Thursday, September 12, 2013

A bicycle's trigger shifters collide with user-centered design

I work in information technology for a living, and I just bought a new bike. A hybrid. I mostly ride to get around town, mainly to commute to work and my gym.

I loved my old bike, vintage 1984, a much more rugged conveyance (heavy, fat and knobby tires) that came to me as a hand-me-down (hand-me-up?) from my younger brother. I've been riding it for more than twenty years. I replaced it only because the frame broke in two places, where the right-side chain and seat stays meets (met) the rear axle ... it must have been metal fatigue, 'cuz I rode the thing like a little old lady.

I've had my new bike less than a week, but I'm already smitten (plug: Mike's Bikes!). I could tell you all about it, but this very geeky post is going to focus on shifting gears. Bike gears, I mean. If you're not a bike geek or an IT geek -- or, minimally, some kind of geek -- I can't guarantee you're going to fathom why I'm spewing so many words on so focused a topic.

So. Shifting gears.

My new bike has index shifters, which is pretty much standard equipment these days. On my bike, the indexed shifters are operated by triggers -- little levers mounted on the handlebars -- that a rider pushes or pulls, hence they're called "trigger shifters."

The old-style friction shifters required a rider to move a lever just the right amount to cause the derailleur to align with the next sprocket up or down the bike's front crankset or rear cassette. Friction shifters often require a bit of post-shift adjustment to get the derailleur (and the chain whose position it governs) into just-right alignment with the desired sprocket otherwise the drive train rattles and chatters, and the chain might even slip, because it's not properly engaged. You might think of this as analog shifting.

The nifty, new index shifters are a good conceptual fit to this digital age. Index shifters are calibrated to pull in or let out a small, fixed length of cable each time the shifter is activated. These fixed lengths of cable move the derailleur over just the right distance, into just-right alignment every time. No further fiddling is necessary (or even possible, except by adjusting the cable tension: a tune-up procedure, not something a rider does while on his or her merry way).

The nifty, new index shifters that came with my bike are made by Shimano, which is the same high-quality manufacturer of my old bike's friction shifters. The model of the shifters on the new bike happens to be the Shimano EF-51; I believe that "EF" is an abbreviation for "EZ Fire."

So Shimano EF-51 trigger shifters have two triggers, which the rider pushes or pulls to shift ... that is, to move the derailleur, and thus cause the chain to move over to the next sprocket.

One trigger (activated by pushing from the rider-side of the shifter, generally with a thumb) moves a derailleur in one direction; the other, smaller trigger (activated by pulling from the forward side of the bike back toward the rider, generally with an index finger) moves a derailleur in the other direction. A push moves the derailleur so that the chain moves to the next larger sprocket. A pull moves the derailleur so that the chain moves to the next smaller sprocket.

If you're with me so far, you have a pretty good understanding of how bicycles work. Which would come in handy if you were to ride my new hybrid bike with nifty, new index shifters, because, IMHO, the way they work is counterintuitive, as we like to say in the world of "user-centered design" as applied to software, and especially to the web site design world. These worlds are where I make a living.

Counterintuitive? How so?

Here's a key quality of multi-geared bicycling that you want to keep in mind to understand what I mean:

On the crankset (the pedal end of the chain), using a bigger gear makes the bike harder to pedal than a smaller gear. On the rear cassette (rear wheel end of the chain), it's just the opposite: using a bigger gear makes the bike easier to pedal than a smaller gear.

[Another way of thinking about harder and easier in terms of bicycle mechanics is that the rider has to exert more force (pedal harder) when the rear wheel rotates further for each rotation of the crankset (= rotation of the pedals); the rear wheel rotates more times per rotation of the crankset when a larger sprocket is engaged on the crankset and as a smaller gear is engaged on the rear wheel. My apologies if that's just not helpful at all...]

So what all the verbiage boils down to is this:

For a rider, pushing the near/inside trigger on the right side shifter, controlling which sprocket is engaged on the rear cassette, makes the bike easier to pedal. And performing the same action on the left side shifter -- pushing the trigger that controls which sprocket is engaged on the crankset -- makes the bike harder to pedal.

Same rider motion (push the trigger), opposite experience for the rider (easier vs. harder to pedal).

This is not good, user-centered design. It's counterintuitive. Assuming I don't know anything about how bikes work, I'll expect the same effect (harder or easier) if I push the trigger on one side or the other of the handlebars.

Yes, of course I can get used to the peculiar way the bike's shifters actually work, but if user-centered design principals were employed in making these devices I wouldn't have to.

That's my point.

Is design that rides roughshod (as it were) over a bicyclist's expected experience actually necessary due to mechanical constraints? Well, keep reading, we'll get to that.

First let's take a closer look at what "user-centered design" means. Here's Wikipedia's definition, with my emphasis added in bold:
In broad terms, user-centered design (UCD) is a type of user interface design and a process in which the needs, wants, and limitations of end users of a product are given extensive attention at each stage of the design process. User-centered design can be characterized as a multi-stage problem solving process that not only requires designers to analyse and foresee how users are likely to use a product, but also to test the validity of their assumptions with regard to user behaviour in real world tests with actual users. Such testing is necessary as it is often very difficult for the designers of a product to understand intuitively what a first-time user of their design experiences, and what each user's learning curve may look like.

The chief difference from other product design philosophies is that user-centered design tries to optimize the product around how users can, want, or need to use the product, rather than forcing the users to change their behavior to accommodate the product.
(For a definition tailored to web site user interfaces, you can check out the definition of usability on the Neilsen Norman Group's site; NNG is usability guru Jakob Neilson's firm.)

So all this cerebration came to me on the very first day I rode my new, nifty bike to work, happily pushing and pulling shifter triggers willy-nilly to put the machine through its paces. Nice ride ... really, I enjoyed going to work. Imagine!

