10x Programmers
Article by Peter Blain, CIO CoderCred
It’s often asked whether 10x programmers really exist. These are the guys who supposedly write code ten times faster than the average programmer. I recently answered a question on Quora about this and got a lot of up-votes, so it seems worthy of elaboration here.
The short answer is yes, 10x programmers exist. They are like chess grandmasters who can play ten or more local chess club players simultaneously. These local players are no chumps, or they wouldn’t belong to a chess club. They arrange their boards in a circle surrounding the grandmaster. The grandmaster walks from board to board spending only a few seconds on each game, while the local players have as long as it takes the grandmaster to walk around the entire circle before making their move. Typically the grandmaster beats them all. The 10x programmer is similar in that he or she needs very little time to comprehend the existing code and come up with a sound course of action. The average programmer needs more head scratching time and often a dose of trial-and-error.
I’ve actually seen better than 10x. I managed a team of developers at a company that subcontracted one of its projects to a large and well respected external organisation. The external provider allocated one manager and a couple of programmers at a cost of about two million dollars over 3 years, but failed to deliver a satisfactory product. The project was still dragging on when I hired a new developer. The new guy wasn’t allocated to the beleaguered project but he informally asked me a few questions about it in passing. I drew a rough depiction of the problem on a white board, and a few days later the new programmer came to me with a working solution that is now used in production. He created his working prototype in just half a day — something that the external organisation hadn’t been able to achieve in three years! This is more like 4000x. The developer’s solution relied on distilling the problem to it’s essence, stripping the superfluous, and stringing together some open source products. This is an extreme case and you could argue that his productivity was less about programming and more about architecture, but I think this gives insight into one of the ways a 10x programmer achieves the speed. If you can understand the problem and see the simplest solution, sometimes there’s not much programming to be done. In other, words the 10x programmer may, in some cases, be eliminating 90% of the task. Other times it’s more about deep insight into the framework or programming language. Either way, these guys definitely exist.
Written by
Peter Blain
For more articles from Peter and the CoderCred team, visit here.
