Author Topic: Pbp finishers list  (Read 9397 times)

LittleWheelsandBig

  • Whimsy Rider
Re: Pbp finishers list
« Reply #50 on: January 08, 2020, 04:35:38 pm »
That is a surprising graph. I hadn't anticipated that so many riders were out of time at so many controls.

The graph is cropped of course, so the (comparatively few?) riders with many hours in hand at the finish aren't visible. That might skew the appearance a little but doesn't change the facts.
Wheel meet again, don't know where, don't know when...

frankly frankie

  • I kid you not
    • Fuchsiaphile
Re: Pbp finishers list
« Reply #51 on: January 08, 2020, 05:15:23 pm »
And PBP themselves have traditionally been more concerned about the time you LEAVE a control, not just the time you arrive.  After all the purpose of a control closing time is to get that control closed - and you can't do that with unconscious bodies all over the floor.

The intermediates times thing is interesting because I have long campaigned for a relaxation within AUK's Regs - to better match what actually happens in reality - along the lines of "Intermediate Controls where staffed have predetermined opening and closing times" - because that's what control times is all about, limiting the time that helpers have to stay committed.  No staff, nobody cares.  Queries are common about the importance of intermediate control times on Perms - and of course the Regs say one thing whilst everyone in the real world say another.

With this prededent from PBP it might be easier to get a tweak like that through.
It's not dark yet but it's getting there.

Re: Pbp finishers list
« Reply #52 on: January 08, 2020, 06:18:32 pm »
Here's a chart I've been wanting to make since August of exactly how late homologated riders get to their controls:



HI RES VERSION

The horizontal axis is the closing time of each control. Above the line is on time, below the line is late (according to chip time).

Lines are coloured according to final outcome (green = Homologated, orange = HD, red = Abandonded, purple = Not Homologated.

There's an awful lot of green in that lower right quadrant.

  ;D
It is what it is. It's not what it's not, so it must be what it is.

Re: Pbp finishers list
« Reply #53 on: January 09, 2020, 12:29:43 pm »
Nice to know I can ignore the control times and enter the 80 limit event, slerpnproperly and catch up during the day.

I suspect if you were to analyse time of day for all those below the line you would find most out of time people at the first control they reach in the morning, and everyone in green above the line by the end of each day, even by lunchtime.

I finished in just under 80 hours, my average speed was never less than 15kmh but I would have been out of time 4bor 5 times if my limit was 80 hours. Front loading the control times is the root cause of all these people being behind IMO

Eddington  96miles

simonp

  • Omnomnomnipotent.
Re: Pbp finishers list
« Reply #54 on: January 09, 2020, 12:46:01 pm »
Over four hours out of time and making it up. I was worried about being 20 minutes late at two controls.

Controls were probably staffed for another 12 hours or more for 90 and 84 riders though.

LittleWheelsandBig

  • Whimsy Rider
Re: Pbp finishers list
« Reply #55 on: January 09, 2020, 01:26:33 pm »
And those responses show why this issue will become more widespread in 2023. For the past few PBPs, there has been evidence of a significant (and increasing?) number of riders arriving at intermediate controls out of time but getting to the finish in time and being homologated. Jo (formerly OTP) created a similar graph last PBP http://gicentre.org/pbp2015/ but it didn't look nearly as dramatic as this one.

In '99 and '03, some full value friends that had left intermediate controls as they closed mentioned how noticeable it was that riders were pushing hard to get their brevet cards stamped inside time. Apparently, as the intermediate controls closed for the 80hr and 90hr massed start groups, there was an obvious reduction in the number of riders arriving. With the more recent staggered starts, this behaviour wouldn't be obvious, except in the online timing, but it appears that riders are becoming more relaxed about being out of time at intermediate controls.

The original purpose of PBP Randonneur's intermediate control opening and closing times was to approximate the original Audax average pace schedule (now set at 22.5 km/h on the road). That was why early PBP randonneurs had to sign out of each control in addition to signing into controls. When the randoneur average speed limits were explained to new riders, they were told to imagine that one car was travelling along the route at the maximum allowable average speed and another car was doing the same at the minimum average speed. The rider's aim was always to stay between those imaginary cars, regardless of eating, sleeping, mechanicals or anything else. As soon as the second imaginary car passes the rider, the rider is out of the brevet. A similar system applies to riders and the broom wagon at cyclo-sportives and the Tour de France.
Wheel meet again, don't know where, don't know when...

Re: Pbp finishers list
« Reply #56 on: January 09, 2020, 01:28:27 pm »
The graph is cropped of course, so the (comparatively few?) riders with many hours in hand at the finish aren't visible. That might skew the appearance a little but doesn't change the facts.

