Landis' aerobars

in message <[email hidden]>, RonSonic

(') said:

On Fri, 9 Jun 2006 23:13:34 -0700, "Robert Chung" <[email hidden]>

Quoted message said:

Ryan Cousineau said:

In article <[email hidden]>,
"Robert Chung" <[email hidden]> wrote:

> Ryan Cousineau wrote:
>>
>> The other thing, is that what I really admire about the linked
>> article is how the Angry Asian--not you; I mean James Huang*--got
>> someone else to pony up for not one but two custom-built carbon/Ti
>> hardtails. So, did he scam Cyclingnews, Seven, or a bit of both?
>> Either way, kudos to them for going along with the con.
>
> Well, I'm not sure kudos are deserved. From my point of view, it's
> sort of vaguely related to the syndrome that I referred to above,
> where engineers mostly think they can engineer out differences
> between two alternatives. The funny thing is that he leaves out the
> one thing that would let him actually measure a difference. It's an
> extraordinarily pooly conceived experiment that someone's paying
> for. Maybe that was his idea all along. In that case, kudos to him.

Oh, the SRM power meter. Yeah. The worst part is all they would have
had to do was change the bikes to V-brakes on the rear wheel.

Well, they could have:
1. used the SRM crank, or
2. used the Ergomo bottom bracket, or
3. used the Polar power module, or
4. considering that the bikes were -->custom-made<-- added conventional
brake studs and then use the PT.

Quoted message said:

Would you agree that the best way to do this experiment would have
been to ask Seven (or any other 26/29 hardtail maker) to build two
bikes for the same rider, using the maker's preferred geometries,
make components as similar as possible (same
saddle/post/cranks/pedals/bars/brakes), and measure some power
outputs?

Well, that'd certainly be a better experiment. When you have head lice
you have to use a nit comb; you can't use a regular comb. When you're
looking for a small and subtle effects, you have to use a measuring
device that's sensitive enough to pick up small and subtle differences.
With current technology, only an on-bike power meter is going to be
sensitive enough to detect small effects. By leaving out the power
meter, all you end up with is yet another subjective opinion.

Quoted message said:

One thing that does attract me to Huang's experiment is the promise
of repeated trials of these two bikes over multiple courses. I have
the sense that with numerous runs, we may asymptotically approach the
truth.

I'd say repeated trials over one course would suffice if it were
sufficiently varied--the SRM samples at about 1 second intervals, so as
long as you go over bumpy stuff and sandy stuff and downhills and easy
and hard climbs (that is, a normal MTB ride) you'll get enough
observations to compare. Multiple courses sounds good, but it's
actually probably a red herring. Large sample size doesn't make up for
a poor experimental design.

Take a look at Dave Harris' original 26 vs. 29 comparison. The bikes
were probably too different, but other than that the experimental
design was reasonably good. I think it could have been even better if
the race course had been shorter and he'd been able to do more laps,
i.e., more replicates of the same stretches of the course.

From a scientific standpoint I'll agree with you. But, I very strongly
suspect that the real difference will be one of personal preference.
I'll betcha the measurables end up so close that feel and subjective
opinion will outweigh them.

One of the things that's wrong with Huang's experimental design is he's
using identical cassettes and chainsets on both bikes. As one has a
drive wheel 10% bigger than the other, it needs different gearing to
achieve the same speed at the same cadence.

Not, of course, that you /can/ precisely scale chainsets by 10%, but you
could approximate it.

--
[email hidden] (Simon Brooke) http://www.jasmine.org.uk/~simon/

;; in faecibus sapiens rheum propagabit

Simon Brooke said:

One of the things that's wrong with Huang's experimental design is he's
using identical cassettes and chainsets on both bikes. As one has a
drive wheel 10% bigger than the other, it needs different gearing to
achieve the same speed at the same cadence.

Yeah, this is the sort of slavish matching that is completely brainless.
OTOH, with 27 speeds he can probably come close. If he'd used a recording
bike computer he could have determined the gearing post-ride (except,
perhaps, for the Ergomo).

