A new nomenclature for bike fit and sizing
Originally 2.19.03 by Dan Empfield
Updated in February, 2007
(www.slowtwitch.com)

CHAPTER ONE (BELOW)
CHAPTER TWO
CHAPTER THREE

Earlier in the decade I wrote an article on how bikes are sized, and the gist was that we ought to be looking at bikes from the point of view of head tubes and top tubes, and that's it. -- no more seat tube sizing.

I wrote this because it doesn't matter where the seat tube is. It used to matter, and I'm old enough to remember when it was useful. When I started racing -- in the era when we nailed our cleats onto the bottoms of our shoes -- all road bikes were made the same way, with 1 1/8" diameter down tubes and 1" top tubes. Top tubes were horizontal and were not "dropped." There was no such thing as "compact geometry." All stems were "minus 17 degrees," which is to say they were parallel to the ground. The only sort of bike was the standard steel road race bike. Neither aluminum nor carbon were available to frame makers and the one and only titanium bike made was the Teledyne Titan, out of commercially pure titanium.

The world has changed. We have global warming, terrorist attacks on U.S. soil, and bike fit nomenclature that makes no sense. Heck, in my last years at QR our sizing system was "center to air," which meant we measured from the center of the bottom bracket axle to the center of some vapor point we made up in our heads. We didn't want to spend half our day answering questions about our fit nomenclature, so we just figured that if a particular size was made for a guy who would ride a 59cm normal bike we'll call it a 59cm. It was a minor detail to us that the bike didn't actually measure 59cm anywhere on it. And we were right. Very few people called us on it

In fact, some of our competitors found the solution elegant, and adopted it themselves. Nowadays virtually no tri bike maker today offers a dimensional sizing scheme where the dimensions match the size name. In other words, you may be looking in vain for a 57cm bike that measures 57cm anywhere on it.

Alternatively, there are companies that have opted for T-shirt sizing, but that's not entirely useful either. What does "Large" mean in terms of bike size?

That's why a half-decade ago I whined for a system that simply measured the bike's top tube and head tube. But with the latter measure, you've got to take into consideration the bottom bracket placement and the wheel size. In other words, if the bottom bracket is placed 7cm below the centerline of the bike (a line drawn through the two wheel axles), and another bike has 5.5cm of drop, then these two bikes will fit differently even if they have the same head tube length. The rider will have a more aggressive position on the bike with the 5.5cm of drop, all other things equal, because
what matters is the head tube's top relative to the bottom bracket. And, of course, a 650c-wheeled bike is going to have a much longer head tube than a 700c bike with the same basic geometry. So simply measuring a bike's head tube length leaves questions unanswered.

All this came home to roost when I was compiling the manuals and curriculum for our first-ever F.I.S.T. tri bike fit workshops. There are three fundamental tasks of a bike fitter. First is to determine what sort of geometry a subject needs—simply put, ought the subject be fitted more along the lines of a road bike, or is a more aggressive, tri-specific position appropriate?

The second task is to find those points in space that determine where the rider's weight ought to rest. Forgetting front-end measures for a moment (clip-on length, armrest width, base bar tilt, stuff like that) a subject's proper bike geometry and size is determined by where his weight will rest in space. That weight is displaced over three spots, the bottom bracket, the saddle, and the armrests. You place your subject on a fit simulator, like the one at right. You move the saddle up and down, fore and aft, same thing with the adjustable stem, and presto, you've got your subject positioned. Once you know where the saddle and armrests ought to be in relation to the BB, the position's dialed. You're done.

Yes, you're done in theory, but you've only generated a determination of where the rider ought to be "in space." To be sure, this is the most important part of the fit. But what bikes are made in production that conform to these points in space? What custom geometries flow from these points in space? This was fairly easy in the old days because, as I said, everything was very standard. But not now.

In 2003 I proposed an entirely new nomenclature for bike sizing, and in the intervening period a lot of bike makers have started to provide these metrics (we're compiling a geometric database that includes these measures). It's built around the idea that there are two parameters that really, truly, determine a frame's length and its height: how high a frame's head tube top sits above the bottom bracket and how far the head tube top sits in front of the bottom bracket. These measures are what we call stack and reach.

CHAPTER TWO

Let us consider two tri-geometry bikes, made by competing manufacturers. Both come in 700c in 58cm. Both bikes have 76-degree seat angles. Each of these bikes has a top tube measurement of between 58cm and 58.5cm. Carbon copies, right?

In actuality, if you owned one of these bikes and moved the parts over to the other's frame, you'd be in for a big surprise. One has a head tube length of 13cm and it's built to accept an integrated headset, which means most of the headset hides inside the head tube. The other has a 17.5cm head tube, and does not use an integrated headset. The significance of the headset is that a standard headset adds about 2cm of stack above the head tube. Therefore, your stem is going to sit about 7cm higher on one than on the other.

And this is not a hypothetical construct, these are two bikes that until recently were offered for sale in these geometries. Since they both measured 58cm, with almost identical top tubes and seat angles, only a closer inspection would've uncovered how incredibly differently these bikes fit. But, if you knew the stack and reach of these two bike frames you'd be able to tell precisely what the differences were and be prepared for them.

