RR alone can’t solve drafting problems at IM races. IM would have to make other changes too, like altering start procedures and bike courses. Courses that are more than one loop can be very challenging. Even one loop courses can be tough. I did IMFL and there was a long section pretty early in the race with almost no shoulder and lots of cars-- a very tough situation for any strong cyclist who’s a marginal swimmer. Stay completely out of the drafting zone and you don’t have time to pass before the next monster truck comes along.
Just about every IM race I’ve done I’ve seen only a handful of truly blatant drafters. These are the guys who are sitting inches off someone’s wheel. I see many others who are technically drafting, many of whom seem to be doing it because it’s hard to avoid given course conditions. I really only care about the blatant drafters-- the ones that make me pray an official is sitting silently behind us. If RR can punish those guys, I’d be satisfied.
With the financial focus of IM, I am thinking this will not be widespread in implementation. There are just so many factors that go into AG drafting that they would literally just be collecting data and reviewing disputes. And who among us wouldn’t dispute a penalty?
I would guess that I’m forced to draft approximately 5 mins in a 70.3 just for safety purposes. Add on that the 3-4x wide groups, people sitting on sideways, etc. And RR would have an incredible task of figuring out whats going on. RR is made for use in a pace line of 50-60 people, and IM could only possibility use this for a full 140.6 with few participants and few loops.
Do they have gps data on them? You could say, ignore drafting in the first five to ten minutes of movement, but how do you know when the end is let alone a steep grade?
Someone has to program all this stuff in and let’s not assume there’s a Google sized team of developers working on every aspect of it.
There really are a lot of challenging to pulling all of this together and they only get a handful of times to field test it, reprogram them all and the try again.
Yes it is very doable from a technical perspective. The units do have GPS on them. If the steep grade comes at mile 10 of a race, then they can just disable them. I promise you, it’s all doable.
This gets back to who is doing this stuff. Not a huge team of engineers at google. You want them to have the architecture to support preprogrammed gps routes in these things, and imagined they have the communication and storage protocol to facilitate downloading that data, and using it in real time while on the road.
Yes, we all know it’s “doable” in the sense that cell phones exist.
But all the comments on this product seem to imagine a sophisticated operating system primed to do every task imagined for it. My impression is they are writing the code to do tasks from scratch and likely have all kinds of processor, storage, and power limitations.
Adding to that: It’s going to be an imperfect product. There’s no way to output 1000 devices that accurately measure and record range to each other for hours at a time. Well there is, but it would be prohibitively expensive. Getting it 90% right takes half the effort. 99% right takes the same amount, 99.9% takes that same amount of effort again.
So what happens when someone erroneously gets a penalty because of race ranger error? There’s obvious errors, like 0.0m away from a rider for 100seconds, but what about when it ‘sticks’ and a rider worked through the draft range in 25sec but the data shows 40sec? We’ve all had GPS watches that jump around sometimes. It’s nbd for training, but easy to imagine something as simple as a thick tree lined section causing false positives.
Is there going to be a team of officials going through each and every penalty? And without officials on the road what will the protest process be? This tech makes sense for 100 pros spread over 1 hour with 10 officials. I don’t think it makes sense for 1000 people spread over 10 hours with the same 10 officials
I like the thought of RR but on a mass scale it needs to be very well implemented with the course in mind. There needs to be very logical dead zones where it doesn’t count drafting like out of T1 and long ascents/descents, and similar areas where people are going to get bunched up due to course design.
Also the total time in the draft zone needs a lot of logical thinking behind it. I’ve had races where I was in one of the last waves, and as a good cyclist, I probably spent an absurd amount of time legally in the draft zone from the constant passing. There was one race where I was in the very last AG wave and ran the numbers. I forgot the exact number now but I think from T1 to T2 I passed like 1,000 people. I was passing non-stop the entire bike course. Ride up to someone’s wheel, make pass, rinse and repeat for 56 miles. I would have reached the five minute allowable draft time in probably the first 10 miles.
Flip side is the last time I did the Syracuse 70.3. I was in the first male AG wave. I think it was MPro, FPro, F50+, M40-44 (my AG at the time). Once I got through all the old ladies ( ) I hardly saw anyone else except for a few of the other guys in my AG. So my total time in the draft zone was very low.
