Where are your hands are placed while riding? If you answered, "Relative to what?" I'd say that's a wise retort.
In this case, relative to the head tube top, or top bearing of the headset. Let's "place" your hands as compass points around the fulcrum of the steering axis (typically in front of the head tube top, maybe northeast and northwest; maybe directly to the side, as in east and west; maybe even behind the head tube top).
I'm not writing about aerobars at the moment. That's a separate topic, different dynamic, that comes later. Today it's steering and handling when the weight of your upper body is supported by your hands, rather than your elbows in the armrests.
I'm talking about every kind of bike except for aero (tri) bikes and recumbents: road, track, cross, MTB, just about everything else. In fact, I'm also talking about tri bikes, but only when your hands are in the pursuits.
While I have been making bikes; writing about bikes; riding bikes; thinking critically about geometries and fitting and handling for a long time, there's an awful lot about what makes bikes work the way they do that I don't know. In fact, for every new bit of knowledge I think I acquire I'll bet there's something else I assumed was the case that turns out to be a false assumption.
My approach here is simply observational, like John Muir when he wrote about natural history and geology. He was classically trained in neither, yet his observations, which birthed views charged with striking common sense, changed the way we now think about glaciers, and trees.
Whether my sense is common (let alone striking), or makes only rare appearances, is a matter of opinion! Nevertheless...
In simple terms, a steering "lever", in my parlance, is the distance between the steering axis (a line through the middle of the fork steer column), and the point where the turn is actuated by the hands. In general, the longer a lever the less force it takes to actuate a turn, but the less effect there is to a given movement.
Imagine a huge steering wheel in your car. If you had no power steering the huge steering wheel would help you turn the wheels with less force applied. But if you need to make a u-turn your hands would be very busy turning that large-diameter wheel! Less strength is required to turn the large wheel, but a point on that steering wheel would travel a long distance before the u-turn is successfully made.
It's not quite that simple on a bike. Two wheels support you rather than 4, and this dictates a number of design imperatives. Your bike is "hinged" and your weight is supported both in front and behind the hinge. The front wheel moves left and right relative to the frame. The hinge consists of the headset bearings that keep the fork attached to the frame. Most of your weight is behind that hinge, sitting on the saddle or standing on the pedals. Some of your weight is in front of that hinge, because your hands rest on the handlebars (although, as we shall see, there might be cases when all your weight sits behind that hinge).
Things like fork offset, head angle, trail, wheel flop, front-center, all are geometrics that live inside pretty close allowable ranges. Not much wiggle room there. Otherwise, there's expansive wiggle room in how you can set a bike up for riding.
Imagine the bike cut in half three different ways, with lines passing through the top of the head tube. Just like splitting the body into the sagittal, coronal and transverse planes we could split the bike into 3 planes, with each plane passing through the head tube top.
I can imagine a road bike with a 100mm stem, and 40mm wide road handlebars. Imagine riding that bike, with your hands on the hoods. Blue sky, white puffy clouds, no wind, 75 degrees. Nice. You might have 32cm of distance between the steer column and the center of your hands, depending on the point on the steerer we choose to measure to (maybe the head tube top, or it could be a line perpendicular to the steerer out to the hand position).
If it were as simple as just that "lever" being longer or shorter around the fulcrum of the circle, like the steering wheel of a car, this would be easy to describe and calculate. It would be as easy as my description of a car making a u-turn.
But what if your hands were further in front of you, but only 21cm apart instead of 42cm apart? What if both hands, instead of northeast and northwest of the head tube top, were pointed almost straight north? You'd still have that 32cm steering lever in this example. The steering lever might be exactly the same, but you'd have way, way less control! It's clear that it's not just the case of a steering lever of a certain length, but that the distance away from the centerline splitting the bike between its left and right halves (its sagittal plane) is also meaningful.
Let's again cut the bike into two parts, passing a plane through the head tube top, this time splitting the bike into front and back. What happens if we move the stem far in front, or instead back toward the plane, or even behind the plane. We might do this with a stem that's long or short or even turned around facing the wrong way. Or, with swept back handlebars on a cruiser bike.
