The null hypothesis and a heretical suggestion about brushes?

I've read several threads about how good or bad various brushes are.
What I haven't seen is any evidence for those claims. If readers know
of any such tests, please let me know - while I find race stories
interesting, they aren't evidence in the sense I'm using the word.
The null hypothesis, as most of you know, simply states that there is
no demonstrable, statistically significant, difference between A and B.
In this case I'll call A the copper brush and B a similar but
non-copper brush (bronze, soft steel, etc.). To REJECT the null
hypothesis and show that A and B have different effects on ski speed
you would have to test them against each other, e.g., with matched
skis/waxes, etc., through a speed trap and carry out the appropriate
statistics on the test results.
The companies that sell these brushes such as Toko and Swix, just to
name two, have been testing waxes for years - at least 60 years in
Swix's case. And I have no bone to pick with either company - I use
and like their products.
But I think it would be simple for them to test brushes the same way
they test waxes: brush several pairs with A, several pairs with B and
test them with the protocols they use when testing waxes. The skis
could be re-waxed and re-brushed multiple times, they could be skied-in
by skiers, they could be tested over a period of months or even years.
I assume they haven't done that (to my knowledge - if the data are out
there, please let me know) and appear instead to rely on advertising to
persuade us to buy one brush or the other. Or they did the tests,
couldn't find a difference, and could see no point in telling us that -
an entirely understandable decision.
Indirect "tests" may exist - these threads often refer to one brush or
another creating microscopic hairs that slow the ski and then seal the
base when heated. Again, has anyone ever tested those claims?
Many of you probably recall the debates that fueled the audio magazines
(and may still for all I know) about whether a separate amp and pre-amp
would improve sound quality over that of a decent integrated tuner.
Stereo Review did some tests with the ABX procedure and was not able to
reject the null hypothesis: tuners produced sound indistinguishable
from separates.
I don't know of a comparable test-oriented magazine for skiers and the
brush manufacturers provide us with advertising but not test data.
Until there is some evidence, for or against one brush, isn't it
justifiable to assume that there isn't any difference?
BTW, I'm not putting anyone down and I'm just as susceptible to ads as
the next guy, as my brush collection shows.

Well, show us the data when you finish your research. With so many
variables to control for, you ought to be finished about the same time
that global warming kills the sport.

Randy
wrote:
I've read several threads about how good or bad various brushes are.
What I haven't seen is any evidence for those claims. If readers know
of any such tests, please let me know - while I find race stories
interesting, they aren't evidence in the sense I'm using the word.
The null hypothesis, as most of you know, simply states that there is
no demonstrable, statistically significant, difference between A and B.
In this case I'll call A the copper brush and B a similar but
non-copper brush (bronze, soft steel, etc.). To REJECT the null
hypothesis and show that A and B have different effects on ski speed
you would have to test them against each other, e.g., with matched
skis/waxes, etc., through a speed trap and carry out the appropriate
statistics on the test results.
The companies that sell these brushes such as Toko and Swix, just to
name two, have been testing waxes for years - at least 60 years in
Swix's case. And I have no bone to pick with either company - I use
and like their products.
But I think it would be simple for them to test brushes the same way
they test waxes: brush several pairs with A, several pairs with B and
test them with the protocols they use when testing waxes. The skis
could be re-waxed and re-brushed multiple times, they could be skied-in
by skiers, they could be tested over a period of months or even years.
I assume they haven't done that (to my knowledge - if the data are out
there, please let me know) and appear instead to rely on advertising to
persuade us to buy one brush or the other. Or they did the tests,
couldn't find a difference, and could see no point in telling us that -
an entirely understandable decision.
Indirect "tests" may exist - these threads often refer to one brush or
another creating microscopic hairs that slow the ski and then seal the
base when heated. Again, has anyone ever tested those claims?
Many of you probably recall the debates that fueled the audio magazines
(and may still for all I know) about whether a separate amp and pre-amp
would improve sound quality over that of a decent integrated tuner.
Stereo Review did some tests with the ABX procedure and was not able to
reject the null hypothesis: tuners produced sound indistinguishable
from separates.
I don't know of a comparable test-oriented magazine for skiers and the
brush manufacturers provide us with advertising but not test data.
Until there is some evidence, for or against one brush, isn't it
justifiable to assume that there isn't any difference?
BTW, I'm not putting anyone down and I'm just as susceptible to ads as
the next guy, as my brush collection shows.

wrote:
Well, show us the data when you finish your research. With so many
variables to control for, you ought to be finished about the same time
that global warming kills the sport.

