4 ms·

> Because the kook that refuses to explain merge sort might be better at the job than the guy who desperatly jumps through hoops. Who cares if the kook is bett

> Because the kook that refuses to explain merge sort might be better at the job than the guy who desperatly jumps through hoops. Who cares if the kook is better if the other one is good enough? At some point, "better" hits diminishing returns and is no longer worth it. It's the reason "we only hire the top X%" attitude is wasteful bullshit for most companies.

JanneVee 11y ago
>At some point, "better" hits diminishing returns and is no longer worth it. Well the whole point throwing out CS trivia questions is to find the better candidate. It is the diminishing returns... by simply restricting your candidate pool to people that have the time to jump through hoops.
beat 11y ago
One of my co-founders made a really interesting observation the other day... the 10x programmer often gets his 10x by leeching the performance of other team members. 10x programmers don't talk through their decisions with teammates, or think about the consequences of their rapid changes on other members, or indeed, the long-term health of the project. If a 10x programmer turns a 3x team into a 1x team, that super-genius isn't worth having. And this is why I hate working with primadonnas who think they're brilliant.
- JoeAltmaier 11y ago
  How do you feel about hard-working brilliant folks who are humble? This sounds more like resentment to me. "Its hard to keep up with brilliant folks" is not a reason to avoid them. They may make more work, but you get to a different place than plodding along with the herd.
  - beat 11y ago
    It's not a matter of "It's hard to keep up with the brilliant folks". It's that they are often actively destructive to the productivity of the team. Hence my co-founder's comment. Brilliant people can often be easygoing and a great benefit to their teams. They do it by moving the whole team forward. They're never 10x individually in that case, but they may elevate a 3x team to 5x. edit: Find me an engineer who is both brilliant and humble, and I'll bend over backwards to hire that person.
    - eonw 11y ago
      ive only met one in my life, and i wouldn't recommend him to anyone... if he was available and i could afford him i would in do it a heartbeat first. he was a 10x with more diplomacy and kindness then any co-worker ive ever had. rare breed indeed.
    - JoeAltmaier 11y ago
      Hm. It could be both. And probably the 'cowboys' we've all experienced are not all brilliant. It doesn't take brilliance to torpedo a team. And I suggest that 3X a small team may be less than 10X the one person. Maybe a better team is one that supports a brilliant developer, making a deliverable out of their prototype. That's another way to manage it.
- dasil003 11y ago
  To make this a bit more subtle, I consider 10x to be contextual. Some developers will legitimately be 10x in some circumstances but not others. The kind of prima donna you are talking about (assuming they are truly brilliant and not just a Dunning-Kruger case) can probably only be 10x on small solo projects. But there are definitely those programmers who are 10x or even 100x by being a combination of brilliant plus collaboration-minded. The thing about programming is that it is a relatively easy form of automation of data processing, so whereas a single developer can have multiples of impact far greater than a traditional office worker, and sometimes even greater than the impact of a traditional office manager (because software is more scalable than people), you can get a multiplicative effect when you combine great programming with great programmer management and collaboration. This is where you'll find a lot of my programming heroes. For example, what Yehuda Katz did with Bundler in terms of gathering community input and hammering something into shape to solve a whole host of hairy problems with RubyGems which no single individual could completely grok is inspiring. The amount of impact a project like that has on the productivity of thousands is impossible to quantify. I guess Linus Torvalds would be the quintessential example of that, for his ability to channel the efforts of thousands of developers in a cohesive direction while maintaining an impressive standard of quality. So being 10x, 100x, 1000x is possible, but it depends on working on the right problems with the right people.
- shados 11y ago
  Yeah, we had that at one of my previous employer. Some guy was really a 20x dev. But he did it at "all costs". That meant working from home, not attending meetings, not answering emails, ignoring the PMs, not mentoring junior devs, not explaining how he did stuff, etc. Now, we all hate a lot of the above stuff, and many companies handle these things better than others, but it did mean the only reason he was "20x" was because he didn't do 90% of the stuff that was expected of everyone else. Maybe the company culture needed to change, but until it did, everyone else looked really bad compared to him, even though they were essentially spending their time trying to deal with his crap without help.
- greenyoda 11y ago
  "the 10x programmer often gets his 10x by leeching the performance of other team members. 10x programmers don't talk through their decisions with teammates, or think about the consequences of their rapid changes on other members, or indeed, the long-term health of the project." I wouldn't call that person a 10x programmer. I think that the only meaningful way to define a 10x programmer is someone who has ten times the contribution to the company's bottom line of an average programmer (i.e., someone who would actually deserve to be paid ten times as much as an average programmer, but could probably be hired for far less than that). However, if someone produces 10x more code than the average programmer but "turns a 3x team into a 1x team", then their net productivity is negative and they should be immediately fired.

