4 ms·

It lost me at the assertion of the 100x productivity differences between programmers. There's not a single study that shows this to be true (that I'm aware of,

It lost me at the assertion of the 100x productivity differences between programmers. There's not a single study that shows this to be true (that I'm aware of, happy to hear if anyone has one). I don't think programmers are fungible, there are differences in productivity, as you'd expect, but even the more-commonly-cited "10x" is bullshit.

Veuxdo 4y ago
> It lost me at the assertion of the 100x productivity differences between programmers. Is there not a trivial counter-example: the programmer from an awful boot camp on his first day on the job vs., say, John Carmack.
- marcus_holmes 4y ago
  > and has thus observed first-hand the well-known variance of a factor of one hundred in productivity between the most able programmers and the merely competent. from the text - it excludes the trivial counter-examples. No-one expects complete newbies to be productive, or even capable of performing the task set. The comparison is always between normal, professional programmers and those "10x" alpha-male (they're always male) exceptional individuals who break all the rules (and therefore don't have to obey them). I've worked with a few self-described "10x" programmers, and (like the few self-described "alpha male" guys I've worked with) they're always toxic assholes who don't understand how teams work. In every case the team would have worked better, faster, without them. But that's anecdata.
  - Veuxdo 4y ago
    Are there any 10x professions? There are certainly 10x products and companies (e.g. Oreo vs. Hydrox, Google vs. Lycos). They got that why by making better choices. Many more than "10" x better choices. Could this apply to developers also?
  - jmillikin 4y ago
    > The comparison is always between normal, professional programmers > and those "10x" alpha-male The commonly-cited 10x difference is between the best and worst engineers, not best and average.
    - Veuxdo 4y ago
      The worst engineers provide negative value, the metaphorical guys rowing the wrong way. Certainly some managers have experienced this.
      - jmillikin 4y ago
        The study that produced the "10x engineer" finding was not about an abstract sort of business value, it was about programming. The claim that 10x engineers exist is: Given a group of programmers who are capable of planning and executing some project, the most effective member will be approximately 10x "better" (at speed / correctness) than the least effective member. I've been on both ends of that 10x, but have never experienced a situation where all the programmers were approximately the same. There's always someone who is clearly more capable, and someone who is not.
        Nomentatus 4y ago
        Hmmm. I'm gonna agree the slow way. I take it you are not referring to lines written/hour but, lines of code that survive; or time to working code that meets set goals, and maybe ten percent of the vicious, business-killing bugs. But then again, maybe the mediocre guys spend a whole lot of time looking for code that they can cut, paste and modify. That certainly used to happen with hardware circuits. Given that most large projects don't make it to the finish line, someone who has consistently pushed projects over the finish line looks like an outlier to me. If your 10x gal (or guy) can choose a far better algorithm, that's not just 10x; that's maybe clear success vs total failure. (So you'll be dividing by zero if you want a ratio, there.) To switch to the military context, it's funny how all the people I know who don't think there are any 10x Generals also firmly believe there are minus 10x Generals. Maybe even lots of them. And can somehow maintain both those beliefs at the same time. Both history and personal experience (not as a general) tells me that true competence is difficult to achieve and that most people are far too lazy to get there, or have personalities that block them. Everybody thinks the have an above-average sense of humor and are amongst the best drivers, but they know that they aren't 10x for humor and driving. Maybe, to maintain their illusion, they have to say there are no 10x ers, cause then they don't look above average anymore.
- woooooo 4y ago
  Even then, there's about 200 workdays in the year. There are probably scenarios where I'd rather have 2 days of Carmack than a year of a newbie, but they are rare and specialized.
CapitalistCartr 4y ago
I have personally seen programmers that were only marginally useful. More than once. Being ten times more effective isn't a great achievement. Management is the same way.
tux3 4y ago
You'd expect people who are much more productive than average, if they exist, to be relatively rare. So it's understandably not necessarily easy to round up a population of them and do a comparative study. But even without a study, some people seem noticeably hyper-productive, beyond noise or luck. I like to work on personal projects, but if you tried to compare my productivity to your local Fabrice Bellard-class programmer however you want to measure it, I'll have to admit the ratio's probably pretty far from 1x.

