Please pick further...
 

  You don't need to be an 'investor' to invest in Singletrack: 6 days left: 95% of target - Find out more

[Closed] Please pick further holes in this non scientific test

22 Posts
17 Users
0 Reactions
41 Views
Posts: 1503
Full Member
Topic starter
 

i'm contemplating getting carbon rims, just cos, and don't know whether 26mm or 30mm I.D. rims will better.

Therefore I've so far done 2 laps of a 90 minute route and the 30mm rims were substantially quicker despite me being far more tired (it was 7am).

different hubs & rims, same tyres & pressures.

How unreliable is this as an assessment?


 
Posted : 05/08/2018 11:59 am
Posts: 7076
Full Member
 

I think it's about 93.7% reliable.


 
Posted : 05/08/2018 12:08 pm
Posts: 12507
Free Member
 

Very.

"Tired" at 7am i think you mean sleepy. Totally different. You've been lying down resting..

Its 4mm man! Look at it on a ruler. Marginal gains perhaps but "noticable" not a chance its the rims width.


 
Posted : 05/08/2018 12:11 pm
Posts: 12482
Free Member
 

Substantially quicker over 90 minutes, so about 10 minutes quicker ?  If so that is not going to be the rim width.


 
Posted : 05/08/2018 12:14 pm
Posts: 0
Free Member
 

How unreliable is this as an assessment?

Extremely.


 
Posted : 05/08/2018 12:23 pm
 geex
Posts: 0
Free Member
 

I once rode fort william WC track with a D521 (21mm) on the front and a EX729 (29mm) on the rear.
It was awful. The rear wheel kept trying to overtake the front in corners and made the front wheel pop into a manual all by itself with it's substantionally quicker acceleration on the straights.


 
Posted : 05/08/2018 12:51 pm
Posts: 0
Free Member
 

The rear wheel kept trying to overtake the front in corners and made the front wheel pop into a manual all by itself with it’s substantionally quicker acceleration on the straights.

That's why I run a 30 mm rim on the front and a 15 mm on the rear. It's like having 4WD, the front wheel just pulls you out of corners if things start getting out of shape at the back.


 
Posted : 05/08/2018 1:05 pm
Posts: 2978
Full Member
 

I'm a bit behind the times...why would a rim with a wider internal diameter be any quicker?


 
Posted : 05/08/2018 1:27 pm
Posts: 9069
Free Member
 

Compare power output of the two rides to the time for a ballpark comparison, if terrain and weather (especially wind direction and strength) conditions were similar.

But being quicker is not often more fun or more comfy.


 
Posted : 05/08/2018 1:44 pm
Posts: 0
Free Member
 

why would a rim with a wider internal diameter be any quicker?

https://www.pinkbike.com/news/rim-widths-comparison-test-mountain-bike.html


 
Posted : 05/08/2018 2:04 pm
Posts: 7656
Full Member
 

Ideally you would need to blind the experiment. Get someone else to choose which wheels to put on before you ride. Not sure how obvious visually it would be though.

In my experience tiredness doesnt always result in riding slower if it is more technical stuff.


 
Posted : 05/08/2018 2:04 pm
Posts: 2978
Full Member
 

why would a rim with a wider internal diameter be any quicker?

https://www.pinkbike.com/news/rim-widths-comparison-test-mountain-bike.html

Interesting...but they only seem to be talking about descending.   What about climbing?


 
Posted : 05/08/2018 2:22 pm
Posts: 65918
Full Member
 

Same pressures bollockses it- wider rims means more volume which has an effect broadly similar to increasing pressures- you usually want to run a high volume tyre at lower pressure.


 
Posted : 05/08/2018 2:24 pm
Posts: 0
Free Member
 

Basically, it would be extremely complex to research. There are multiple variables involved. Obviously, it would best be done blind, but I think that's the least of the problems. First, you need multiple laps under each condition, so a 90 minute loop is probably too long. If you're doing multiple laps on the same day, there are issues such as fatigue, how long it takes to warm up, trail conditions and familiarity, and so on. So you need to run it over multiple days, with each condition tried at different times (i.e. first lap of the day, second, etc). But other factors such as temperature and weather conditions can have a huge effect. If it's cold, it can take a lot longer for the rider to warm up, but if it's hot, fatigue might set in earlier. If it rains during the day, the trail conditions can change drastically. If you run it over multiple days, you need to be consistent about taking rest days, nutrition, sleep, taking a pre-ride dump, etc.

Then there are factors such as tyre pressures. You need to try each rim at a range of different pressures because the optimum pressure may not be identical for different widths. There may also be tyre factors. Some tyres might work better with a wider rim, others with a narrower rim.