Simply as a function of it, the seat angle makes absolutely no difference in any of this. Let's take Trek's Team Time Trial bike as an example. Its virtue is in its very low front end. It's head tube measures 11.3cm in its large size. But it's got a very slack seat angle of 73.7 degrees. The nose of my saddle needs to be straight over the bottom bracket on my tri bike. At 73.7 degrees, and at 80cm of seat height, my saddle is going to be roughly 7cm behind the BB. Obviously, I can't ride this bike in triathlons. Or can I?

If I find the right set-back seat post, and flip it around forward, and move the saddle forward on its rails, perhaps I can find this 7cm for forward saddle movement. All I really care about is, once it's there, is the bike going to fit me from the BB forward? It's certainly low enough for me, but is it long enough?

In fact, it won't work for me very easily. While it's low enough -- it's got a small enough stack -- it hasn't got enough reach. You might find this funny, since it's got a 57.7cm top tube length and my own tri bike only has 56.0cm of top tube. But in fact the reach of this bike is quite small, because its top tube starts so far behind the BB. If you normalized the seat angle at 78 degrees -- that is, if you draw an imaginary line through the BB at a 78 degree angle, and measured the to tube length of the Trek at the point it intersected with this imaginary line -- that 57.7cm top tube might only be 52cm or 53cm. It was this problem that caused Trek to revamp the design, and come out with the much better designed Equinox TTX.

Accordingly, seat angle is immensely important to the question of your preferred rider position, but it is theoretically immaterial to the question of bike fit, because you can move the saddle all forward or backward to find your saddle's desired "point in space" relative to the bottom bracket. What you cannot change is the head tube top's orientation to the bottom bracket. Or, to put it another way, the way you add to a frame's stack is to add headset spacers. The way you add to a frame's reach is to put on a longer and longer stem. But you also change the frame's structural integrity when you do so, and you also change its handling characteristics. So if you want a frame to handle properly, you'll start with one that has a stack and reach that's as close to appropriate as possible.

But let's go back to seat angle for a moment. No, this is not an important issue on paper, and it hasn't been so until recently. The reason it's important now is that most desirable tri bikes have proprietary aerodynamic seat tube and post complexes, that is, you can't change out the stock OEM seat post. So the seat angle is now fairly important, in that your bike's saddle might not be able to adjust to its proper "point in space" unless the frame is fairly close to the seat angle you need.

In the adjacent diagram, you can see that your stack and reach is not going to change one iota regardless of what the seat angle is. Neither is the chainstay, nor is the front/center. If you've got a slacker angle, no big deal. A forward seat post will put your saddle in the same position. Two frames can have different seat angles and have the same stack and reach. Each will have exactly the same front-to-rear weight dispacement, as long as the chain stay and front/center are the same. They'll both handle the same, as long as the steering geometry is the same. But the seat angle becomes important today, with today's tri bikes, because you can't simply put a forward seat post on the bike -- the bikes today come with posts you can't swap out.

CHAPTER THREE

There are a couple of ways to remake sizing nomenclature so that it means something. What I'm presenting here is "stack" and "reach." Another way to get to the goal is via a method constructed by Enduro Sport in Toronto, a shop that is further down the road in tri bike fit than most in North America.

Enduro Sport has catalogued all the tri-geometry bikes it offers in terms of head tube length and top tube. But they've gone one step further. They've "normalized" all the seat angles at 78 degrees. In other words, let's say one bike has a 76-degree seat angle and a 58cm top tube in its 58cm size, and another bike that size has a 80-degree seat angle and a 55.5cm top tube in that size. Let's further say that both bikes have 13cm head tubes. In Enduro Sport's nomenclature, both these bikes might be the same for purposes of fit. Both might have 56.75cm top tube lengths when "normalized" to 78-degrees. Here's another way to look at reach: it's the bike's top tube length with the seat angle normalized at 90 degrees.

Back to Enduro Sports' sizing nomenclature: you might object with, "These bikes aren't the same, they have different top tube lengths." But they really don't. What is fixed in stone is whatever your proper position is relative to the BB. You can't be led by the geometry of the bike. You have to determine where you need to be "in space" and then find the right bike to place underneath you. Let's say your correct seat angle happens to be 78 degrees. If you buy the 76-degree frame you'll move the saddle forward on the rails, and/or change out the set-back seat post to a zero-offset post. In so doing you've effectively shortened the top tube length. Likewise, you'd have to move the saddle back on the rails of a set-back post on the 80-degree bike to get to 78 degrees, and you artificially lengthened the top tube. Therefore, in both cases you've really come up with the same bike, from the point of view of fit.

That's the elegance of Enduro Sport's approach, and the approach of any process that reduces the measures of any and all bikes to a sort of "Esperanto" that allows the entire industry to speak one meaningful language. Stack and reach is precisely that.

Why not simply use Enduro Sports' sizing scheme?
This nomenclature would work fine for me, but for a couple of things. First, if you are just considering head tube length you are omitting two parameters, bottom bracket height and wheel size. You must know the wheel size before you can use the head tube dimension properly. Another way to view it is that you must also, perhaps, "normalize" the head tube to 700c (or to 650c if you prefer).

So, by using stack and reach you don't need to normalize anything. The information about the bike's size you need and don't have with stack and reach is whether the headset is integrated. If it is not, you must account for the head parts that stick up above the head tube top.