Speed typing test for programmers

Question

It's pretty basic knowledge that as a programmer, our output is channelled through a keyboard upon which we must type. If we type slower than we think, then it's a bottleneck for our output. Steve Yegge and our beloved dictator have both blogged about this in recent months.

I challenged my girlfriend to a friendly game of Typing Of The Dead, and found my ass sorely kicked. I'm only able to do about 70 WPM to her 95, but I did notice that on "special keys" (ie: punctuation), I was much better. I then started paying attention to how she was typing - correctly, with her fingers on the home row - compared to me: with my right hand angled so my little finger always rests on the shift key. This is obviously because as a programmer, I'm constantly typing a large assortment of punctuation marks { } ( ) [ ] ? < > / $ % # * &, most of which require shift.

Getting to my point now, if you were interested in their typing speed (for reasons outlined in the blogs above), how valid would it be then to ask a potential recruit their WPM, given that normal typing tests examine your proficiency at typing in English rather than in code? If you believe it's not appropriate to use standard typing tests for this, do you know of any method to find their coding-typing speed?

Edit: I know that a programmer's typing speed is waaay down the list of Things Which Are Important. Obviously no one's ever going to hire a crappy programmer just because they can type fast. It's a theoretical, ok?

Answer 1

Although typing speed is important. Coding is more about design and implementation that entering code blindingly fast.

Answer 2

Getting to my point now, how valid would it be then to ask a potential recruit their typing speed...?

Not at all valid.

If you believe it's not appropriate to use standard typing tests for this, do you know of any method to find their coding-typing speed?

I know of no typing speed test that would be relevant to a software development job.

I do know how to touch-type. I was the second-fastest person in my typing class in high school, and I've been touch-typing for 25 years. But if you asked me any such question in an interview, I would not consider taking the job.

Answer 3

In Typespeed you may choose to try your speed with DOS/Unix commands or programming functions. I assume there must be even better typing speed tests for programmers.

Answer 4

Keyboard? How quaint. (Scotty, Star Trek IV) ..

Sorry, this was obligatory.

Edit:

Changed to community wiki, I did not expect this to get voted up and it is a non answer.

Answer 5

Getting to my point now, how valid would it be then to ask a potential recruit their typing speed...?

What do you want? People genearating random code really fast, or people generating quality code

Answer 6

I agree with whats been said before. Typing in the code is not the bottleneck for programming unless you're writing an application almost exactly like one you've written in the past and the only way you know how to type is hunt and peck.

In some programming methodologies, they put two programmers to a single computer. Two sets of eyes and two brains will be more productive when it comes to the real roadblocks that take up a majority of a program's development.

Answer 7

I really don't see the relevance either. Maybe if I had 2 applicants that seemed equally qualified, but I bet you'd get equally valid results by flipping a coin. In my experience, typing speed really doesn't significantly improve productitivity, a slow typist that can write 10 lines of solid code I feel would be a lot more productive than one that could burn up the keyboard yet wrote 1000 line of buggy, impossible to maintain code.

Personally, I'd probably walk out of the room feeling rather offended, I'm a software developer not a clerk/typist. If a company feels typing speed is a significant enough measure of programming skills to test for it, is this someplace I really would want to work? No, I'm not a fast typist, even though one of my secretaries told me I was the fastest and most accurate two finger typist she'd ever seen.

Answer 8

I'd say stopping and thinking about what you are coding yields a lot of time saved afterwards in a good design and less bugs. But I type fast anyways although I'm getting rsi on my pinky from shift-ctrl-shift-ctrl pivoting all the time.

Answer 9

Typing speed is probably less important now that it ever was.

Unless you're hardcore and code in an IDE or text editor with no intellisense (or the non-copyrighted word for it), you're probably not typing a large chunk of it anyway.

I'm a C# developer using VS 2008 and I know personally I use a hell of a lot of shortcuts. I probably hit tab more than anything else when I'm writing code.

To answer the question though, I think that you're probably wasting your time demanding a typing speed test of a potential recruit. It's really not that important in the scheme of things.

Answer 10

Despite other answers in this post, I don't think fast typing necessarily equals buggy, sloppy code. I find the faster i type, the more I enjoy programming as I can more quickly materialize my ideas in reality.

To all those pointing out typing is not relevant these days, here is quote I like from the Yegge post:

If you really think refactoring tools are a substitute for typing, it's like you're telling us that it's OK for you to saw your legs off because you have a car. We're not fucking buying it.

But to answer the questions; No, I don't think you should ask a potential recruit for a WPM score on either english or programming text. If you really care how productive they are, ask them to write some code to solve a problem.

If you just want to beat your girlfriend at typing speed and accuracy, perhaps you should play round two with some random code in your language of choice. Most typing tests let you use arbitrary text.

Good luck.

Answer 11

programming != data entry

I can understand this would be a valid test/concern for a position which involved straight data entry (where speed & error rate are serious and valid KPIs), but I seriously can't see this being an important factor in the hiring process for a programmer or software engineer. If it is.. perhaps you don't want to work at that company anyway?

Answer 12

Software Engineering Facts and Fallacies:

When you combine this fact:

Fact 22. 80% of software work is intellectual. A fair amount of it is creative. Little of it is clerical.

With this one:

Fact 2. The best programmers are up to 28 times better than the worst programmers.

You see how typing speed is of such little importance.

Intellisense and refactoring tools decrease the difference even further.

Answer 13

nickf is absolutely on the money to ask his question, which was totally ignored.

It's amazing to watch coders at conferences make errors in every other word as they type before a group; it's not that they're nervous...these are people who do the conference presentations every other day. Just kick back and watch the cursor go forward five characters, back two, correction, forward 8, back 3, correction; rinse and repeat. All this while they babble about their topic.

Coders are, lots of times, lazy spastics. Much of the time spent correcting code is actually correcting typos.

The emphasis of this discussion was on typing speed. Everyone's right that typing speed doesn't matter [excluding the retarded two finger typists, of course, who can't be bothered to learn to type, or claim to be 'dyslexic' which is a comprehension problem and not a typing problem]. However, learning to type well is just about speed...it's about accuracy. As regards accuracy, I'm not going to bother even explaining the importance of that.

Nickf...just search on stenograph* test* and you'll be in. And absolutely...test for typing speed. Do not, for sure, listen to the legions of coders who defend their ridiculously lazy approach. There's a scarcity of coders now, so they set the rules. Once web based tools wipe out 60 percent of the market (which is mostly very basic products like small business pages, etc.) skill and speed will come to bear.

Answer 14

28 times better...I believe it, actually. Where is this stat from?