JanneVee 11y ago

>At some point, "better" hits diminishing returns and is no longer worth it. Well the whole point throwing out CS trivia questions is to find the better candidate. It is the diminishing returns... by simply restricting your candidate pool to people that have the time to jump through hoops.

beat 11y ago

One of my co-founders made a really interesting observation the other day... the 10x programmer often gets his 10x by leeching the performance of other team members. 10x programmers don't talk through their decisions with teammates, or think about the consequences of their rapid changes on other members, or indeed, the long-term health of the project. If a 10x programmer turns a 3x team into a 1x team, that super-genius isn't worth having. And this is why I hate working with primadonnas who think they're brilliant.

JoeAltmaier 11y ago
How do you feel about hard-working brilliant folks who are humble? This sounds more like resentment to me. "Its hard to keep up with brilliant folks" is not a reason to avoid them. They may make more work, but you get to a different place than plodding along with the herd.
- beat 11y ago
  It's not a matter of "It's hard to keep up with the brilliant folks". It's that they are often actively destructive to the productivity of the team. Hence my co-founder's comment. Brilliant people can often be easygoing and a great benefit to their teams. They do it by moving the whole team forward. They're never 10x individually in that case, but they may elevate a 3x team to 5x. edit: Find me an engineer who is both brilliant and humble, and I'll bend over backwards to hire that person.
  - eonw 11y ago
    ive only met one in my life, and i wouldn't recommend him to anyone... if he was available and i could afford him i would in do it a heartbeat first. he was a 10x with more diplomacy and kindness then any co-worker ive ever had. rare breed indeed.
  - JoeAltmaier 11y ago
    Hm. It could be both. And probably the 'cowboys' we've all experienced are not all brilliant. It doesn't take brilliance to torpedo a team. And I suggest that 3X a small team may be less than 10X the one person. Maybe a better team is one that supports a brilliant developer, making a deliverable out of their prototype. That's another way to manage it.
dasil003 11y ago
To make this a bit more subtle, I consider 10x to be contextual. Some developers will legitimately be 10x in some circumstances but not others. The kind of prima donna you are talking about (assuming they are truly brilliant and not just a Dunning-Kruger case) can probably only be 10x on small solo projects. But there are definitely those programmers who are 10x or even 100x by being a combination of brilliant plus collaboration-minded. The thing about programming is that it is a relatively easy form of automation of data processing, so whereas a single developer can have multiples of impact far greater than a traditional office worker, and sometimes even greater than the impact of a traditional office manager (because software is more scalable than people), you can get a multiplicative effect when you combine great programming with great programmer management and collaboration. This is where you'll find a lot of my programming heroes. For example, what Yehuda Katz did with Bundler in terms of gathering community input and hammering something into shape to solve a whole host of hairy problems with RubyGems which no single individual could completely grok is inspiring. The amount of impact a project like that has on the productivity of thousands is impossible to quantify. I guess Linus Torvalds would be the quintessential example of that, for his ability to channel the efforts of thousands of developers in a cohesive direction while maintaining an impressive standard of quality. So being 10x, 100x, 1000x is possible, but it depends on working on the right problems with the right people.
shados 11y ago
Yeah, we had that at one of my previous employer. Some guy was really a 20x dev. But he did it at "all costs". That meant working from home, not attending meetings, not answering emails, ignoring the PMs, not mentoring junior devs, not explaining how he did stuff, etc. Now, we all hate a lot of the above stuff, and many companies handle these things better than others, but it did mean the only reason he was "20x" was because he didn't do 90% of the stuff that was expected of everyone else. Maybe the company culture needed to change, but until it did, everyone else looked really bad compared to him, even though they were essentially spending their time trying to deal with his crap without help.
greenyoda 11y ago
"the 10x programmer often gets his 10x by leeching the performance of other team members. 10x programmers don't talk through their decisions with teammates, or think about the consequences of their rapid changes on other members, or indeed, the long-term health of the project." I wouldn't call that person a 10x programmer. I think that the only meaningful way to define a 10x programmer is someone who has ten times the contribution to the company's bottom line of an average programmer (i.e., someone who would actually deserve to be paid ten times as much as an average programmer, but could probably be hired for far less than that). However, if someone produces 10x more code than the average programmer but "turns a 3x team into a 1x team", then their net productivity is negative and they should be immediately fired.