Stop Spreading the 10x Productivity Myth and Focus on Improving
Developers should focus on improving instead on the difference in productivity
Have you come across an article that states that there is a 10:1 difference in productivity between the best and average developers? Did you, after reading the article, asked yourself where you stand on that scale?
I did and didn’t know the answer.
I have been developing professional IT solutions for over 20 years. During that time, I have never noticed such a difference in productivity.
But is it important? Isn’t it better to focus on improving instead of comparing yourself against others?
In this article, I will dive into the origin of this productivity myth and show you why this should be considered a myth.
What is the origin of the myth?
Let’s look at the original study “Sackman et a. 1968” that is generally considered the original study, which supported the 10x claim. The conclusion of that study is:
Statistically significant results of both experiments indicated faster debugging under online conditions, but perhaps the most important practical finding involves the striking individual differences in programmer performance.
Ok, so the study concludes that there is a striking difference in the performance of programmers. I took the table below from the original research. The table shows the results of the experiments held by Sachman.
Many articles, use the 28:1 difference in productivity from the second row in table 1, “Debug hours Algebra”.
But what the report does not tell is how Sackman conducted these experiments.
Thomas Dicky, 13 years later exposes the myth
Thirteen years later, Thomas Dickey researched the original data from the Sackman study. He wrote an article about the study in Proceedings of the IEEE (Volume: 69, Issue: 7).
He found that the original 28:1 ratio was observed because “subject 7 required 170 hours to program the ‘algebra’ program in a batch environment, in machine language (while) subject 3 required 6 hours to program the same problem in JTS (ALGOL) in a time-shared environment”.
Sackman compared the productivity of a programmer using ALGOL, a high-level programming language against another programmer using machine language. Besides the difference in the programming language, they were also working on entirely different environments, batch-oriented vs. a time-shared environment.
You can’t compare those! This is comparing apples to oranges.
At the end of the report, Thomas Dicky concludes that
“Sackman’s stats didn’t checkout”.
There are other research studies that point out other flaws.
If you want to know more, I can recommend the Leprechauns of Software Engineering from Laurent Bossavit. In his book, Laurent spends two chapters discussing many studies. He points out many flaws in the studies.
So, Is there no difference between the productivity of developers?
Sure, you know, like me, that there is. But the importance and scale have been exaggerated.
Based on the findings in the studies, I conclude that there is absolutely no empirical evidence that supports the 10x productivity claim. I think that these kinds of statements are misleading.
We have to stop selling this myth to other programmers. Don’t believe that there is something magical that you can do or learn to become ten times more productive.
Instead of focussing on the difference, we have to ask how we can improve as a developer. See my article on improving as a developer.
Thank you for reading.
My name is Patrick Kalkman. I work at Hoogendoorn, where we build all-in-one process computers that deliver integrated solutions for climate-, water- and energy management.