Simon Brooke said:

in message <[email hidden]>, RonSonic

(') said:

On Fri, 9 Jun 2006 23:13:34 -0700, "Robert Chung" <[email hidden]>

Quoted message said:

Ryan Cousineau wrote:
> In article <[email hidden]>,
> "Robert Chung" <[email hidden]> wrote:
>
>> Ryan Cousineau wrote:
>>>
>>> The other thing, is that what I really admire about the linked
>>> article is how the Angry Asian--not you; I mean James Huang*--got
>>> someone else to pony up for not one but two custom-built carbon/Ti
>>> hardtails. So, did he scam Cyclingnews, Seven, or a bit of both?
>>> Either way, kudos to them for going along with the con.
>>
>> Well, I'm not sure kudos are deserved. From my point of view, it's
>> sort of vaguely related to the syndrome that I referred to above,
>> where engineers mostly think they can engineer out differences
>> between two alternatives. The funny thing is that he leaves out the
>> one thing that would let him actually measure a difference. It's an
>> extraordinarily pooly conceived experiment that someone's paying
>> for. Maybe that was his idea all along. In that case, kudos to him.
>
> Oh, the SRM power meter. Yeah. The worst part is all they would have
> had to do was change the bikes to V-brakes on the rear wheel.

Well, they could have:
1. used the SRM crank, or
2. used the Ergomo bottom bracket, or
3. used the Polar power module, or
4. considering that the bikes were -->custom-made<-- added conventional
brake studs and then use the PT.

> Would you agree that the best way to do this experiment would have
> been to ask Seven (or any other 26/29 hardtail maker) to build two
> bikes for the same rider, using the maker's preferred geometries,
> make components as similar as possible (same
> saddle/post/cranks/pedals/bars/brakes), and measure some power
> outputs?

Well, that'd certainly be a better experiment. When you have head lice
you have to use a nit comb; you can't use a regular comb. When you're
looking for a small and subtle effects, you have to use a measuring
device that's sensitive enough to pick up small and subtle differences.
With current technology, only an on-bike power meter is going to be
sensitive enough to detect small effects. By leaving out the power
meter, all you end up with is yet another subjective opinion.

> One thing that does attract me to Huang's experiment is the promise
> of repeated trials of these two bikes over multiple courses. I have
> the sense that with numerous runs, we may asymptotically approach the
> truth.

I'd say repeated trials over one course would suffice if it were
sufficiently varied--the SRM samples at about 1 second intervals, so as
long as you go over bumpy stuff and sandy stuff and downhills and easy
and hard climbs (that is, a normal MTB ride) you'll get enough
observations to compare. Multiple courses sounds good, but it's
actually probably a red herring. Large sample size doesn't make up for
a poor experimental design.

Take a look at Dave Harris' original 26 vs. 29 comparison. The bikes
were probably too different, but other than that the experimental
design was reasonably good. I think it could have been even better if
the race course had been shorter and he'd been able to do more laps,
i.e., more replicates of the same stretches of the course.

From a scientific standpoint I'll agree with you. But, I very strongly
suspect that the real difference will be one of personal preference.
I'll betcha the measurables end up so close that feel and subjective
opinion will outweigh them.

One of the things that's wrong with Huang's experimental design is he's
using identical cassettes and chainsets on both bikes. As one has a
drive wheel 10% bigger than the other, it needs different gearing to
achieve the same speed at the same cadence.

Not, of course, that you /can/ precisely scale chainsets by 10%, but you
could approximate it.

He can always go up or down a cog and get pretty close. That's only the
beginning of his errors in designing this test, in my unhumble opinion. I want
to know if a properly designed 29 is better than a properly designed 26. There's
no telling what is wrong with the geometry of either or both bikes in this silly
attempt to make them otherwise identical.

Let's say, as an example, that the 29er steers slow and feels trucklike on tight
singletrack - is it the 29 wheels or is it the 26 geometry they're bolted to?

Does the rider's size matter?

I'm 6'2 and 220 would a 29er suit me a lot better than it does a little guy?

Is this the rider to tell us that?

Oh well, the man got two bikes out of this, at least one of which he's gonna
really like.

Ron

Robert Chung said:

Simon Brooke said:

One of the things that's wrong with Huang's experimental design is he's
using identical cassettes and chainsets on both bikes. As one has a
drive wheel 10% bigger than the other, it needs different gearing to
achieve the same speed at the same cadence.

