Author Topic: Controls + Distances  (Read 8620 times)

marcusjb

  • Full of bon courage.
    • Occasional wittering
Re: Controls + Distances
« Reply #100 on: August 23, 2019, 09:24:28 am »
^ plenty of jiggery-pokery goes on it seems - as above, many people seem to finish at 90:00 exactly....

I honestly would not worry Grams - you were well in time at the end; you will be homolowhatsitted no problem.
Right! What's next?

Ooooh. That sounds like a daft idea.  I am in!

Re: Controls + Distances
« Reply #101 on: September 24, 2019, 10:33:56 pm »
I've just managed to get the data into a processable format, and my first attempt at analysis says that of the 4102 riders with both a complete set of scans and who finished under the time limit, only 3085 actually made it to every control on time.

There are 702 people who were at least an hour behind at least once, and 458 of them were at least two hours behind! I even found a couple of riders who were 9 hours behind at one point, and somehow finished on time. Tinteniac on the way back seems to the worst place.

It's entirely possible I cocked something up, or the data's wonky, but if these numbers are right then they can't enforce intermediate controls without disqualifying rather a lot of riders.

Re: Controls + Distances
« Reply #102 on: September 25, 2019, 07:39:50 am »
At some of the controls you rolled over the timing mat on the way in as well as on the way out.  I'd bet that if the timing mat didn't register their visit on the way in then the time registered might be for the way out.  Just after the ride I noticed that one of the times in my tracking data was registered on the way out of the control but looking at it now I couldn't tell you which one it was.

Re: Controls + Distances
« Reply #103 on: September 25, 2019, 09:57:08 am »
When the tracking site first came up but before the start there was a note saying it would use riders average speed to estimate location between controls, to do this they would have to time going out as well as going in. When the ride started the note was changed to say it would only show location of last control entered, perhaps they didn't have enough mats for all controls for in & out. 2023 is a long time away in technology so perhaps there will be a magic app that will track and tell each rider individually their personal time in hand etc.

Ps I actually kept a copy of the note
https://storage.ning.com/topology/rest/1.0/file/get/3423227588?profile=original

Re: Controls + Distances
« Reply #104 on: September 25, 2019, 10:15:51 am »
I find all this analysis fascinating. You might expect that more data, and more investigation of that data, would result in better performance, with more people finishing. The opposite seems to be happening, with more HDs, and more DNFs.

I'd see it as a cybernetic problem. The basic control language for a French event should be French. But the field contains fewer French speakers in each successive edition. The next default is English, but there are a large number of dialects being used. That means that mathematics becomes an important feedback path.

But if you've got the legs, French is more important than maths, as it it makes sense of the controls, and of the shops. Tablets and smartphones carry the promise of a single world culture, but you are still in France.

Phil W

Re: Controls + Distances
« Reply #105 on: September 25, 2019, 05:37:29 pm »
I find all this analysis fascinating. You might expect that more data, and more investigation of that data, would result in better performance, with more people finishing. The opposite seems to be happening, with more HDs, and more DNFs.

I doubt that many of the 6000-7000 riders are analysing data from previous PBPs at all.  There's little if any official analysis and publication I'm aware of. Nothing in the dossier that I'm aware of. What would be interesting (to me) is to see if the amount of time stopped is increasing and if so, by how much each edition? How much does your start affect time stopped? PBP really does suck time at controls like no other.  Are on the road speeds dropping or are controls just leeching away time?

zigzag

  • unfuckwithable
Re: Controls + Distances
« Reply #106 on: September 25, 2019, 05:50:03 pm »
controls are what they were in previous two editions - i haven't noticed any big changes. some are better, some less good, but for their basic function (stamp, snack, drink) they were all pretty good and efficient (tip: avoid paying with contactless card).

Re: Controls + Distances
« Reply #107 on: September 25, 2019, 05:59:01 pm »
I was thinking more of the analysis by the organisers. LEL attempted to draw lessons from previous editions. In the end there were more DNFs following the analysis than before.

I'd say it's because the net is being thrown wider than before. Fewer of the participants belong to clubs with PBP experience, and fewer have experience of riding in groups, and of riding TTs. They think that can be offset by spreadsheets, consumer electronics, novel frame materials, electric gears and 1x chain sets. In addition, fewer speak French.

