Armstrong is hated

[Warren]

In article <[email protected]>, Nick Burns <[email protected]> wrote:

> "Norm T" <[email protected]> wrote in message

> > Ullrich is not really much better this year than in '01, its mainly Lance who has just
> > gotten OLD.
>
> No, you are wrong. Ullrich is significantly stronger. Armstrong is older than last year, but
> aren't you too? Who do you know that is not a year older than last year?

Cher.

 

[Daniel Connelly]

Daniel Connelly wrote:
>
>
> Kurgan Gringioni wrote:>> He rode 2 minutes slower at Alpe'd'Huez. If he had that sort of
> form this
>> year, his lead now would be 6+ minutes as in previous years.
>
>
>
> Under much different conditions.
>

Followup:

Ferrari provides climbing rates:
http://www.cyclingnews.com/riders/2003/diaries/ferrari/tour.php?id=ferrari031

L'Alpe:
----------
In fact, en excellent Iban Mayo climbed the final ascent in 39:05 (VAM of 1735 m/h), two minutes
slower than Marco Pantani's record.

Farther back come Armstrong and the group of the best riders (VAM of 1640 m/h).They were all
controlling each other, trying to avoid risks to keep something in reserve for the next few days,
beginning with tomorrow¹s 9th stage, with Lautaret and Izoard, ending in Gap, one of the hottest
places in France.

Peyresourde -------- ---- Vinokourov and Mayo climbed the 13 km of Col de Peyresourde in 30'20",
with a VAM of 1800 [vertical] m/h. The group with Ullrich, Armstrong, Basso and Zubeldia climbed in
31'15", performing a VAM of 1747 m/h.

Tuormalet, Luz Ardiden:
-------------------------
On the previous mountain, the Tourmalet, Ullrich set another fierce tempo with his typical full
and powerful pushes on the pedals: a time of 44'30", with a VAM of 1700 m/h. Armstrong replied
with a VAM of 1740 m/h towards Luz-Ardiden: a total time of 35'05".

 

[Dieter Buerssne]

Daniel Connelly wrote:

> Good idea... I did that for time, but with a log x axis, have problem with nonpositive places, so
> kept that axis unchanged.

You could use a linear x-axis, of course. Or give Botero/Ullrich place 0.5.

Dieter

 

[Daniel Connelly]

Well, maybe a sgn(x)sqrt(abs(x)) axis. I want to emphasize low placings....

Dan

Dieter Buerssner wrote:
> Daniel Connelly wrote:
>
>
>>Good idea... I did that for time, but with a log x axis, have problem with nonpositive places, so
>>kept that axis unchanged.
>
>
> You could use a linear x-axis, of course. Or give Botero/Ullrich place 0.5.
>
> Dieter

 

[Jenko]

Daniel Connelly wrote ...
>
> The spread in the time trial results suggests otherwise:
>
> http://www-tcad.stanford.edu/~djconnel/cycling/ITT1.pdf

Hardly convincing. Very different TTs. Much less riders went full gas this year. Let me provide a
counteranalysis. Assume the top 25 riders in the GC ride the TT near their maximum, and only take
them into account. Consider Ullrich an outlier. Compute mean, stdev, and normalize Armstrong time
(the most negative the value, the better). What gives? 2002, 1st TT: -1.65 (-1.71 if Botero is not
taken into account) 2003, 1st TT: -1.31 (-1.79 if Ullrich is taken into account)

Interpretation: the gap between GC contenders grew more year-to-year than the gap between Armstrong
and those contenders.

I know that limiting the analysis to the top 25 is somehow flawed, given that this year's TT suited
GC riders more than last year's flat one. But it's IMHO closer to an apples to apples comparison
than considering the whole field - if you do, the result confirms your conclusion ...
2002: -2.38 (-2.34 without Botero)
2003: -2.47 (-3.05 with Ullrich)

You've gotta love statistics!

Jenko

 

[Davey Crockett]

On Tue, 22 Jul 2003 05:43:12 -0700, Jenko wrote:

> Daniel Connelly wrote ...
>>
>> The spread in the time trial results suggests otherwise:
>>
>> http://www-tcad.stanford.edu/~djconnel/cycling/ITT1.pdf
>
> Hardly convincing. Very different TTs. Much less riders went full gas this year. Let me provide a
> counteranalysis. Assume the top 25 riders in the GC ride the TT near their maximum, and only take
> them into account. Consider Ullrich an outlier. Compute mean, stdev, and normalize Armstrong time
> (the most negative the value, the better). What gives? 2002, 1st TT: -1.65 (-1.71 if Botero is not
> taken into account) 2003, 1st TT: -1.31 (-1.79 if Ullrich is taken into account)
>
> Interpretation: the gap between GC contenders grew more year-to-year than the gap between
> Armstrong and those contenders.
>
> I know that limiting the analysis to the top 25 is somehow flawed, given that this year's TT
> suited GC riders more than last year's flat one. But it's IMHO closer to an apples to apples
> comparison than considering the whole field - if you do, the result confirms your conclusion ...
> 2002: -2.38 (-2.34 without Botero)
> 2003: -2.47 (-3.05 with Ullrich)
>
> You've gotta love statistics!
>
> Jenko

Figures don't Lie, but Liars can sure Figure

It's a bike race fer goodness' sake.

Anything can happen.

So reject the Hypothesis that the Best Rider will win and accept the NULL which I think is the
accepted statistical parlance:

In layman language:

Even a Wheel Sucking ****** like Armstrong can win it.

 

[Jenko]

Davey Crockett wrote ...
>
> It's a bike race fer goodness' sake.
>

Ok, I'll make a bike racing observation. Ullrich didn't feed properly in the Tourmalet descent. Only
took something near the end of it. He was approaching bonking when Lance attacked. That explains his
time loss and final recovery.

You've gotte love cycling!

Jenko

--
Posted via Mailgate.ORG Server - http://www.Mailgate.ORG