Yeah, this is the sort of slavish matching that is completely brainless.
OTOH, with 27 speeds he can probably come close. If he'd used a recording
bike computer he could have determined the gearing post-ride (except,
perhaps, for the Ergomo).

It only makes a difference if he's reaching his limits on gearing,
either spinning out your big gear on the "26", or at suboptimal cadence
in the granny on the "29".

in message <[email hidden]>, Dan

Connelly ('d_j_c_o_n_n_e_l@i_e_e_e.o_r_g') said:

Robert Chung said:

Simon Brooke said:

One of the things that's wrong with Huang's experimental design is
he's using identical cassettes and chainsets on both bikes. As one
has a drive wheel 10% bigger than the other, it needs different
gearing to achieve the same speed at the same cadence.

Yeah, this is the sort of slavish matching that is completely
brainless. OTOH, with 27 speeds he can probably come close. If he'd
used a recording bike computer he could have determined the gearing
post-ride (except, perhaps, for the Ergomo).

It only makes a difference if he's reaching his limits on gearing,
either spinning out your big gear on the "26"

And you never do that on a 26 with MTB gearing?

Quoted message said:

, or at suboptimal cadence in the granny on the "29".

On a 26 you usually fall off the back - or at least can't keep the front
wheel on the deck - before that happens IME, don't know about a 29.

--
[email hidden] (Simon Brooke) http://www.jasmine.org.uk/~simon/

Morning had broken, and there was nothing we could do but wait
patiently for the RAC to arrive.

Dan Connelly said:

It only makes a difference if he's reaching his limits on gearing,
either spinning out your big gear on the "26", or at suboptimal cadence
in the granny on the "29".

Right, but the only way to know whether that ever happens is if you're
recording cadence and speed.

Dave Harris' original "experiment" was a race. He configured each bike in
the way he learned (during other races and training) was fastest for that
particular bike's characteristics. He did laps on each bike. His main
interest was to figure out which bike was faster for him on that course.
IOW, Harris configured each bike optimally and recorded everything. Huang
is configuing each bike identically and recording who-knows-what.

"Robert Chung" <[email hidden]> wrote in message
news:[email hidden]...

Quoted message said:

Dan Connelly said:

It only makes a difference if he's reaching his limits on gearing,
either spinning out your big gear on the "26", or at suboptimal cadence
in the granny on the "29".

Right, but the only way to know whether that ever happens is if you're
recording cadence and speed.

Dave Harris' original "experiment" was a race. He configured each bike in
the way he learned (during other races and training) was fastest for that
particular bike's characteristics. He did laps on each bike. His main
interest was to figure out which bike was faster for him on that course.
IOW, Harris configured each bike optimally and recorded everything. Huang
is configuing each bike identically and recording who-knows-what.

I don't much believe you can tell much either way. Of course on some courses
perhaps one wheels size would be favored while on another some other size
would be favored but I doubt that there's enough difference to be
measureable between the bikes themselves using either method. There's more
variation in the rider's ideas.

Tom Kunich said:

I don't much believe you can tell much either way. Of course on some
courses perhaps one wheels size would be favored while on another some
other size would be favored but I doubt that there's enough difference
to be measureable between the bikes themselves using either method.
There's more variation in the rider's ideas.

http://teamhealthfx.com/blogs/dave_harris/archive/2006/06/08/1258.aspx

Robert Chung said:

Tom Kunich said:

I don't much believe you can tell much either way. Of course on some
courses perhaps one wheels size would be favored while on another some
other size would be favored but I doubt that there's enough difference
to be measureable between the bikes themselves using either method.
There's more variation in the rider's ideas.

http://teamhealthfx.com/blogs/dave_harris/archive/2006/06/08/1258.aspx

Robert, your problem is that you're a member of the
reality based community. Tom is one of history's
actors, and when he acts, he creates his own reality.

Ben
Yeah, I know, the same callback _started_ this sub-thread.

Robert Chung said:

Tom Kunich said:

I don't much believe you can tell much either way. Of course on some
courses perhaps one wheels size would be favored while on another some
other size would be favored but I doubt that there's enough difference
to be measureable between the bikes themselves using either method.
There's more variation in the rider's ideas.

http://teamhealthfx.com/blogs/dave_harris/archive/2006/06/08/1258.aspx

Anything to see your name in print huh?

