Actually the accuracy and precision are very high. How directly it applies to riding on a asphalt road surface is a fair question. It certainly provides an accurate and objective comparison of the tires, as they are all tested in highly repeatable conditions with high accuracy.

And it provide comparison for GP4000s2 for 23mm, 25mm and 28mm

Can you explain why you believe that is the case?

Are you sure that's the tire and not the wheel?

Sure, you can question whether the absolute .crr values he comes up with are correct and/or relevant to riding on asphalt or concrete. When you look at the rankings and relative differences between different tires, though, his results generally jibe with others who've tested the same tires.

---

"They're made of latex, not nitroglycerin"

Yep

Because I removed the rubbing tyre (of course) and installed another one (GP Attack 22mm), and now the clearance is constant (and larger), so wheel is perfect. And new tyre also.

The GP4000s2 stretched a bit in one place only, probably slightly more than 1mm.
Not an issue, really, I will reuse it on back wheel.
But no more 25mm on the P3 front wheel.

Based on his published testing protocols and, as others have noted, the consistency of his numbers when compared to other testing sources.
Can you explain why you are skeptical of his accuracy? If your issue is the relevance of the numbers to real world use on asphalt roads, I understand your point, but there is no practical way to resolve this as there are infinite variations in hardness and textures of road surfaces.
IMO this data is much more useful than cdr numbers for bike frames, or any component that is part of a much greater system and is tested on it’s own.

I'm not concerned about mimicking asphalt. The apparatus that holds the wheel has a sliding fit with one tube inside another (friction). I'm not sure how the load is applied, but as I recall that was not a simple weight, but rather a load cell of some kind. And measuring motor input and subtracting an efficiency is not the best way to determine power. There are too many opportunities for random errors and a drift in calibration.

The "consistency of numbers compared to other sources" argument doesn't hold weight. Errors big enough to be obvious in that respect would need to be pretty huge. Since he is testing and comparing tires over a span of years, I'd like to know if he gets the same results with the same tires over that span. For instance does the GP4000 measured now have the same Crr as 4 years ago?

That would be nice, but how do you propose he preserve the tire over a 4 year period from environmental changes? Bottom line, I trust his testing and have had good results picking tires from his data. If you are looking for absolute verified perfection, good luck :-) BTW- he has many results posted on his site of the same tire used in multiple tests, you can certainly review those and decide if it meets your standards. If not, and if you have a better source for more reliable tire data, I would greatly appreciate you sharing, as would many other on ST

If he has Tire A as 8% higher .crr than Tire B, which is 5% higher than Tire C which is 7% higher than Tire D, and then you see another tester come up with the same order and very similar relative differences, and then a third, you don't think that gives some validation to his methods?

---

"They're made of latex, not nitroglycerin"

 
I agree with gary p and IntenseOne that BRR is making very accurate evaluations of tires, both for dimensioning, puncture resistance and rolling resistance.

Did not found better source in the web.

Of course the margin error can be discussed, but from comparison with others tests, it appear margin error is lower than differences between most tyres.

And I'm not worried about this margin error, because, anyway, when hitting real road, Crr will significantly change, and possibly differently for different tyres and different types of roads...
... plus, the impedance effect will also reverse the lower sides of the Crr/pressure curve, differently based on weight, road, speed, ...

So, real life Crr will be very different, and is much more multi-parameter and complex than the curve given by BRR.
BRR possible margin error is far from being significant then...

IMO, BRR curves are an excellent basis. Not perfect, but really good enough as a first evaluation.

Then you have to make your OWN evaluation for :
1) butyl, light butyl, latex (BRR supply good evaluation for these differences)
2) tyre size (BRR also supply good evaluation for these differences) - taking into account bike clearances
3) optimal pressure (most complex parameter because of impedance "reverse curve" effect depending on pressure, but also road, tyre, speed, ...)

From my personal experience, real life rolling resistance measured are up to 2,3 times BRR figures, on "not good" asphalt.

Show me.

The reason I questioned his setup in the first place (many years ago) was because his numbers did not match other sources. The apparatus is inherently susceptible to error, particularly over time. If there isn't a way accurately calibrate it, and no attempt is made to quantify the degree of error... then it's hard to have faith in the results.

RR just published its results on the TL GP5000 - pretty awesome!

very nice indeed. Link to results =>

---

---

Save: $50 on Speed Hound Recovery Boots | $20 on Air Relax| $100 on Normatec| 15% on Most Absorbable Magnesium

Blogs: Best CHEAP Zwift / Bike Trainer Desk | Theragun G3 vs $140 Bivi Percussive Massager | Normatec Pulse 2.0 vs Normatec Pulse | Speed Hound vs Normatec | Air Relax vs Normatec | Q1 2018 Blood Test Results | | Why HED JET+ Is The BEST value wheelset

Looks like the TL GP5000 is basically the same as the standard GP5000 with a latex tube.

(assuming the ~2w latex delta is similar to the GP4000 tests).

That is what I am coming up with - I think I am trying tubless for the first time ever this season based on the various results we have seen!