Yesterday, I finally worked through some final details of our wheel deflection measurement jig. The jig lets us precisely measure the lateral deflection of any wheel with a quick release hub that fits 9mm or 10mm dropouts. This covers most wheels, including every wheel we've ever sold (fronts are 9mm, rears are 10). After way too long of a day yesterday, we were able to measure several front wheels and are already learning a bunch. It will take time, but the plan is to evaluate everything; rims, hubs, spoke count, spoke type, lacing - everything.
The remaining problem with the whole deal is developing an effective way to quantify and communicate the differences between wheels. As it is now, we could take a wheel that deflects, say .092" in our protocol, and compare that to another wheel that deflects .125", and say that the first wheel is 25% stiffer than the second (difference between the two, divided by the deflection of the second). We could even say the first is 35% stiffer, by taking the difference between the two and dividing it by the deflection of the first. This second way is the way that's used whenever you buy any consumer product that claims to give you "25% MORE FREE!" or whatever. Some similar methodology is used when, for example, a 2015 model year crankset is compared to a 2014 model year crankset, in order to show the greatest improvement. A 30% increase in stiffness is impressive, right?
Unsurprisingly, we don't plan to do that. The goal of this whole exercise is to help us to guide you in selecting the best wheel for you and the plans you have for it. We know, through a few years of subjective testing, that there are wheels that are empirically "stiff enough for whoever wants to do whatever with it," and that there are wheels whose lack of stiffness compromises their performance. So if we take the former wheel and call that a 10, and we take a wheel that's just at the margin where any further loss of stiffness would really hamper it and call that a 3, that seems more useful to us. It leaves some space below the range for any true bombs we find, and provides enough granularity that the differences between one wheel and another actually show up. There's no established protocol, and no standardization of measurement tools, so any comparison of absolute values is the height of worthlessness. We have measurement precision, but slightly fuzzier relative values are much more valuable in this case.
Added to what we've learned (and continue to learn) in the wind tunnel, this all gives us a profound ability to help you decide which components and build are right for you. The development of this testing tool has been an obsession of mine for about a month, and now that it's done I've got the latitude to start measuring more parameters. Brake heat development in real world conditions and time to stop measurement are up next. Mike's a little freaked out about the bills I rack up at McMaster, so it's definitely time to pick some low-hanging fruit for a while (fortunately I already bought the heat scanner, so I don't have to worry about that one).
A couple of quick codas to add. First, having made something as cool as our rig is, you might expect us to blow up the internets with photos. Nope. This thing took a long time, a pile of money, a bunch of wrong turns, and several dropped curse words to develop, and having it is an advantage for us. Second, while we have no plans to go out and buy a pile of wheels (or even any wheels) specifically for the purpose of comparing what we do to what's on the market, if you possess a wheel that you'd like to get measured, we're happy to do it. Contact us and we'll arrange it. Last, thanks to Justin at Octo for helping me to figure out how to put this contraption together. The notes I've got from this round give a huge headstart on the next few measurement fixtures.