5 ms·

I've done 3 interviews in the last couple weeks where candidates couldn't do a simple Python dictionary lookup. These were people with a Masters in CS and have

I've done 3 interviews in the last couple weeks where candidates couldn't do a simple Python dictionary lookup. These were people with a Masters in CS and have Python as their primary language. I wouldn't worry about being an Impostor.

tombert 8y ago
My favorite is when people put "concurrency" as a feature on their resumes. When I ask them their opinions on actors vs. CSP, without fail, they don't know what the hell I'm talking about. Apparently "concurrency" to a lot of trained people just means "spinning up threads and use a lock occasionally".
vonmoltke 8y ago
I have done dozens of interviews in my career and have yet to meet anyone who was this bad. What am I doing wrong?
- mohaine 8y ago
  Probably working for a company with sane resume filtering. We outsourced that.
  - vonmoltke 8y ago
    Oh, yeah, I could see that. Most of my experience was with a company where my team and a dedicated technical recruiter did all the filtering, and where the engineers could override the recruiter decisions (in either direction).
  - foobiekr 8y ago
    Some of it is that, but I can say from my own experience that the company itself can draw large numbers of this kind of candidates. For various reasons ($) I’ve worked at a company that people view as a sort of safety school; you get a much higher fraction of incompetents at companies like this. You don’t see them at startups or at companies with solid technical reputations.
ramphastidae 8y ago
Am I missing something, or do you mean they didn't know `myDict['key'] `?
- mohaine 8y ago
  Yes. One candidate remembered to do it that way after some searching, the other never did. Both started out by looping over the items in the dict and stopping at the correct one. Asking if there was a simpler way to get values out of dict didn't help. The third candidate really didn't get far enough to get the dict wrong so I counted them as well but they may have known how to do a dict lookup. Here is the exercise: Given these dicts: {"a":[1,2],"b":[3,4],"z":[5]} {"a":[66],"b":[77]} Write a program in language of your choice that sums the common keys. a and b are common so result is: {"a":1+2+66,"b":3+4+77} so output should be: {"a":69,"b":84} This exercise is stupid simple. Note this is in an IDE/Editor, not a whiteboard.
  - tombert 8y ago
    I remember how I didn't believe that Fizzbuzz eliminates some crazy-high percentage of candidates until I started interviewing. I think most of the people I have interviewed are relatively smart people, but a lot of them really haven't had a lot of need to practice their algorithms, so they don't. I've had people fail the "tell me if these two strings are anagrams" thing, with one person's response being to have 26 counters, one for each letter.

tombert 8y ago

My favorite is when people put "concurrency" as a feature on their resumes. When I ask them their opinions on actors vs. CSP, without fail, they don't know what the hell I'm talking about. Apparently "concurrency" to a lot of trained people just means "spinning up threads and use a lock occasionally".

vonmoltke 8y ago

I have done dozens of interviews in my career and have yet to meet anyone who was this bad. What am I doing wrong?

mohaine 8y ago
Probably working for a company with sane resume filtering. We outsourced that.
- vonmoltke 8y ago
  Oh, yeah, I could see that. Most of my experience was with a company where my team and a dedicated technical recruiter did all the filtering, and where the engineers could override the recruiter decisions (in either direction).
- foobiekr 8y ago
  Some of it is that, but I can say from my own experience that the company itself can draw large numbers of this kind of candidates. For various reasons ($) I’ve worked at a company that people view as a sort of safety school; you get a much higher fraction of incompetents at companies like this. You don’t see them at startups or at companies with solid technical reputations.

ramphastidae 8y ago

Am I missing something, or do you mean they didn't know `myDict['key'] `?

mohaine 8y ago
Yes. One candidate remembered to do it that way after some searching, the other never did. Both started out by looping over the items in the dict and stopping at the correct one. Asking if there was a simpler way to get values out of dict didn't help. The third candidate really didn't get far enough to get the dict wrong so I counted them as well but they may have known how to do a dict lookup. Here is the exercise: Given these dicts: {"a":[1,2],"b":[3,4],"z":[5]} {"a":[66],"b":[77]} Write a program in language of your choice that sums the common keys. a and b are common so result is: {"a":1+2+66,"b":3+4+77} so output should be: {"a":69,"b":84} This exercise is stupid simple. Note this is in an IDE/Editor, not a whiteboard.
- tombert 8y ago
  I remember how I didn't believe that Fizzbuzz eliminates some crazy-high percentage of candidates until I started interviewing. I think most of the people I have interviewed are relatively smart people, but a lot of them really haven't had a lot of need to practice their algorithms, so they don't. I've had people fail the "tell me if these two strings are anagrams" thing, with one person's response being to have 26 counters, one for each letter.