View Full Version : Tour '04...showcase of speed?
Things are looking good for the flat stages this year.
Peta, Boonen (who doesn't think Petacchi poses a problem
for him to win 2-3 stages and the green), McEwen, Cooke,
and the aging superstars, Zabel and Cipo. All these teams
together provide a lot of horsepower and motivation for
bunch finishes.
This could be the fastest TdF ever (until the AAA-unlimited-fuel-
class professional league is created).
jim gravity wrote:
>
> This could be the fastest TdF ever
http://anonymous.coward.free.fr/temp/tdf.png
2004 TdF is 3395 km.
Thanks, I was looking for that.
This is a really interesting plot. I notice, though, that
there is a pretty good correlation of Tour length with date--
more recent Tours have been shorter--as well as the clear
correlation of average speed with length. So it's hard to be
sure whether racers can average a higher speed when the race
is shorter (which seems plausible), or whether racers have
gotten faster over the years because of better
training/equipment/nutrition/whatever.
Glenn
"Robert Chung" <me2@privacy.net> wrote in message news:<2gpja9F5gimnU1@uni-
berlin.de>...
> jim gravity wrote:
> >
> > This could be the fastest TdF ever
>
> http://anonymous.coward.free.fr/temp/tdf.png
>
> 2004 TdF is 3395 km.
Glenn Cooper wrote:
> This is a really interesting plot. I notice, though, that
> there is a pretty good correlation of Tour length with date--
> more recent Tours have been shorter--as well as the clear
> correlation of average speed with length. So it's hard to
> be sure whether racers can average a higher speed when the
> race is shorter (which seems plausible), or whether racers
> have gotten faster over the years because of better
> training/equipment/nutrition/whatever.
>
> Glenn
>
It's straightforward enough to do a bivariate regression to
separate the two. Barring that, eyeball it -- take vertical
slices of the plot, and you'll observe a clear date
dependence.
Dan
"Robert Chung" <me2@privacy.net> wrote in message
news:2gpja9F5gimnU1@uni-berlin.de...
> jim gravity wrote:
> >
> > This could be the fastest TdF ever
>
> http://anonymous.coward.free.fr/temp/tdf.png
>
> 2004 TdF is 3395 km.
What's the correlation with the multiple road improvements
over the last 40 years? The top half of many of those climbs
were broken up or even gravel up until the early 70's and
some of them into the 90's if memory serves. Bad asphalt
makes a pretty dramatic difference versus good asphalt in
average speed. So does the increasing general fitness of the
peloton which allows more breakaways. And that changes the
tactics to generate still more early breakaways that tend to
hold the speeds higher.
Hinault could do long breakaways but who has been able to do
that since then? Chiapucci when no one believed he had a
chance and so didn't pursue with any seriousness?
> It's straightforward enough to do a bivariate regression to
> separate the two. Barring that, eyeball it -- take vertical
> slices of the plot, and you'll observe a clear date
> dependence.
It gets fuzzy around 4200-4500km though.
So, 1999 was the last time Cipolini rode the tour...it was a pretty fast year, given it's length.
And the protour is starting next year. I figure that is going to slow things down.
Dan Connelly wrote:
> Glenn Cooper wrote:
>> This is a really interesting plot. I notice, though, that
>> there is a pretty good correlation of Tour length with
>> date--more recent Tours have been shorter--as well as the
>> clear correlation of average speed with length. So it's
>> hard to be sure whether racers can average a higher speed
>> when the race is shorter (which seems plausible), or
>> whether racers have gotten faster over the years because
>> of better training/equipment/nutrition/whatever.
>>
>> Glenn
>>
>
> It's straightforward enough to do a bivariate regression
> to separate the two.
Distance and year are highly collinear.
http://anonymous.coward.free.fr/temp/tdf.csv
Scatterplot matrix:
http://anonymous.coward.free.fr/temp/tdf2.png
Dan Connelly wrote:
> Glenn Cooper wrote:
>
>> This is a really interesting plot. I notice, though, that
>> there is a pretty good correlation of Tour length with
>> date--more recent Tours have been shorter--as well as the
>> clear correlation of average speed with length. So it's
>> hard to be sure whether racers can average a higher speed
>> when the race is shorter (which seems plausible), or
>> whether racers have gotten faster over the years because
>> of better training/equipment/nutrition/whatever.
>>
>> Glenn
>>
>
> It's straightforward enough to do a bivariate regression
> to separate the two. Barring that, eyeball it -- take
> vertical slices of the plot, and you'll observe a clear
> date dependence.
>
> Dan
>
Well, I don't have a proper multivariate regression code,
but I did my best with Open Office calc.... (the gnumeric
solver wouldn't work for me). Maybe I'll try Excel when I'm
on widoze.
But the model seems to be close to :
speed = s0 + a (y - d) s0 = 36.05 kph y == (year - 1970) /
(10 years) d == (distance - 4000 km) / 1000 km a = 0.926 kph
Dan
Tom Kunich <cyclintom@yahoo.com> wrote:
> "Robert Chung" <me2@privacy.net> wrote in message news:2gpja9F5gimnU1@uni-
> berlin.de...