I think the rolling start IM uses for most/all races now might mitigate the issue of having 2,000 people on the course ahead of a strong cyclist, unless said cyclist is just a godawful swimmer.
I believe in the other thread where the creators of RR were posting, they said that time in the zone doesnt accumulate until you have expired the time to pass allotment. So if you are going by everyone in the 25 to 30 or whatever is allowed seconds, you would not be getting any time in zone. And I would guess that it is easy to do a restart each time to the next person since all the RR’s are unique individual units…
I think this would be great technology to use now in AG racing. Not to start by using just the data for penalties and such, but to alert folks when they are in the danger zones and to help the referee to make good calls out on the course.
Eventually when the programming gets very specific and the data is considered very good and factual, then you can do a virtual penalty system, allowing a lot of leeway for any gray areas…
The RR technology doesn’t need to be 100% accurate. It just needs to be better than the current ‘human’ technology. And we know that is really really low (can’t have bikes everywhere, even when there they aren’t right, and I say this as someone who has been a ref on a motorbike).
Yesterday I raced the NZ middle distance champs. Part of the course was no overtaking single file due to roadworks narrowing the available space for about 1km. Obviously this led to a queue forming behind one slow rider (throwing no shade, she kicked my arse by swimming past me and getting a 5min+ lead, and this is triathlon). All good. But then later in the day I heard a club mate got a draft penalty within that no passing zone. She was sat up and freewheeling as she was behind a back marker on lap 1, but still got a 3min penalty which was madness.
So I think RR can produce fairer races. BUT. I think it’s a bad idea.
And my reasoning is that if we have rules that start to be written around RR, and the thread above shows how thinking is going with cummulative draft time that is only possible with RR, then it makes grass roots racing which is the feeder to the events that are currently using RR prohibative. And so RR should be for pro/eite racing only.
I would argue it needs to be basically “fool proof” if we are going to go away from essentially “humans” to basically an AI data driven penalty system. And not that I think this is going away from human refs, but what I mean is that if your going to give someone retro active penalty or however they want to give it to an athlete, it’s gotta be above reproach.
And I’m sure they can eventually get to that level, I would guess it would just take a few years of getting the kinks out so to speak. And of course it’s going to fuck up at some point, that’s almost a given. But I think overall you need it to be almost fool proof to then go with any additional penalties from it.
But who knows maybe it’ll just be more of an “on course” marker for both athlete and official to sorta in real time have a cue on drafting for AG ranks.
I do think they can go “eye in the sky” RR data penaltiespur within pros though.
I would guess they could use the data to get officials in the correct positions quicker along the course vs using RR to hand out data driven penalties, and/or “target” certain people if they suddenly keep racing with 15 mins of drafting data, etc. But again I think they could use it with confidence to use data driven penalites much more realistically in the pro fields.
I’m with @BDoughtie on this one. A computer red that is better than human refs is better, but it needs to be perfect on order to replace them. It’s not logical, it’s just the way people are.
Re driverless cars. Much safer on average than human driven cars, but every single death caused by a driverless car is the days big story, on the same day where hundreds are killed by human drivers. People demand perfection from their machine replacements
Not sure why both you guys only think of absolutes, especially with your user name… (-; We can walk and chew gum at the same time. We can have the technology and have it be used by the human refs. We can use its data to target different areas of races, different groups, etc. that show a propensity to be in the red. And eventually like the driverless cars, we can turn the lions share of work over to the programs and hardware.
Yes driverless cars scare the shit out of me right now, but they have already proven to be many times safer than their human counterparts. And when virtually all cars on the road are of that nature, it will be 100’s of times safer. Dont let the one off’s here and there skew the actual math of the problems. There will be a time when pure RR will outperform humans by 100 times too. And just the threat of them is going to change how bike rides in triathlons will be conducted forever.
It will be virtually the same as what draft legal did for eliminating the problems with trying to enforce this seemingly unenforceable issue.
I basically have said nothing in absolutes monty, lol. I basically said in “absolutes” that in order to use it as an actual data driven penalty pathway that yes they will need to make it foolproof. I literally said how they can use the current setup to in real time locate drafting hotspot groups to alert officials out on the course and/or “target” specific individuals if they suddenly keep racing and have 15 mins of drafting in every race they do.