And yet: I wasn't the first person ever to notice the counterintuitive design of trigger shifters. The oddity I noticed has been discussed for years -- for example, in a May 2010 posting, excerpted below:
Jim, 05-21-10, 02:08 PM

This shifter is on my new 2010 Electra Townie 21D. I love it's simplicity, but get frustrated that the shifters function opposite of each other (ie. on the right shifter the upper lever is to upshift and the lower lever is to downshift - but on the left shifter the upper lever is to downshift and the lower lever is to upshift). I would have preferred the left to match the right. [...]

Wanderer, 05-21-10, 03:13 PM

Actually, they are both the same - clicking the shifters, in the same manner, moves the chain the same direction - toward the smaller sprocket, or toward the larger sprocket - in both instances....... start thinking in "sprockets" and all will be well.

Jim, 05-21-10, 03:33 PM

Yes, but using a larger sprocket on the rear is gearing down, and using a larger sprocket on the front is gearing up. Therein lies my conundrum.

mtnroadie, 05-23-10, 05:01 AM

It takes a little more force to go up to a larger sprocket. This is why the thumb lever is used on both the front and rear sprockets when shifting to a larger sprocket. Just ride it for a while. It will become second nature in no time.

Jim, 05-23-10, 04:51 PM

That's exactly what I assumed, after putting some thought into it. I spend more time going up and down the rear gears anyway. I'm sure I'll get used to it eventually. It's just kinda like when you get a new car and someone puts Reverse, down and to the right instead of, up and to the left where it has always been. Ugh.
To the question whether my shifters needed to be designed the way they are for mechanical reasons, "mtnroadie" claims that "It takes a little more force to go up to a larger sprocket. This is why the thumb lever is used on both the front and rear sprockets when shifting to a larger sprocket." Maybe. But I'm skeptical. Seems to me that's a problem that engineers have been solving with torque and appropriately lengthened levers for a very long time.

And I'm also thinking Jim shouldn't have caved so easily. It's not "kinda like when you get a new car." There's not an intuitively consistent place where automotive engineers put the reverse gearshifter position (though there may be longstanding customs observed by some or many manufacturers).

What's counterintuitive about the trigger shifters Jim and I find odd is that performing the same action on different sides of the same handlebars of a single bike produces different results -- from a rider's (user's) perspective.

Like I said: I'll get used to the shifters on my new bike just the way they are. The fact that there are nifty little numbers that show up in a nifty little window to show which sprocket is engaged means I won't be tempted (as I was on my old bike) to look down, away from the lunatic drivers careening-while-texting mere inches from my frail and vulnerable flesh, so I can watch which way the chain is moving. That's seriously nifty!

But it would be niftier still if a next generation of trigger shifter engineers took the experience of naive bike riders into account, and didn't design principally for gearheads.

I'd love to hear from the geeks who made it all the way to the bottom of this post in the comments below!! No spamvertisements, pleeze.

Related posts on One Finger Typing:
Bike parking fail
Sharrows and stripes: bike lanes for a common good
Fixing flat tires

Thanks to Keithonearth via Wikimedia Commons for the image of a bicycle drivetrain (adapted by the author of this post by adding a plain white background and converting to JPG format; anyone is welcome to use the adapted image under the terms of the original creator's license, Creative Commons Attribution-Share Alike 3.0 Unported).


  1. My current bicycle has cylindrical shifters at the inside end of each hand grip. Similar to the trigger shifters, if I want to make it easier to pedal I can rotate either shifter, but they do need to be rotated in opposite directions from each other to achieve the similar results. I did quickly adapt to this and it seems now to make total sense to me. If I'm going up a hill and want to downshift I turn my hands in opposite directions. It's as if I had a bottle in one hand and a twist cap in the other and I were tightening the cap onto the bottle. So I associate this with tightening which feels right for pedaling more and going up a hill. If I'm going down hill and want to upshift then I rotate my hands in opposite direction in a 'loosening' motion, which makes sense because going fast feels more open.

    I think one clue as to why the manufacturers set up the derailleurs in this 'counter-intuitive' way has to do with the spring. Both the front and back derailleurs have springs which constantly pull the derailleurs toward the smaller gears. With trigger shifters, when you use the thumb lever you move the lever in a pretty long sweep because you are physically using your thumb to work against the spring and move the chain over to the next bigger gear. When you use the small lever it's just a short movement because you're really just releasing the derailleur such that the spring can move the chain over to the next smaller gear. This is true for both the front and back derailleurs.

    I can see that it makes sense to always have the thumb working against the spring, but I don't see why the spring would always have to be set up to move the chain only toward smaller gears. Since the derailleur only moves a chain laterally it doesn't seem like moving to larger or smaller gears would make a difference in strength needed.

    At any rate, for my cylindrical rotating shifters it really doesn't feel easier or harder for a hand to move the shifter in one direction or the other.

    I am curious if the lever shifter of the old style actually had this same issue. Did one go up for easier pedaling and the other down? I can't remember, but I suspect that they were counter-intuitive in exactly this same way...

  2. Great article! I agree. I am finding it difficult to remember which is which while waiting for the muscle memory to figure it out. I feel I want to thumb push to make it easier to peddle and finger pull to make it harder.

  3. Great article! I agree. I am finding it difficult to remember which is which while waiting for the muscle memory to figure it out. I feel I want to thumb push to make it easier to peddle and finger pull to make it harder.

    1. Thanks for the note, Jason... I've found that muscle memory does kick in, though imperfectly for me ... I'm still riding my now-two-year-old bike, and *mostly* push/pull the correct lever -- but every once in a while I still get it wrong.