When the hand position sits behind the steering axis it doesn't matter how the hand-hold got there: whether by a stem facing backward or via a front end of a bike like the Felt Highland Park above. There are all kinds of associated problems with a kooky front end like a stem facing backward, but most of them are clearance problems. Knee clearance, stuff like that. Otherwise, the bike might not handle in an ideal fashion, but it won't be untenable. It might even be just fine, as in the case of that sweet cruiser above.
Maybe there's some magic in the hand position being a given distance away from simply east and west of the head tube top. Maybe for cruisers like the one above what's optimal is a given distance back of that point, just as in road bikes it's a given distance in front of that point. Maybe it makes no difference at all in steering per se, it's just that the hand position informs body position, which informs weight displacement, especially when out of the saddle. I don't know.
Let's cut the bike in half again, this time creating upper and lower halves, bisected by a plane (transverse) passing horizontally, parallel to the ground, and passing through the head tube top. What if the hands are well above that point? Or below that point? What difference does it make?
In my opinion and experience the only real steering problem with the hands too high above is speed wobble associated with a system (bike plus stem, spacers, handlebars, etc.) that was overly flexible and prone to a harmonic oscillation. (Of course, I think you might have a bike fit problem if your hands are high above the bike, or you might have the wrong bike for your proper bike position, but that's not the issue we're tackling here.)
There's a thread on speed wobble on the Slowtwitch Reader Forum right now, and Trek engineer Carl Matson offered the following observation:
He placed "a chunk of steel [simulating] a pair of full 28oz bottles and [other typical stuff], cantilevered out behind the saddle as part of ride evaluation. The bike went into speed wobble almost instantly when I removed one of my hands from the hoods early in the descent. The same bike, sans the simulator, has never speed wobbled with or without hand movement in any other condition at speeds in excess of 55mph."
I have experienced the same on bikes where my handlebars were too high above the frame's main triangle.
Carl theorizes that "cantilevering a bunch of mass off of a "corner" of this system of (damped) springs" is the problem and I agree. Too much weight cantilevered too far in front of the "system" and you make a very small version of the Tacoma Narrows bridge.
Does it make a difference in which direction the weight is cantilevered? Is too far "out," as in the use of a longer stem, as bad as a weight placed too high "up?" I don't think so. Not in my experience. I suspect there's a special problem with mass high above the ground rather than perched lower to the ground, in front of or behind the bike the same distance. But I don't know.
Hands too low, this causes me a particular annoyance, and it's especially the case with tri bikes, especially if ridden steep, especially if ridden with the hand positions well in front of the steering axis.
The best example I can think of was the old Vision Tech Supermax one-piece aerobar from 15 or more years ago. This bar had a hand position very low, swept very far forward of the steering axis. It made steering cumbersome, like the front wheel was nailed to the ground. Why? I think it was all the weight on the front wheel when hands were out on the extensions. Weight in front of the steering axis informs the action of the steering.
Steep seat angles places more weight in front of the steering axis. So does a swept forward position, i.e., long steering lever forward of the steerer. Likewise a low hand position, which causes you to bend over at the waist, and perhaps move more forward yet.
Conversely, hands further back shifts your body weight back, unweighting the bars and the front wheel, giving the steering a lighter touch. Maybe in some cases too light a touch, causing the bike to oversteer when, say, climbing out of the saddle.
Summing Up
It's remarkable how rideable bikes are. How much you can bend convention. How much you can get away with. Tri bikes tend to work best with pursuit positions that are further back toward the rider, not swept forward, because there is so much weight forward of the "hinge" spoken of above. For this reason I think a number of pursuit bars have an annoying tendency to sweep too far forward for their (and our) own good.
The width of the bars is much more impactful on steering performance and the feeling of confidence than, for example, any rearward placement of the hands versus that hinge. As long as the spatial relationship between the body and the hands remains tenable – if your hands move back, your body ought to move back so that the distance remains appropriate – there doesn't seem to be a big downside.
But, as opposed to road or MTB, the pursuit position is hanging out there, unused, for the great majority of the time, so there is some urgency to make that position narrower for aerodynamic reasons.