But using this as an actual data driven penalty in AG ranks? That’s going to be years and years before it’s even considered or up for real debate. And again your race company has to even want to do it, which is probaly going to be a bigger problem to solve than anything. (It’s not going to be cheap, especially if you are IM who has to order what likely 10k units easily considering your talking about races over 3 main different continent areas…they aren’t going to all be able to use the 1 single “3k race bulk order” of units and just ship those specific units from race to race like they do at T100 in the pro ranks).
But I guess IM can go to the local cities and show how they can get a few more required hotel stays cus fitting 3k units at a race is going to be an all day task, so now they’ll have even more “bike drop off days” I suppose.
I’m agreeing with you on this. But the masses don’t. “People are dumb, panicky dangerous animals and you know it.”.
Every tri website including this one is going to have headlines when the first AGs are penalized automatically without a human physically seeing the infraction. And the forum (a highly engaged and thoughtful fraction) won’t just be people saying it was correct.
Also, I’m not sure how exactly it gets reviewed by human refs. Post race penalties would be wildly unpopular, so it needs to be in near real time. Which means a wireless uploading mechanism. From there how does a human determine penalty or not? All they have on front of them is data. What interpretation can the human use? If it’s a set standard then why have the human at all? A computer can just yes/no against a standard. Furthermore, this also means hiring additional refs or taking existing refs off the road, exacerbating the issue
Like I said before, this will happen over time, not tomorrow. So it gives folks time to ease into whatever is going to happen, and it will of course be tweaked along the way to work better and better.
But I can easily imagine that along with the rear blinking lights on the seat post, there is one on the bars. Then I can imagine a 3 light system that informs the rider that they are approaching certain metrics in the drafting zones. So for shits and giggles lets say 10 minutes in the zone is a DQ. When you get to 5 minutes you get a yellow light, when you get to 9 minutes you get an orange one, and when you go over it goes red. That would be a real time assessment of your riding with plenty of warnings to let you know you must change your habits.
You could also have a ref on the screens watching all of this, and making sure there are no anomalies causing this information to be incorrect. As I said before, just the fact that every bike in a race will have this, will be the #1 determent to groups forming or intentional drafting. Its like the same with drug testing, the threat of a test is the biggest deterrent to folks trying cheat the system. Only this will be more black and white, until some smart hacker gets his way into the system, and it has to be revamped to accommodate the habitual cheaters…
Sorry, this isn’t driverless cars killing pedestrians. This is an amateur athlete being given a 2 min penalty on a weekend. As it is no-one ever agrees a human ref has it right, and complain that it’s subjective. So here we’re saying the ‘robots’ will apply data. At the moment you can’t appeal a ref’s judgement call, and I would suggest that in RR ‘auto mode’ then there could be the grounds to appeal on specific cases. For example, I’d expect RR to have deactivated zones on hills, tight corners, etc, just as per the application of the rules at the moment. Same f two people have a puncture/mechanical and are stopped beside the road not moving. Technically within 10m but clearly not drafting.
What this does is take away the bile directed at our volunteers / officials from athletes.
Still, the issue of affordability/availability is there. The biggest races need a few thousand of the units, which is a scale/cost issue. Smaller ‘club’ races with a couple of hundred athletes once a year don’t have the funds/capacity to pay for units and the team to run it.
That’s not gonna work tho. Poor swimmer, good cyclist. Spends the first half hour of the race almost exclusively in the draft zone passing good swimmer/bad cyclists. Are they just not allowed to pass anymore once they use up their 10 min? And how does that will not motivate people to “use” their 10 min of drafting towards the end of the bike leg.
Protocol for this has to require the TX/RX units (both on the fork blade and the seatpost) to identify the rider ID ahead/behind, and up to 25 seconds is discounted provided the identities swap (ie the pass was made). This addresses the Long type: all their passes don’t contribute to the ‘margin’ (provided within 25 secs (40 for 20m draft zone).).
I think this type of analysis (from the central hub) will allow effective targeting by motorefs rather than remote data driven decisions.
) I hardly saw anyone else except for a few of the other guys in my AG. So my total time in the draft zone was very low.