Link Details

Well they both make valid points but experience DOES matter

You do what you want with your time. How can I measure what you did? What implies that more time equals more knowledge and vice-versa? What about those who can learn in less time - aren't they going to be rewarded? This is against merit.

The people who read Dzone regularly are not the best group to sample, but even the author acknowledges that "experience and skill in programming may not be directly correlated". There's no "may not", they /are not/ highly correlated. That doesn't mean that outliers don't exist. My gut feel is that about 2% of developers actually do benefit from experience in the way the author describes. To focus on years of experience would lead to making decisions based on the 98% case where this is not relevant. A big mistake. In any case, it's what you know that matters rather than years of experience. When someone emphasizes years of experience (except as an interview ploy for those who value it), it makes me wonder how much they really know.

The lack of an obvious correlation is not due to the myth that experience doesn't matter, but rather due to the fact that some people start at different levels than others and some people gain more from experience than others. Those two factors are why its perfectly possible for some young developers to be better than some of their more experienced counterparts. It certainly does not follow that experience doesn't matter, in fact it should be clear to anyone with half a brain that most developers benefit from experience. That doesn't mean experience should be the most important thing when considering a candidate for a job; as I said its perfectly possible for a younger developer to be a better candidate than an older one. It just means that younger developer should get better as he (or she) gets more experience.

And no, I'm not saying this because I'm some old fart who wants to keep alive the notion that I'm better than my younger peers. In fact I've only been in this industry for 3 years now. I just recognize that I still have plenty of room for improvement.

As someone with quite a bit more experience than you, I predict that when you gain the additional experience you'll agree that although someone "should" get better with experience, the relationship between skill and experience is often (but not always!) very sublinear.

No one here is claiming that experience does not matter in any absolute sense. That's a strawman argument. What I am saying is that if you have two job candidates with skills in a particular technology and one has 8 years of experience and the other has 4 years of experience, you really know nothing about the actual skill levels of those developers.

>>As someone with quite a bit more experience than you, I predict that when you gain the additional experience you'll agree that although someone "should" get better with experience, the relationship between skill and experience is often (but not always!) very sublinear.
I never claimed that the relationship was linear. Obviously it isn't. Obviously you will usually learn a lot more when you are first starting, and that goes for virtually any subject. So you arguing against the idea that the relationship is linear is a perfect example of the strawman fallacy you erroneously tried to label my argument.

>>No one here is claiming that experience does not matter in any absolute sense.
He may not be "here", but Jeff Atwood (whose blog entry provoked this one) certainly was arguing that when he said "It's been shown time and time again that there is no correlation between years of experience and skill in programming". Yes, he then contradicts himself in the final paragraph, but all that proves is that he didn't think through his argument very well before he started writing.

"Time and time again" is not the same as "always". It means that it's often or frequently true. There's no contradiction in his final statement unless you view it in black-and-white terms (which is the essence of the straw man I mentioned earlier).