Then there are questions of different types of trails and riding style. If it's an XC trail with most time spend climbing and gentle, sweeping singletrack, you might get different results to enduro style stuff with much more rugged descents that load the tyres up much more.

So basically, you would need to time yourself over hundreds of laps on different trails over months and months to be able to get a general answer. Even then, your conclusion would probably be "It depends".


 
Posted : 05/08/2018 2:42 pm
Posts: 41642
Free Member
 

At the very least you need to repeat the test.

Several laps on each set, find the average, find the standard deviation of each, learn some statistical mathematics and compare them to determine the level of significance of any result.

Significance is important as even if one wheel is faster than the other its the significance that tells you that it's actually faster and not just a fluke.

E.g. rim A

91 92 93 94 95 min

Rim B

92 93 94 95 92 min

Shows rim A is faster, but its significance testing that tells you whether you have enough data to state that conclusively or whether the 91min result was a fluke.


 
Posted : 05/08/2018 2:48 pm
Posts: 0
Free Member
 

significance that tells you that it’s actually faster

Statistical significance is mostly a function of your sample size. You need lots of data to get significance, even with a large effect size. With a large dataset, you will nearly always get statistical significance, even with a very small effect size. Ultimately, it's the effect size that is of interest. Significance is just a measure of how much confidence you have in that, and often quite arbitrary. Results that have one chance in 19 of being random are rejected, but 1 chance in 20 are accepted uncritically.


 
Posted : 05/08/2018 2:57 pm
Posts: 1503
Full Member
Topic starter
 

in somethingion the feedback is twofold:

a) it's a bit tricky to assess this kinda stuff in a sciencey way

b) 30mm rims are definitely better

3) unless 26mm rims are you preference, in which case they're waaaaaayyyy betterer

p.s. I need to buy a new bike


 
Posted : 06/08/2018 10:13 am
Posts: 7544
Free Member
 

100% of my testing on wider rims has identified that they puncture more. I swapped to 30mm rims last year and had way more punctures on the same brand of tyres (including news ones, even running a Huck Norris). Reading the interview with the chap from BMC on Pinkbike suggests this is something to do with what wider rims do to the sidewalls.

Other than that it'll make cock all difference.


 
Posted : 06/08/2018 10:19 am
Posts: 17209
Full Member
 

Ride a lap with narrow front wide rear and vice verse for a balanced design. Your endpoint of time is not helpful without adjustment for other variables such as power or heart rate, ie effort .

So basically it is completely unscientific. Anecdotal is the word you are looking for.

BTW I do testing of components and clothing all the time. Constant power laps with time as the endpoint. My laps are 2:30 not 90 minutes, and I repeat them five times. If you could find a short loop of a few minutes and ride that, the green lap at Swinley is a good example, five times per wheelset, then I’d be interested.

Experimental design (in humans) is part of my day job.


 
Posted : 06/08/2018 10:31 am
Posts: 27
Free Member
 

^^^ WARNING: condescension levels critical ^^^


 
Posted : 06/08/2018 11:22 am
Posts: 9306
Free Member
 

So basically, you would need to time yourself over hundreds of laps on different trails over months and months to be able to get a general answer. Even then, your conclusion would probably be “It depends”.

Exactly. Which is why so much of the pseudo-science the bike media come up with to test stuff is pointless. People buy and ride stuff they like the look of, is lighter or feels better somehow, or suits their perceptions etc. eg I have a really fast 29er. Others think it's slow. What works for me and motivates me to ride well isn't the same as another rider.

4mm on the rims - try some blind testing, get the rims covered in masking tape or something like that and ride at 3 different pressures per rim size, keeping tyres the same. Repeat until you think you can decide based on feel.

Or do roll-down tests, also on different pressures per rim. That will tell you most of what counts as 'faster', the rest is grip and that's one you'll need to judge based on experience.

Experimental design (in humans) is part of my day job.

Interesting - does that include anything on perception/feedback loops etc and if so can I PM you on that? Not a commercially related free info request, just personally interested.


 
Posted : 06/08/2018 11:25 am
Posts: 7433
Free Member
 

I find riding blind introduces significant complications into my own scientific testing.


 
Posted : 06/08/2018 11:43 am
Posts: 943
Free Member
 

if you want 'more reliability' then just repeat the test.

if you are asking if the test is suitable then you just need to control all the other variables (weather, power, fatigue, terrain, blindness etc.) which clearly isn't that easy.


 
Posted : 06/08/2018 12:18 pm

6 DAYS LEFT
We are currently at 95% of our target!