Here's the full uncropped "interactive" chart (may take a while to render, probably won't be much fun on phones):
http://www.fondantfancies.com/pbp2019tihchart.html

You can mouse over the datapoints to see bib numbers, although most are too close together.

Some of the more interesting lines on there will be scanning or more likely HTML-scraping errors.

I suspect if you were to analyse time of day for all those below the line you would find most out of time people at the first control they reach in the morning, and everyone in green above the line by the end of each day, even by lunchtime.

You can see quite obvious downward dips around Carhaix in both directions, which (having not done the analysis) likely correspond to people sleeping Monday and Tuesday nights.

Re: Pbp finishers list
« Reply #57 on: January 09, 2020, 01:50:55 pm »
Very well done on the data processing and representation.
I wonder if you'd consider doing one with just the NHs on, as I for one cannot differentiate their lines, even on the HiRes version.

Re: Pbp finishers list
« Reply #58 on: January 09, 2020, 03:44:58 pm »
Very well done on the data processing and representation.
I wonder if you'd consider doing one with just the NHs on, as I for one cannot differentiate their lines, even on the HiRes version.

Here:
http://www.fondantfancies.com/pbp2019tihchartnh.html

I can't see any pattern.

(BTW if you go to pbpresults.com and tick the "DNF" and "Has complete control data" buttons you get 22 of the 24 NHes, and the pace graphs there are more useful than my graph. Have a look at T094...)

Phil W

Re: Pbp finishers list
« Reply #59 on: January 09, 2020, 04:30:17 pm »
Here's the full uncropped "interactive" chart (may take a while to render, probably won't be much fun on phones):
the analysis) likely correspond to people sleeping Monday and Tuesday nights.

You want to use a chart library that draws on the canvas element rather than SVG.  Far far faster for ten or hundreds of thousands of data points.  Then it should render interactive in way under a half a second.

Re: Pbp finishers list
« Reply #60 on: January 09, 2020, 06:59:45 pm »


(BTW if you go to pbpresults.com and tick the "DNF" and "Has complete control data" buttons you get 22 of the 24 NHes, and the pace graphs there are more useful than my graph. Have a look at T094...)

That’s quite an impressive upturn in speed from T094! Others are less clearly suspicious. What reasons do people think they are NH? Lack of a stamp? How would they have no stamp but have a chip time, could they have given their chip to someone else?!

LittleWheelsandBig

  • Whimsy Rider
Re: Pbp finishers list
« Reply #61 on: January 09, 2020, 09:46:15 pm »
A couple of NH folk didn't get stamps at secret controls, with no electronic timing happening.
Wheel meet again, don't know where, don't know when...

Re: Pbp finishers list
« Reply #62 on: January 09, 2020, 09:55:23 pm »
A couple of NH folk didn't get stamps at secret controls, with no electronic timing happening.
Good point, I’d forgotten about those controls!

Alex B

  • Headwind specialist
    • Where is there an end of it?
Re: Pbp finishers list
« Reply #63 on: January 10, 2020, 01:15:06 pm »
I spoke to some riders at Carhaix who had abandoned because they were out of time at the control and assumed the rules would apply as written (actually, what is chapter & verse on this issue for PBP?)

Seems a bit rough if the reality is more relaxed than people are led to expect ...

LittleWheelsandBig

  • Whimsy Rider
Re: Pbp finishers list
« Reply #64 on: January 10, 2020, 01:24:37 pm »
From http://www.paris-brest-paris.org/index2.php?lang=en&cat=randonnee&page=reglement

Article 14 : Opening and Closing Time of the Controls
Passage of the participants within the schedule of closure indicated on brevet cards is compulsory for every control. Opening hours will be also indicated but for information only. Only a serious material incident may be accepted as justification for late arrival, and the delay must be recovered at the latest within the next two controls.

Is this a 'tragedy of the commons' sort of thing? A few folk edging past the strict rule and being homologated is bearable but commonly flouting the rule isn't?
Wheel meet again, don't know where, don't know when...

Alex B

  • Headwind specialist
    • Where is there an end of it?
Re: Pbp finishers list
« Reply #65 on: January 10, 2020, 01:30:05 pm »
Article 14 : Opening and Closing Time of the Controls
Passage of the participants within the schedule of closure indicated on brevet cards is compulsory for every control. Opening hours will be also indicated but for information only.

As I recall the Brevet card just had the control's open time for the entire start group a rider was in. I wonder if it was intended at one point to have personalized opening/closing times too?