Phil W

Re: Controls + Distances
« Reply #108 on: September 25, 2019, 06:24:48 pm »
I was thinking more of the analysis by the organisers. LEL attempted to draw lessons from previous editions. In the end there were more DNFs following the analysis than before.


But the analysis of the 13 data wasn't of DNFs not was it in order to reduce DNFs. That was more about trying to help smooth the peaks of riders in controls. But there's many factors, including who has entered, the weather, (human physcology you can nudge but no more) you can't control, which may render the efforts null and void. then add in changes in numbers and more hills.  You can see that the rider start time in 17 did not have a significant impact on finish rates. The last start did have the best finish rate but then you start asking who is more likely to opt for a late start.

I'd be wary of expecting an outcome of an analysis which was never intended to achieve it.

I'm not sure that PBP changed much in their format from 15 other than the forced move to Rambouillet.

Re: Controls + Distances
« Reply #109 on: September 25, 2019, 06:51:13 pm »
I'm surprised that analysis isn't aimed at reducing DNFs. I suppose that might accord with a desire to make the event 'tougher'.

The main change to the format from 2015 for PBP was the way the entry panned out. Initially it was the same as 2015, but the dramatic fall in the proportion of French pre-qualifiers caused an increase in the overall allocation. As it turned out, there wasn't an increase in the field beyond the envisaged 6,000, and certainly no increase in validations. Meanwhile, French participation fell below 25%.

Phil W

Re: Controls + Distances
« Reply #110 on: September 25, 2019, 07:17:57 pm »
I'm surprised that analysis isn't aimed at reducing DNFs. I suppose that might accord with a desire to make the event 'tougher'.

I think it's just a recognition that's there's only so much an organiser can do, and there are many factors outside their control which impact DNF rates much more.

The most interesting factor that can be influenced is human physcology.  So the don't worry about control times northbound meant a number were taking it easy going north. A number didn't take advantage of the tailwind in the fens and didn't think about headwinds on their return etc.  Riders who reached Louth after dark were increasingly reluctant to push on to Pocklington. So 2021 provide a shelter control midway by Humber bridge can influence that desire to ride into the night etc.  In fact there was a distinct drop off in riders on the road once it got dark. But then in 13 the problem was Pocklington not Market Rasen the equivalent of Louth in that edition.

It is  easy to think an organiser can control the DNF rate, but I think that's an illusion unless they massage those out of time into the finishers, to match some kind of idealised figure. Changing the rules after the fact. That get increasingly hard for them to do with online tracking.

LittleWheelsandBig

  • Whimsy Rider
Re: Controls + Distances
« Reply #111 on: September 25, 2019, 07:39:00 pm »
So the current lot of audaxers/ randonneurs aren’t a patch on those of a generation or two ago, because the current lot DNF more?
Wheel meet again, don't know where, don't know when...

Re: Controls + Distances
« Reply #112 on: September 25, 2019, 08:33:27 pm »

It is  easy to think an organiser can control the DNF rate, but I think that's an illusion unless they massage those out of time into the finishers, to match some kind of idealised figure. Changing the rules after the fact. That get increasingly hard for them to do with online tracking.

We had electronic tracking in 1999. It wasn't online, except via Minitel in France. It was done with magnetic-strip cards. By 2003 it was available in real time online. 1999 and 2007 had two hour time extensions, as did LEL 2009. There's always a bit of confusion with those, ensuring that participants are aware of the changes at the time they happen.

There's possibly a case for a short extension due to the roadworks diversion this year, but a bit unfair retrospectively.

Re: Controls + Distances
« Reply #113 on: September 26, 2019, 01:40:00 am »
So the current lot of audaxers/ randonneurs aren’t a patch on those of a generation or two ago, because the current lot DNF more?

LOL! To be fair I think 2019 was one of the tougher PBPs with headwinds both ways and cold nights. By contrast I found 2015 the easiest that I’ve completed with hardly any wind and milder nights. 2015 was also dry for me as I finished before it started raining.

Anyone not finishing the last PBP or LEL shouldn’t be too hard on themselves as they were both very tough rides. They should just learn from their experiences and have another go next time. The main advice I would give would be to not squander any time in hand they build up as it is more precious than gold.