Veuxdo 4y ago

> It lost me at the assertion of the 100x productivity differences between programmers. Is there not a trivial counter-example: the programmer from an awful boot camp on his first day on the job vs., say, John Carmack.

marcus_holmes 4y ago
> and has thus observed first-hand the well-known variance of a factor of one hundred in productivity between the most able programmers and the merely competent. from the text - it excludes the trivial counter-examples. No-one expects complete newbies to be productive, or even capable of performing the task set. The comparison is always between normal, professional programmers and those "10x" alpha-male (they're always male) exceptional individuals who break all the rules (and therefore don't have to obey them). I've worked with a few self-described "10x" programmers, and (like the few self-described "alpha male" guys I've worked with) they're always toxic assholes who don't understand how teams work. In every case the team would have worked better, faster, without them. But that's anecdata.
- Veuxdo 4y ago
  Are there any 10x professions? There are certainly 10x products and companies (e.g. Oreo vs. Hydrox, Google vs. Lycos). They got that why by making better choices. Many more than "10" x better choices. Could this apply to developers also?
- jmillikin 4y ago
  > The comparison is always between normal, professional programmers > and those "10x" alpha-male The commonly-cited 10x difference is between the best and worst engineers, not best and average.
  - Veuxdo 4y ago
    The worst engineers provide negative value, the metaphorical guys rowing the wrong way. Certainly some managers have experienced this.
    - jmillikin 4y ago
      The study that produced the "10x engineer" finding was not about an abstract sort of business value, it was about programming. The claim that 10x engineers exist is: Given a group of programmers who are capable of planning and executing some project, the most effective member will be approximately 10x "better" (at speed / correctness) than the least effective member. I've been on both ends of that 10x, but have never experienced a situation where all the programmers were approximately the same. There's always someone who is clearly more capable, and someone who is not.
      - Nomentatus 4y ago
        Hmmm. I'm gonna agree the slow way. I take it you are not referring to lines written/hour but, lines of code that survive; or time to working code that meets set goals, and maybe ten percent of the vicious, business-killing bugs. But then again, maybe the mediocre guys spend a whole lot of time looking for code that they can cut, paste and modify. That certainly used to happen with hardware circuits. Given that most large projects don't make it to the finish line, someone who has consistently pushed projects over the finish line looks like an outlier to me. If your 10x gal (or guy) can choose a far better algorithm, that's not just 10x; that's maybe clear success vs total failure. (So you'll be dividing by zero if you want a ratio, there.) To switch to the military context, it's funny how all the people I know who don't think there are any 10x Generals also firmly believe there are minus 10x Generals. Maybe even lots of them. And can somehow maintain both those beliefs at the same time. Both history and personal experience (not as a general) tells me that true competence is difficult to achieve and that most people are far too lazy to get there, or have personalities that block them. Everybody thinks the have an above-average sense of humor and are amongst the best drivers, but they know that they aren't 10x for humor and driving. Maybe, to maintain their illusion, they have to say there are no 10x ers, cause then they don't look above average anymore.
woooooo 4y ago
Even then, there's about 200 workdays in the year. There are probably scenarios where I'd rather have 2 days of Carmack than a year of a newbie, but they are rare and specialized.

CapitalistCartr 4y ago

I have personally seen programmers that were only marginally useful. More than once. Being ten times more effective isn't a great achievement. Management is the same way.

tux3 4y ago

You'd expect people who are much more productive than average, if they exist, to be relatively rare. So it's understandably not necessarily easy to round up a population of them and do a comparative study. But even without a study, some people seem noticeably hyper-productive, beyond noise or luck. I like to work on personal projects, but if you tried to compare my productivity to your local Fabrice Bellard-class programmer however you want to measure it, I'll have to admit the ratio's probably pretty far from 1x.