Re: Pbp finishers list
« Reply #66 on: January 10, 2020, 02:45:11 pm »
Numbers for riders scanning late who went on to be homologated:
Quote
Rambouillet:   0
Villaines-la-Juhel:   17
Fougères:   6
Tinténiac:   8
Loudéac:   33
Carhaix:   316
Brest:   228
Carhaix:   275
Loudéac:   431
Tinténiac:   841
Fougères:   693
Villaines-la-Juhel:   534
Mortagne-au-Perche:   532
Dreux:   268
Rambouillet:   6

That's 1,164 homologated riders with at least one late scan. Whatever the intention when the rules were written, I presume no one fancied disqualifying a thousand riders.

(I'm using each rider's start scan as their start time for cutoffs. Most riders started very close to their allotted time, so it doesn't make a huge difference. Also Villaines had the scanning mat at the exit on the way out, which might explain the 17 late scans)

simonp

  • Omnomnomnipotent.
Re: Pbp finishers list
« Reply #67 on: January 10, 2020, 03:05:19 pm »
My official time was adjusted for my start scan, by 4-5 minutes. This seems to have been taken into account, and the time shown is the chip time not the time on the card.

The numbers quoted above are an underestimate for lateness at the controller's desk. Many people will have been tight at the chip antenna and will then have taken time (at most controls probably 5-10 minutes) to park up and get to the controller's desk. I was in time by something like 10 minutes at Tintineac on the return, and my stamp was exactly at the control close time (sticky clock?).



Re: Pbp finishers list
« Reply #68 on: January 10, 2020, 03:16:40 pm »
And while I've got the spreadsheet out, the 90th percentile scan times for homologated riders in the 90 hours group:
Quote
Rambouillet   n/a
Villaines-la-Juhel   139 minutes early
Fougères   198 minutes early
Tinténiac   194 minutes early
Loudéac   191 minutes early
Carhaix   7 minutes late
Brest   20 minutes early
Carhaix   24 minutes early
Loudéac   3 minutes late
Tinténiac   118 minutes late
Fougères   91 minutes late
Villaines-la-Juhel   32 minutes late
Mortagne-au-Perche   44 minutes late
Dreux   7 minutes early
Rambouillet: 55 minutes early

That is, the time when 90% of riders who went on to be homologated have scanned in, and 10% still haven't.

Re: Pbp finishers list
« Reply #69 on: January 10, 2020, 03:26:49 pm »
I'm genuinely amazed at how close some people ran it overall but, also, how little concern there is about being out of time at intermediate checks.

FifeingEejit

  • Not Small
Re: Pbp finishers list
« Reply #70 on: January 10, 2020, 04:10:48 pm »
I have a vauge memory of discovering during the ride that it had been announced that due to weather conditions  ??? ??? ??? ??? you only had to be on time at Brest and Ramboulliet.

I'm still baffled by the apparently existence of hills, head wind and the fabled section with a rough surface.

Re: Pbp finishers list
« Reply #71 on: January 10, 2020, 04:51:29 pm »
Numbers for riders scanning late who went on to be homologated:
Quote
Rambouillet:   0
Villaines-la-Juhel:   17
Fougères:   6
Tinténiac:   8
Loudéac:   33
Carhaix:   316
Brest:   228
Carhaix:   275
Loudéac:   431
Tinténiac:   841
Fougères:   693
Villaines-la-Juhel:   534
Mortagne-au-Perche:   532
Dreux:   268
Rambouillet:   6

That's 1,164 homologated riders with at least one late scan. Whatever the intention when the rules were written, I presume no one fancied disqualifying a thousand riders.

(I'm using each rider's start scan as their start time for cutoffs. Most riders started very close to their allotted time, so it doesn't make a huge difference. Also Villaines had the scanning mat at the exit on the way out, which might explain the 17 late scans)
Suggest large numbers on the 3x400km plan, sleeping at loudec in both directions, maybe with a hotel room booked for 2 nights.

Eddington  96miles

Re: Pbp finishers list
« Reply #72 on: January 10, 2020, 05:20:01 pm »

I'm still baffled by ......... the fabled section with a rough surface.
Was it in the courtyard at the finish?

Eddington  96miles

FifeingEejit

  • Not Small
Re: Pbp finishers list
« Reply #73 on: January 10, 2020, 06:43:00 pm »

I'm still baffled by ......... the fabled section with a rough surface.
Was it in the courtyard at the finish?

Nah apparently it was a descent somewhere and someone was apparently posted there to warn you of it.
I suspect if it was a rough surface it was considered so under French standards not Scottish (The B7076 is smooth.... honest)

Alex B

  • Headwind specialist
    • Where is there an end of it?
Re: Pbp finishers list
« Reply #74 on: January 10, 2020, 06:48:59 pm »
... a descent somewhere and someone was apparently posted there to warn you of it.

That was for a descent before Loudéac on the return. It wasn't really a bad surface (by UK standards anyway). Didn't some people get routed through a field to route around an incident with emergency services in attendance (or something)?