So what's 3w anyway?

The theme for this year seems to be 3w. That's the amount of savings that wind tunnel tests say you save by using a Zipp 404 instead of an RFSW3 (or Al33 custom build) or an FSW3. Those are also at par with a Zipp 303 - 3w down to a 404. Then there are the other wheels right behind that group - HED Belgium+ and FLO30. I'll bend one of our rules and backstop the Easton R90SL with the HED/FLO group. Since that test in February, we've had a lot of conversations that focus on this 3w concept. But what exactly does it mean?

Ostensibly, 3w in a wind tunnel means around 9 seconds over the course of a 40k TT, assuming no other variables and a 30mph bike speed. More subjectively, it could mean the difference between basking in victory and choking on the salty tears of defeat (though I'd note that no Olympic or World TT margin of victory in recent memory would have been affected by the winner using box section wheels versus whole-hog aero wheels). It could be the increase in your FTP after a few hard weeks of training. Or it could be noise in the system as the simple test in this CyclingTips review of Roval wheels would indicate. Personally, I'm sold that it's more the latter, with a side of "though I try to stay in shape and would much rather be thought of as that sort of fast debonair older gentleman than that slow old fart I just don't see 3w being that difference in any case." 

Does my shadow make me look old?

We've studied tires quite a bit before (like what became this blog, for which I was nearly burned at the stake as a heretic for concepts which we all take for granted 3 short years later), but this year it's been a big focus for us. With sites like Bicycle Rolling Resistance doing all the interesting work that they do, plus articles like this one in VeloNews getting more attention, it's really evident what a big roll (I'll be here all week, ladies and gentlemen!) tires play in your speed. WAY more than 3w on offer there. Heck you can reverse that 3w delta by using latex tubes versus regular weight butyl tubes, which is worth 3w per tire (and you're not supposed to use latex tubes with carbon because you might die!). 

In addition to the speed gains from good tires, there's a comfort and enjoyment enhancement with really supple tires. I've gone from being a general "life is too short to ride crappy tires" guy to being an incurable tire snob, and I'm happier for it.

As so often happens, we're going to counter the bike industry and general market's proclivity to see things as "black and white, one way or the other" with our own completely muddled "well, it depends..." Are there aerodynamic differences between wheels? Sure. Can small amounts matter? The Sahara desert is made up of darn near microscopic particles, so yeah. But are there a whole bunch of other things that convolute the discussion? Yes, that's the big point. 

Back to blog


Hi Jon,

The problem with building alloy wheels much deeper than 32-ish or so is with the way they’re made. When you make an alloy rim, you extrude a straight piece of metal which then gets rolled into a circle. When the outer dimension of the circle gets SO MUCH bigger than the inside dimension, the extra material from the inside needs a place to put itself. Quite often, the inside edge of the rim gets wavy as the material creates that “place to go” for itself. There are some rims out there that are alloy and deep, and there are some ways to mitigate it, but then you run into your other guess – they do get heavy. The very few 40mm+ deep alloy clincher rims I know of are all 600g or more.

Dave Kirkpatrick

There are now allow wheels (Kinlin’s in FSW’s, Al33’s, etc.) that can be the aero equivalent of a good carbon wheel like a Zipp 303. So I was wondering… Are there any allow wheels, current or on the horizon, that can be in the same aero range as, say, a Zipp 404? Or would that be too much metal and, hence, too heavy? There could be a market there although I don’t know how large. Many people who hang out here (myself included) are happy going alloy when you get similar aero benefits to that Zipp 303, but have a more rugged rim that brakes well. If you’re already getting new allow wheels and those 3 Watts are available, it might be worth going that way…

Jon W

Gummee – At some point everything has a threshold, right? It could be 1w that makes the difference. However, in the situation you describe, you are a) drafting, where aerodynamic impact gets cut by roughly 1/3 b) at low yaw angles, where Zipp wheels do relatively poorly – they do better a 10* c) conflating aero with deep is a mistake – we’ve seen 80+mm deep wheels that SUCKED aero-wise – at basically all angles. And then of course you have to have vetted your tire/rim combo for aero, and compared that to rolling resistance outcomes, etc. And be sure that the aero you’re supposedly getting isn’t being paid for by previous efforts spent due to compromised braking, energy wasted in managing crosswind handling, etc. So you might have a strong placebo response of “I’m using deep wheels, therefore I can hang” but our unvarnished take on your proposition is “nope, no way.” Your mileage and ours clearly vary on that.


That 3w sometimes means hanging on to the back of the really fast guys or getting dropped. AMHIK

I’ll bet that some of its a placebo, but overall, I like my aero wheels when I’m going fast. If I’m going slowly, well… 32h training wheels are where it’s at.


Zack – Right you are. I’ve just never been able to figure out when, outside of a no-draft triathlon when you’re trying to minimize time on course at relatively low speeds with relatively high aerodynamic importance, this dynamic is relevant. And even then, it’s very small. An ultra endurance event? You’re carrying a bunch of stuff that muddies the question and handling and reliability is at a super premium – I’d choose the 3w worse aerodynamics of the AForce there rather than go the 404 route 1000 times out of 1000 on that. Apart from those, maybe you’re doing a recovery ride of set distance with an effort ceiling and you need to get home to mow the lawn by “x” o’clock?

Dave K

Leave a comment

Please note, comments need to be approved before they are published.