>> jim gravity wrote:
>> >
>> > This could be the fastest TdF ever
>>
>> http://anonymous.coward.free.fr/temp/tdf.png
>>
>> 2004 TdF is 3395 km.
> What's the correlation with the multiple road improvements
> over the last 40 years? The top half of many of those
> climbs were broken up or even gravel up until the early
> 70's and some of them into the 90's if memory serves. Bad
> asphalt makes a pretty dramatic difference versus good
> asphalt in average speed. So does the increasing general
> fitness of the peloton which allows more breakaways. And
> that changes the tactics to generate still more early
> breakaways that tend to hold the speeds higher.
> Hinault could do long breakaways but who has been able to
> do that since then?
Thierry Marie, didn't he solo into Brittany (or was it
Brittney, naw that wuz the guy frum the Bayou) for Castorama
in the late 80's? Something like 100+kms? Fignon had another
in his last daze.
Chiapucci when no one believed he had a chance and so
didn't pursue
> with any seriousness?
"Clovis Lark" <clark@steel.ucs.indiana.edu> wrote in message
news:c8b6ou$n2k$4@hood.uits.indiana.edu...
> Tom Kunich <cyclintom@yahoo.com> wrote:
> > "Robert Chung" <me2@privacy.net> wrote in message news:2gpja9F5gimnU1@uni-
> > berlin.de...
> >> jim gravity wrote:
> >> >
> >> > This could be the fastest TdF ever
> >>
> >> http://anonymous.coward.free.fr/temp/tdf.png
> >>
> >> 2004 TdF is 3395 km.
>
> > What's the correlation with the multiple road
> > improvements over the last
40
> > years? The top half of many of those climbs were broken
> > up or even
gravel up
> > until the early 70's and some of them into the 90's if
> > memory serves.
Bad
> > asphalt makes a pretty dramatic difference versus good
> > asphalt in
average
> > speed. So does the increasing general fitness of the
> > peloton which
allows
> > more breakaways. And that changes the tactics to
> > generate still more
early
> > breakaways that tend to hold the speeds higher.
>
> > Hinault could do long breakaways but who has been able
> > to do that since then?
>
> Thierry Marie, didn't he solo into Brittany (or was it
> Brittney, naw that wuz the guy frum the Bayou) for
> Castorama in the late 80's? Something like 100+kms? Fignon
> had another in his last daze.
Nah, I'm talking about a contender. Fignon would NEVER have
been allowed on a long leash when he was a team leader. He
was absolutely great. Too bad his ego was so fragile. Fignon
had a superb 1989 - easily the best ridre of that year for
my money - but lost the Tour by 8 seconds and the World's by
a bike length and never could do much after that.
Tom Kunich <cyclintom@yahoo.com> wrote:
> "Clovis Lark" <clark@steel.ucs.indiana.edu> wrote in
> message news:c8b6ou$n2k$4@hood.uits.indiana.edu...
>> Tom Kunich <cyclintom@yahoo.com> wrote:
>> > "Robert Chung" <me2@privacy.net> wrote in message news:2gpja9F5gimnU1@uni-
>> > berlin.de...
>> >> jim gravity wrote:
>> >> >
>> >> > This could be the fastest TdF ever
>> >>
>> >> http://anonymous.coward.free.fr/temp/tdf.png
>> >>
>> >> 2004 TdF is 3395 km.
>>
>> > What's the correlation with the multiple road
>> > improvements over the last
> 40
>> > years? The top half of many of those climbs were broken
>> > up or even
> gravel up
>> > until the early 70's and some of them into the 90's if
>> > memory serves.
> Bad
>> > asphalt makes a pretty dramatic difference versus good
>> > asphalt in
> average
>> > speed. So does the increasing general fitness of the
>> > peloton which
> allows
>> > more breakaways. And that changes the tactics to
>> > generate still more
> early
>> > breakaways that tend to hold the speeds higher.
>>
>> > Hinault could do long breakaways but who has been able
>> > to do that since then?
>>
>> Thierry Marie, didn't he solo into Brittany (or was it
>> Brittney, naw that wuz the guy frum the Bayou) for
>> Castorama in the late 80's? Something like 100+kms?
>> Fignon had another in his last daze.
> Nah, I'm talking about a contender. Fignon would NEVER
> have been allowed on a long leash when he was a team
> leader. He was absolutely great. Too bad his ego was so
> fragile. Fignon had a superb 1989 - easily the best ridre
> of that year for my money - but lost the Tour by 8 seconds
> and the World's by a bike length and never could do much
> after that.
nothing to do with that partial dislocation of the hip in
the Giro 90? Naw...
Dan Connelly <d_j_c_o_n_n_e_l@i_e_e_e.o_r_g> wrote in message news:<fP8qc.68113$pw5.39904@newssvr25.news.prodigy.com>...
> Well, I don't have a proper multivariate regression code,
> but I did my best with Open Office calc.... (the gnumeric
> solver wouldn't work for me).
If something in gnumeric does not work right please let us
know at bugzilla.gnome.org -- we mean for things to work.
Thanks,
Morten
vBulletin, Copyright ©2000-2009, Jelsoft Enterprises Ltd.
Search Engine Friendly URLs by
vBSEO 3.3.0
