I think I understand what you are trying to say. Maybe you need alternate examples to show what you mean:
Scenario 1)
------------------------
You have 2 candidates 'A' and 'B' in a team. 'A' is actually better than 'B', writes better code, and has the ability to debug issues which 'B' really can't get a handle on. Both
get a call to interview for a company in 2 weeks. 'A' has confidence in his/her abilities, so just goes through an algorithms/data structure book and writes 0 lines of code
in the 2 weeks. 'B' on the other hand, buys couple of programming interview books, goes through common questions (like reversing a linked list) and actually practices coding these questions. On interview day, 'A' is posed a question to merge 2 sorted linked lists. 'A' hasn't worked with linked lists for some time now, takes some time to come up with an answer, and writes up a solution on board that has some syntax errors. The interviewer had other questions to ask, but has no time because 'A' took a lot of time to code this one. ('A' has a cold start, so to speak).'B' when posed with the same question solves it within 15 mins (what with all the practice) With that confidence, 'B' aces the rest of the interviews. 'A' is kind of disturbed, cannot believe linked lists can be so tough, and does OK in the rest of the interviews, but still doesn't exude the confidence of a person who knows his stuff.
No points for guessing who is hired.
Scenario 2)
---------------------------
Instead of short interview times and questions about what everyone calls 'basic data structures', lets say both were given laptops during the interview and were given a program/requirement that neither had seen before and something that did require logical/analytical thinking to solve. Say, instead of 40 mins they were given 2-3 hours to complete the coding/debugging with access to a compiler and to language/API documentation. Who do you think would do better? (If it is not obvious, I back 'A' to do better)
I would really say scenario 2) is a better simulation of actual working conditions. Unfortunately, most interviews are like 1), maybe because 2) is more work for the interviewer. (logistics + framing a good enough problem)
OMG, blink1729, you read my mind. I couldn't have said it better ! Someone from salesforce.com (who I met via HN) told me that software folks are interviewed similarly to Scenario 2)!
Scenario 1) ------------------------ You have 2 candidates 'A' and 'B' in a team. 'A' is actually better than 'B', writes better code, and has the ability to debug issues which 'B' really can't get a handle on. Both get a call to interview for a company in 2 weeks. 'A' has confidence in his/her abilities, so just goes through an algorithms/data structure book and writes 0 lines of code in the 2 weeks. 'B' on the other hand, buys couple of programming interview books, goes through common questions (like reversing a linked list) and actually practices coding these questions. On interview day, 'A' is posed a question to merge 2 sorted linked lists. 'A' hasn't worked with linked lists for some time now, takes some time to come up with an answer, and writes up a solution on board that has some syntax errors. The interviewer had other questions to ask, but has no time because 'A' took a lot of time to code this one. ('A' has a cold start, so to speak).'B' when posed with the same question solves it within 15 mins (what with all the practice) With that confidence, 'B' aces the rest of the interviews. 'A' is kind of disturbed, cannot believe linked lists can be so tough, and does OK in the rest of the interviews, but still doesn't exude the confidence of a person who knows his stuff.
No points for guessing who is hired.
Scenario 2) --------------------------- Instead of short interview times and questions about what everyone calls 'basic data structures', lets say both were given laptops during the interview and were given a program/requirement that neither had seen before and something that did require logical/analytical thinking to solve. Say, instead of 40 mins they were given 2-3 hours to complete the coding/debugging with access to a compiler and to language/API documentation. Who do you think would do better? (If it is not obvious, I back 'A' to do better)
I would really say scenario 2) is a better simulation of actual working conditions. Unfortunately, most interviews are like 1), maybe because 2) is more work for the interviewer. (logistics + framing a good enough problem)