Randy

What's your point?

The OP asks a good question but didn't consider how difficult it is to
control the variables involved, anymore than he investigated Stereo
Review's claim of not being able to hear differences between audio
components (the latter goes back decades and presumably has to do with
the relatively poor sonic quality of mass market components made by
their big buck advertisers).

As for brushes, my sense of it is that top level wax techs do all
sorts of things that marketing departments wouldn't want to advertise.
It would seem that the main thing is to know what you want to
accomplish with each brush or tool, and see it that's occurring. Ask
for advice from other experts who don't have a stake in the sales of
one or another brand.



"Zeke" wrote:


wrote:
Well, show us the data when you finish your research. With so many
variables to control for, you ought to be finished about the same time
that global warming kills the sport.

Randy

What's your point?

I guess I wasn't clear despite a pretty long post. I'll try again:

In the ski world as I know it we prep skis largely by conventional
wisdom: reps, fast skiers, our pals in the ski shop and our buddies
all have ideas about the best way to do it. I'm certain a lot of it is
good advice, just as I'm certain that some of it will be found to be
wrong - same thing in medicine, psychology, any field.

The copper brush is VERY popular (I use one and have nothing against
it) and might even be reaching the status of the white nylon, that is,
you can/should use it just about every time you wax - cold wax, medium
temp wax, warm wax.

If that's true it makes the variables easier, not harder, to control:
test it when you wax cold, when you wax medium, and when you wax warm,
in comparison to the glide speed of the skis brushed with the
non-copper.

If it's truly a superior tool (while the other is, as one person called
the bronze brush, the "brush of death,") it shouldn't take years of
waxing, ironing and brushing to find out. It's the near-universality
of its recommended use (according to the manufacturers and the much
more experienced skiers I know) that should make the test relatively
easy.

(I didn't intend to offend anyone by bringing up Stereo Review and ABX
sound testing - it's not that germane to this.)

As long as you have the same skier on identically flexed skis, the
same conditions, the rest of the wax job identical, the previous speed
of the skis identical, clean brushes with equal wear, brushing
calibrated to be equal relative to the nature of the brushes
(pressure and number of strokes) ... what am i leaving out?


wrote:

I guess I wasn't clear despite a pretty long post. I'll try again:

In the ski world as I know it we prep skis largely by conventional
wisdom: reps, fast skiers, our pals in the ski shop and our buddies
all have ideas about the best way to do it. I'm certain a lot of it is
good advice, just as I'm certain that some of it will be found to be
wrong - same thing in medicine, psychology, any field.

The copper brush is VERY popular (I use one and have nothing against
it) and might even be reaching the status of the white nylon, that is,
you can/should use it just about every time you wax - cold wax, medium
temp wax, warm wax.

If that's true it makes the variables easier, not harder, to control:
test it when you wax cold, when you wax medium, and when you wax warm,
in comparison to the glide speed of the skis brushed with the
non-copper.

If it's truly a superior tool (while the other is, as one person called
the bronze brush, the "brush of death,") it shouldn't take years of
waxing, ironing and brushing to find out. It's the near-universality
of its recommended use (according to the manufacturers and the much
more experienced skiers I know) that should make the test relatively
easy.

(I didn't intend to offend anyone by bringing up Stereo Review and ABX
sound testing - it's not that germane to this.)

I can't imagine ski speed is influenced very much by brush type. The
difference would come in the time it takes to brush your skis out. I
may be off base but I think you could get top notch brushing results
with a quiver of toothbrushes it you had lots of time and patience.

Personally I like the copper because it takes quite a bit of wax out
quickly and like others have said, it doesn't appear to add any
structure to the ski. Follow with white nylon or horse hair then a
polishing brush and call it good. I think most of what is said about
brushes is marketing speak.