Re: Controls + Distances
« Reply #114 on: September 26, 2019, 05:14:49 am »

...but the dramatic fall in the proportion of French pre-qualifiers caused an increase in the overall allocation. As it turned out, there wasn't an increase in the field beyond the envisaged 6,000, and certainly no increase in validations. Meanwhile, French participation fell below 25%.

I wondered why they didn't let the attrition after the initial full registration stand to reduce crowding - not enough French riders on long brevets last year? I also wonder if DNF/Abandon rates correlate to registration date.

Re: Controls + Distances
« Reply #115 on: September 26, 2019, 07:23:15 am »
So the current lot of audaxers/ randonneurs aren’t a patch on those of a generation or two ago, because the current lot DNF more?

LOL! To be fair I think 2019 was one of the tougher PBPs with headwinds both ways and cold nights. By contrast I found 2015 the easiest that I’ve completed with hardly any wind and milder nights. 2015 was also dry for me as I finished before it started raining.

Anyone not finishing the last PBP or LEL shouldn’t be too hard on themselves as they were both very tough rides. They should just learn from their experiences and have another go next time. The main advice I would give would be to not squander any time in hand they build up as it is more precious than gold.

^^^
Wisdom without hyperbole
#makewattsnotwar

Re: Controls + Distances
« Reply #116 on: September 26, 2019, 08:00:20 am »
At the end of my blog (only a few pages of guff to get to the bottom) I added a table of the number of starters and finishers for every country based on the PBP results unofficial website.  It maybe of interest to those with a bent for getting to the bottom of the stats.

With the possible exception of Brazil the closer a country is to the equator the lower the rate of finishing on time. Doubtless its more complicated than that!

https://yorkbadger.wordpress.com/2019/09/07/pbp-2019/

Stephen

Sent from my SM-G950F using Tapatalk


Re: Controls + Distances
« Reply #117 on: September 26, 2019, 09:16:26 am »

...but the dramatic fall in the proportion of French pre-qualifiers caused an increase in the overall allocation. As it turned out, there wasn't an increase in the field beyond the envisaged 6,000, and certainly no increase in validations. Meanwhile, French participation fell below 25%.

I wondered why they didn't let the attrition after the initial full registration stand to reduce crowding - not enough French riders on long brevets last year? I also wonder if DNF/Abandon rates correlate to registration date.

In 2015, we ended up supplying photographs for an article in Cycling Plus, the largest circulation cycling magazine in the  UK. I was approached to approve re-use of that material in mid-December 2014. I assume that most magazines and websites were working to that sort of schedule. Previously, they might have sent a staff writer to tick off PBP, but the pre-qualification seemed to rule that out.

French riders in the Randonnee heartlands in the North seemed to have done the pre-qualifiers. Perhaps we lost those who would have been inspired by 'bucket-list' articles in the media. There are other 'bucket list' events after all, centred on nostalgia or climbing.

We might then expect that the 'successful' applicants would have more time to research aspects such as controls and distances, or likely weather, given their earlier commitment to participation.

So it's interesting to consider if number-crunching can do much to explain the DNF rate, and once explained, do anything to lower it next time. Or whether that is even an aim, given a preference for 'diversity'.

I'd tend to focus on the 'cultural' factors. The main one being a reduction in the number of French-speakers in the field. That will have a bearing on communication between riders and control staff.

Re: Controls + Distances
« Reply #118 on: September 26, 2019, 09:29:17 am »
There are two kinds of France nowadays in regard to PBP.
You have the randonneuring heartlands, Bretagne, North and Ile de France. Riders there are aware of the changes in PBP, they do their brevets every year and keep on top of things. They know that PBP is an international event.
Rides from rural France outside of the 3 mentioned area's. They still think that PBP is as 25 years ago, start their preparations in march of the PBP year and were awakened very rude shortly after that. Many of them don't ride brevets in between PBP years.
Their reactions on the French forum are wildly different. The first group completely agrees with the pre-qualification rules and has prepared for them. The 2nd group feels that French riders are entitled to a start spot and complain because they 'couldn't' do BRM's in 2018. When I suggested that they could also organise themselves I was met with a digital blank stare.

in contrast, the crowds in Bretagne react according to their area. They are fully aware that PBP is an international event and enjoy it.

