4 ms·

I'd have to agree with your point: "Most of them are doable but I think they are probably missing amazing developers that may not know how to solve those probl

I'd have to agree with your point: "Most of them are doable but I think they are probably missing amazing developers that may not know how to solve those problems, but they are capable of solving real-life problems (fix this bug, port this library, refactor this code...)." I do believe some startups are missing out on some amazingly talented developers by making this your rubric. Some people just don't some solve puzzles well, in the same way some people don't perform stellar on tests, but does that mean they are not capable of writing great code. Not at all. I've met some stellar coders who do not solve timed puzzle or timed tests well. Will someone please explain to me exactly why solving a puzzle is a indicator of potential workplace success. Do you want to hire puzzle solvers or actual developers? Should I spend my time improving my puzzle solving skills or churning out code that solves real-life problems?

gvb 15y ago
Puzzles are human benchmarks. Like all benchmarks, their objectives are: 1) They have to be short enough to complete in a "reasonable" amount of time. 2) They have to have enough resolution to distinguish between good, better, and best. 3) They have to be representative of the real life problems that the person / computer will be solving. Most benchmarks only meet objective #1. Further, benchmarks that only meet objective #1 are easily "game-able". Benchmarks which meet objectives #2 and #3 invariably don't meet #1, which makes them hard to use as a benchmark.
devs1010 15y ago
I agree, its utter nonsense for the most part, it seems some companies just want you to jump through a bunch of hoops and to make you feel stupid. I think having you work on a small contracted project or something related to the business domain would be a lot more useful.
- buss 15y ago
  That would probably be more useful, but it doesn't scale. An interview like that takes up a ton of resources and can leave a bad impression on a candidate. Not only are you paying for devs to help and evaluate the candidate, you are doing this over a much longer period than a traditional interview, and you have to pay the candidate for the work they did, even if you're going to throw it out. Nobody wants to force candidates to jump through hoops or make them feel stupid. That's a recipe for bad word-of-mouth and never hiring good devs. When I conduct interviews, I always try to make people feel like they did an OK job. I don't want them to get off the phone thinking they did really poorly. I try hard to seem impartial, but that's often quite difficult to do. The root problem is that nobody knows what the hell they're doing in interviews. We don't know the best way to evaluate candidates, so we try various things that are ineffective or wasteful. Eventually we'll figure it out.
  - devs1010 15y ago
    if its for web dev, make them do something that has a web UI, give specific contraints and then write an automated integration test for it (such as using selenium), run this on the incoming submissions, only review the ones that pass, this is just a quick idea... there are ways to automate this....
atarian 15y ago
I can think of a number of things why it's used: 1) It tests if you're able to think on your feet. Anyone can Google for help/solution/code for a problem, but most people want to know if you can think for yourself and that your first reaction to a problem isn't to go for the browser. They want to know what you're thinking and how you can solve a problem. 2) It tests how you handle stress. Do you blank out and panic? Do you start shaking uncontrollably and can't speak properly? Crunch time gets stressful, and if you're easily stressed out your desired employer may see this and think you're not a good fit. 3) It tests teamwork. A lot of times, interviewers will drop hints or advice on how to solve a problem. I've heard that some programmers will just disregard it or try to fight it and solve it their own way. It's a good time for an interviewer to see if you're someone they'd want to work with. EDIT: If you're going to downvote me, you should leave a reply why you found my comment to be deconstructive or irrelevant.

gvb 15y ago

Puzzles are human benchmarks. Like all benchmarks, their objectives are: 1) They have to be short enough to complete in a "reasonable" amount of time. 2) They have to have enough resolution to distinguish between good, better, and best. 3) They have to be representative of the real life problems that the person / computer will be solving. Most benchmarks only meet objective #1. Further, benchmarks that only meet objective #1 are easily "game-able". Benchmarks which meet objectives #2 and #3 invariably don't meet #1, which makes them hard to use as a benchmark.

devs1010 15y ago

I agree, its utter nonsense for the most part, it seems some companies just want you to jump through a bunch of hoops and to make you feel stupid. I think having you work on a small contracted project or something related to the business domain would be a lot more useful.

buss 15y ago
That would probably be more useful, but it doesn't scale. An interview like that takes up a ton of resources and can leave a bad impression on a candidate. Not only are you paying for devs to help and evaluate the candidate, you are doing this over a much longer period than a traditional interview, and you have to pay the candidate for the work they did, even if you're going to throw it out. Nobody wants to force candidates to jump through hoops or make them feel stupid. That's a recipe for bad word-of-mouth and never hiring good devs. When I conduct interviews, I always try to make people feel like they did an OK job. I don't want them to get off the phone thinking they did really poorly. I try hard to seem impartial, but that's often quite difficult to do. The root problem is that nobody knows what the hell they're doing in interviews. We don't know the best way to evaluate candidates, so we try various things that are ineffective or wasteful. Eventually we'll figure it out.
- devs1010 15y ago
  if its for web dev, make them do something that has a web UI, give specific contraints and then write an automated integration test for it (such as using selenium), run this on the incoming submissions, only review the ones that pass, this is just a quick idea... there are ways to automate this....

atarian 15y ago

I can think of a number of things why it's used: 1) It tests if you're able to think on your feet. Anyone can Google for help/solution/code for a problem, but most people want to know if you can think for yourself and that your first reaction to a problem isn't to go for the browser. They want to know what you're thinking and how you can solve a problem. 2) It tests how you handle stress. Do you blank out and panic? Do you start shaking uncontrollably and can't speak properly? Crunch time gets stressful, and if you're easily stressed out your desired employer may see this and think you're not a good fit. 3) It tests teamwork. A lot of times, interviewers will drop hints or advice on how to solve a problem. I've heard that some programmers will just disregard it or try to fight it and solve it their own way. It's a good time for an interviewer to see if you're someone they'd want to work with. EDIT: If you're going to downvote me, you should leave a reply why you found my comment to be deconstructive or irrelevant.