DMK

On Oct 29, 3:04 pm, wrote:
I guess I wasn't clear despite a pretty long post. I'll try again:

In the ski world as I know it we prep skis largely by conventional
wisdom: reps, fast skiers, our pals in the ski shop and our buddies
all have ideas about the best way to do it. I'm certain a lot of it is
good advice, just as I'm certain that some of it will be found to be
wrong - same thing in medicine, psychology, any field.

The copper brush is VERY popular (I use one and have nothing against
it) and might even be reaching the status of the white nylon, that is,
you can/should use it just about every time you wax - cold wax, medium
temp wax, warm wax.

If that's true it makes the variables easier, not harder, to control:
test it when you wax cold, when you wax medium, and when you wax warm,
in comparison to the glide speed of the skis brushed with the
non-copper.

If it's truly a superior tool (while the other is, as one person called
the bronze brush, the "brush of death,") it shouldn't take years of
waxing, ironing and brushing to find out. It's the near-universality
of its recommended use (according to the manufacturers and the much
more experienced skiers I know) that should make the test relatively
easy.

(I didn't intend to offend anyone by bringing up Stereo Review and ABX
sound testing - it's not that germane to this.)

My two cents:

Brushes are used to remove wax and to polish or work-harden the base.
We're not talking about the type of base modifications that wax and
structure make. All things being equal you'd have a very hard time
finding ski-speed performance differences between brushes unless you
simply didn't manage to do the required job with one of them. That
should be plainly evident to the eye. No need to use a speed trap to
determine that you haven't got the wax off the ski base!

That's not to say there aren't differences between brushes. Generally I
do 90% of my brush work with one brush. My current favorite is the new
fine steel brush from Swix. The Toko copper brush works well also. I'm
not a big fan of the various bronze brushes. The criteria for a good
brush is that it should not physically move base material, it should
quickly remove wax, and it shouldn't clog up. All of these metal
brushes need to be broken-in carefully before you use them on good skis
to meet the first requirement.

Once you've removed the wax a hard nylon brush works well to polish the
base. You can "scrub" in both directions at this point in the process,
and use quite a lot of pressure.

On the whole the process should be quite fast, and somewhat tiring.
I've seen many people ritualize their brushing practices - count
strokes, go through specific orders of brushes and change them for
different waxes. On the whole it's safe to say that this won't make a
difference. Find a good metal brush that won't destroy your base and
use it for 90% of your work. Then find a good polishing brush. And
you're all set!

Zach

In my tests, I found that no matter what steps you take to try and make
good scientific conclusions about ski speed, your final result will not
be believed by a significant proportion of people. People have their
beliefs, and it's hard to sway them from those beliefs.

As for testing skis, almost no one has closely matches pairs of skis. I
did my testing by using two pair (call them pair A, and pair B), and
then intermixing those skis. So one ski from pair A and one ski from
pair B would get Swix HF6, and the other ski from pair A and the other
ski from pair B would get Toko HF Red. This then assumes the left and
right ski in each pair is equal both in flex and grind. (A pretty
decent assumption.) You can only compare two waxes in this way.

In order to glide test the skis with a timer, you have to glide in
tracks. Each run down the tracks makes the tracks glaze. So usually you
alternate the skis and record each time, then graph the results with a
line, and look for a gap between the two lines. The glide also depends
on the hill incline, and the timing marks (magnets in the snow) should
be placed after you're up and gliding and before you slow significantly
at the bottom.

It's fun when people yank your magnets out the snow, look at them, and
then toss them into the deep stuff. (Yes it happens.)

One problem is gusty winds. Another difficulty is trying to start
exactly the same, tuck exactly the same each run, and have good balance
since all those are hidden factors in speed. Another factor is how far
the skis have been skied. Skis should probably be tested for initial
speed, be skied a good 2-5 km, and then tested again. Significant
differences are usually in tenths of a second, and I think it's
important to graph each of the points and simple to SD analysis on the
numbers (since that can hide trends).