Re: Controls + Distances
« Reply #119 on: September 26, 2019, 10:16:06 am »
It maybe of interest to those with a bent for getting to the bottom of the stats.
Why only the supine?

Re: Controls + Distances
« Reply #120 on: September 26, 2019, 12:26:17 pm »
There are two kinds of France nowadays in regard to PBP.
You have the randonneuring heartlands, Bretagne, North and Ile de France. Riders there are aware of the changes in PBP, they do their brevets every year and keep on top of things. They know that PBP is an international event.


I have been mistaken for a Norman or a Fleming at Semaine Federale, those being the biggest and tallest of the French. There are cultural links with the area I'm from, so it's understandable. So the Norman controls, Mortagne and Villaines are my favourites. Villaines is actually very badly laid out. But they provide a map of the site, and a couple of children to guide you through the maze.

I do like Brittany, and the great virtue from an organisational point of view is that there are good, non-toll, roads. There is a mobile volunteer element. The guy with the microphone, who the French term an 'animateur', turns up at various points on the route. It's interesting to disentangle the core functions from the control-based ones.

The electronic timing is a core function, connected to HQ. The stamping of the cards is a local function, carried out by local volunteers, and there isn't actually a link between them. I arrived at Loudeac wanting to know when the first group would arrive back from Carhaix. That information wasn't available, and I cobbled together an estimate from the tracking, and a knowledge that Zigzag had been doing well.

The obvious answer to the appetite for feedback is to put the time in hand on the tracking. Those with a smartphone can then use that. Those who want to use other methods can.

FifeingEejit

  • Not Small
Re: Controls + Distances
« Reply #121 on: September 26, 2019, 01:10:07 pm »
There's quite a lot that could be done with the electronic tracking the problem is, a lot of that would only be used once every 4 years (unless LEL and other organizers got behind it).

Re: Controls + Distances
« Reply #122 on: September 26, 2019, 01:19:52 pm »
LEL had barcodes being scanned at the same point as the signing. So a display of time-in-hand could be entered in a dedicated box on the card, if that is thought to be a good idea. A similar system could be used at PBP to provide a bit more management information for the control.

Re: Controls + Distances
« Reply #123 on: September 26, 2019, 07:22:49 pm »


I'd say it's because the net is being thrown wider than before. Fewer of the participants belong to clubs with PBP experience, and fewer have experience of riding in groups, and of riding TTs. They think that can be offset by spreadsheets, consumer electronics, novel frame materials, electric gears and 1x chain sets. In addition, fewer speak French.

I believe there was a time when even UK riders were new on the PBP scene, had to learn the ropes etc. Once one or two have experience there is a snowball effect and off you go. Look at ACME, a few in 2015 being led by Tomsk and then a whole heap more (12ish) this time round, many relatively new riders (under two years) < this is the difference and it relates to participation; the more participations you can bank, or be around those who have experience, the better off you are. Or indeed watching films about previous editions. I would say that counts vastly more than the spreadsheets, electrics etc which don't make much of a difference either way. As for knowing French it doesn't make a blind bit of difference. I didn't speak a word of it beyond please and thank you. I am all for a wider net - next time more of those people will come back and finish and educate those around them. Was this not how Audax UK started? In that sense I don't think the DNF rate is particularly important, or rather that it might fluctuate with the experience gained by that country over the years. Corollary is that you shouldn't try to make flagship events 'easier'. PBP and LEL are direct experiences that you can't fake, that's what makes them valuable, I don't think anyone really beleives they can be offset with technology do they?

As for novel frame materials you mean the minority on steel?  ::-)

LittleWheelsandBig

  • Whimsy Rider
Re: Controls + Distances
« Reply #124 on: September 26, 2019, 07:51:58 pm »
The early AUKs were mostly long-distance TTers (24hr Fellowship) or hard-riding tourists and had a DNF rate around 10% or below. That certainly isn’t how most newer PBP countries start nowadays, though the Russians and some Eastern European contingents hew close to that model.

ACP historically was very interested in DNF rates, using the PBP plaquette to praise countries with low DNF rates and encouraging other countries to improve.

It is quite noticeable that many French randonneurs and audaxers get ever lighter and more expensive bikes while making jokes about compensation for progressive dilapidation.
Wheel meet again, don't know where, don't know when...