Tom Kunich said:

Robert Chung said:

Tom Kunich said:

I don't much believe you can tell much either way. Of course on some
courses perhaps one wheels size would be favored while on another some
other size would be favored but I doubt that there's enough difference
to be measureable between the bikes themselves using either method.
There's more variation in the rider's ideas.

http://teamhealthfx.com/blogs/dave_harris/archive/2006/06/08/1258.aspx

Anything to see your name in print huh?

That the best you can do? You wrote that you didn't believe there was
enough difference to measure. I didn't know anything about the course,
didn't know anything about the bikes (in fact, I still get mixed up in
which bike is which and what the differences between them are). However,
by looking at the data I could measure a difference. The difference wasn't
evenly spread out over the lap (that is, that one bike was consistently
reading x% high over the entire course) but was concentrated in one
section. Only then did I tell Dave what I'd found. From the mileage
markings he knew immediately which section it was, and what made it
different from other sections of the course.

I only wish there were more "replicates": each lap was 15 miles long and
this was a race, so the sample sizes for each bike are small. Too bad the
race didn't have shorter laps.

Quoted message said:

Robert, your problem is that you're a member of the
reality based community. Tom is one of history's
actors, and when he acts, he creates his own reality.

Speaking of creating one's own reality:
http://news.yahoo.com/news?tmpl=story&u=/060613/481/2d038d09d1f9433dabb44b98b8b60999

RonSonic said:

to know if a properly designed 29 is better than a properly designed 26. There's
no telling what is wrong with the geometry of either or both bikes in this silly
attempt to make them otherwise identical.

Let's say, as an example, that the 29er steers slow and feels trucklike on tight
singletrack - is it the 29 wheels or is it the 26 geometry they're bolted to?

Ron

The first thing I noticed was the identical head tube angle, which of
course results in a huge difference in the trail measurements for the
two bikes, due to the difference in wheel diameter.

That alone (the different trail measurements) would most likely account
for the different handling feel, not the size of the wheels themselves.

I'm surprised the folks at Seven didn't insist on adjusting the
geometry to try to replicate the overall geometry, not just the raw
angles.

Scott

In article <[email hidden]>, "Robert Chung" <[email hidden]>

Quoted message said:

Quoted message said:

Robert, your problem is that you're a member of the
reality based community. Tom is one of history's
actors, and when he acts, he creates his own reality.

Speaking of creating one's own reality:
http://news.yahoo.com/news?tmpl=story&u=/060613/481/2d038d09d1f9433dabb44b98b8
b60999

That must be a size XXXXL helmet to get Tony Snow's hair into it.

--
tanx,
Howard

Never take a tenant with a monkey.

remove YOUR SHOES to reply, ok?

"Robert Chung" <[email hidden]> wrote in message
news:[email hidden]...

Quoted message said:

Tom Kunich said:

Robert Chung said:

Tom Kunich wrote:

> I don't much believe you can tell much either way. Of course on some
> courses perhaps one wheels size would be favored while on another some
> other size would be favored but I doubt that there's enough difference
> to be measureable between the bikes themselves using either method.
> There's more variation in the rider's ideas.

http://teamhealthfx.com/blogs/dave_harris/archive/2006/06/08/1258.aspx

Anything to see your name in print huh?

That the best you can do? You wrote that you didn't believe there was
enough difference to measure. I didn't know anything about the course,
didn't know anything about the bikes (in fact, I still get mixed up in
which bike is which and what the differences between them are). However,
by looking at the data I could measure a difference. The difference wasn't
evenly spread out over the lap (that is, that one bike was consistently
reading x% high over the entire course) but was concentrated in one
section. Only then did I tell Dave what I'd found. From the mileage
markings he knew immediately which section it was, and what made it
different from other sections of the course.

I only wish there were more "replicates": each lap was 15 miles long and
this was a race, so the sample sizes for each bike are small. Too bad the
race didn't have shorter laps.

Geez Robert, don't be so sensitive. I said that there are horses for
courses. Some courses are likely to be biased towards one size and others to
the other size. Perhaps you're suggesting that cyclocross bikes would be
faster as 26" wheel bikes?

Quoted message said:

Robert, your problem is that you're a member of the
reality based community. Tom is one of history's
actors, and when he acts, he creates his own reality.

So you're saying he's a member of the virtual reality community.