With all of that said, I agree with Zach that I don't believe it would
be easy to detect a difference in brushes, and hard to establish a
defined end point for brushing. There would likely be a large
difference for slightly underbrushing a ski as compared to using two
different brushes.

I think the metal brushes have become popular because they really speed
up brushing (and, for the most par, blow away rotobrushes). I again
agree with Zach that it's probably a good idea to break in the Swix
fine metals before using them for race wax jobs. (They're really sharp
when they're new.) If you use the brush exclusively in one direction,
they're very sharp in the opposite direction, even when they're well
used. The Toko copper has less of this effect, but it's still there.

As for rotobrushes, they're great for polishing...

Jay



wrote:
As long as you have the same skier on identically flexed skis, the
same conditions, the rest of the wax job identical, the previous speed
of the skis identical, clean brushes with equal wear, brushing
calibrated to be equal relative to the nature of the brushes
(pressure and number of strokes) ... what am i leaving out?

Jay,
Is it worth while to build and setup a glide down timer? With all the
changing variables during the testing, maybe different
temperature/conditions come race day, conditions changing during the race
(temperature usually warming as the day goes on) and the fairly wide range
waxes will work in.
Zach did glide down tests the day before the Birkie and recorded the
times. Mike Wynn picked the same wax/grind that was the fastest recorded by
the timer, just by using feel. He put one wax/grind ski combination on one
foot and another combination on another foot. I've seen elite skiers pick
their race wax the same way, by just using feel.

--

Paul Haltvick
Bay Design and Build - LLC
Engineering, Construction and Information Technology Services
FSx - Fischer / Swix Racing

"Bjorn A. Payne Diaz" wrote in message
oups.com...
In my tests, I found that no matter what steps you take to try and make
good scientific conclusions about ski speed, your final result will not
be believed by a significant proportion of people. People have their
beliefs, and it's hard to sway them from those beliefs.

As for testing skis, almost no one has closely matches pairs of skis. I
did my testing by using two pair (call them pair A, and pair B), and
then intermixing those skis. So one ski from pair A and one ski from
pair B would get Swix HF6, and the other ski from pair A and the other
ski from pair B would get Toko HF Red. This then assumes the left and
right ski in each pair is equal both in flex and grind. (A pretty
decent assumption.) You can only compare two waxes in this way.

In order to glide test the skis with a timer, you have to glide in
tracks. Each run down the tracks makes the tracks glaze. So usually you
alternate the skis and record each time, then graph the results with a
line, and look for a gap between the two lines. The glide also depends
on the hill incline, and the timing marks (magnets in the snow) should
be placed after you're up and gliding and before you slow significantly
at the bottom.

It's fun when people yank your magnets out the snow, look at them, and
then toss them into the deep stuff. (Yes it happens.)

One problem is gusty winds. Another difficulty is trying to start
exactly the same, tuck exactly the same each run, and have good balance
since all those are hidden factors in speed. Another factor is how far
the skis have been skied. Skis should probably be tested for initial
speed, be skied a good 2-5 km, and then tested again. Significant
differences are usually in tenths of a second, and I think it's
important to graph each of the points and simple to SD analysis on the
numbers (since that can hide trends).

With all of that said, I agree with Zach that I don't believe it would
be easy to detect a difference in brushes, and hard to establish a
defined end point for brushing. There would likely be a large
difference for slightly underbrushing a ski as compared to using two
different brushes.

I think the metal brushes have become popular because they really speed
up brushing (and, for the most par, blow away rotobrushes). I again
agree with Zach that it's probably a good idea to break in the Swix
fine metals before using them for race wax jobs. (They're really sharp
when they're new.) If you use the brush exclusively in one direction,
they're very sharp in the opposite direction, even when they're well
used. The Toko copper has less of this effect, but it's still there.

As for rotobrushes, they're great for polishing...

Jay



wrote:
As long as you have the same skier on identically flexed skis, the
same conditions, the rest of the wax job identical, the previous speed
of the skis identical, clean brushes with equal wear, brushing
calibrated to be equal relative to the nature of the brushes
(pressure and number of strokes) ... what am i leaving out?