerated some incredible amount to close a gap.
What actually happens in big races like the tour or California is the gps stuff is for instant gratification of the tv audience but the truly accurate gaps are delayed until the second group reaches the same point but to do this you have to monitor a single point on the road. The times for radio tour are often done by 1 or 2 guys by radio. I see a point on the road and call to the second bike or Commissaire the landmark and broadcast a start signal 3 2 1 start on the crosswalk at pine and main as an example. The second bike gets there and stops his watch to broadcast the time back to the other Commissaire or radio tour. This does not work well for TV but they are really taking gps positions and estimating the time based on speed and distance. It works pretty well for broadcaster but it is really showing the distance and speed of two vehicles and not the riders. So we have seen gps broadcast transmitters on bikes (now against the rules) except they have a very limited range to be acceptable for a rider to carry. This means there has to be a close by receiver too again inside a vehicle.
It is all about the watch as the simplest and easiest tool to measure a gap. Time is also very important for allowing cars into gaps as we all know a car is a great draft and cars can only be in 1 minute gaps or more.
Besides you can always get distant if you have your speed over time too.
Time works best. It is the simplest and most reliable tool we use but the technology is changing so maybe at some point the airplane for the radio repeater will carry a gps computer tied to a device on each rider and maybe you can have your time distance and speed in some higher tech world.

 

[Caruut]

I don't know. Seems some kind of issue with GPS in the Giro. Can't remember any other races where the real time gaps have been fluctuating so wildly. Don't think it's an issue with gaps being measured by time in general.

LBL - Nibali was allegedly way ahead for ages, then suddenly Iglinsky is on him, and the race officials updated the gap. Possibly an issue with manually updating which moto/car is actually giving the time and everyone getting so excited they didn't realise they hadn't bothered to get readings from Iglinsky's moto.

 

[Caruut]

I was primarily thinking of the moment every breakaway stage when the gap falls from 9min to 6min inside 30 seconds because the peloton decides to wake up, and someone gets shocked about how it is possible.

Or earlier today, when the gap went from 3mins to 5mins for Zaugg inside 2 minutes because the favourite group started freewheeling.


And frankly, several perfectly bright posters tried to claim that the time gap would stay the same when the speed increased in this thread, so it's pretty clear lots of people don't get it.

The real time gap cannot fall from 9 minutes to 6 minutes inside 30 seconds. That is a practical error in measuring it and displaying it on our screens, not a theoretical deficiency. Perhaps it is a lazy guy in the control room, not making sure that the moto he is using the signal from is the moto with the group. Perhaps they switched from a vehicle in behind the peloton to a vehicle in front of it. Maybe they haven't got a signal from them in a while - it is, after all, mountains. Whatever it is, it is not a fundamental error in the concept of time gaps.

People only claimed that it would stay the same until you released the rather perplexing scenario in which they both speed up at the same time .You seem to be the only one who thinks in this situation that the time gap shouldn't decrease. The chasing group has raced a given section faster, so the time gap decreases. That's really not particularly revolutionary.