Person I’d like to debate: @gayle, regarding this comment, which I categorically disagree with: t.co/eAHngQzG7s tqbf tweeted on Mar 09, 2015 06:44

@tqbf "Categorically" is awfully strong, and logically impossible :). gayle replied on Mar 09, 2015 06:46

@gayle Sorry. I was imprecise. I meant, “there was no argument made in that comment I agree with at all.” tqbf replied on Mar 09, 2015 06:46

@gayle … I was glad to see it, by the way! I’ve had a hard time finding *someone* who might disagree about how bad interviews are. tqbf replied on Mar 09, 2015 06:47

@tqbf So you disagree with "The interview process isn't perfect." AND "it actually works fairly well."? gayle replied on Mar 09, 2015 06:48

@gayle The interview process is fatally flawed, not merely imperfect. It does not work at all. tqbf replied on Mar 09, 2015 06:48

@gayle It rejects huge numbers of capable people and routinely hires people who fail at the job. tqbf replied on Mar 09, 2015 06:49

@tqbf So you think Google and Facebook would be equally successful hiring at random? gayle replied on Mar 09, 2015 06:48

@tqbf Having done hundreds of these interviews AND talked with their managers, the people I identify as smart tend to in fact be smart. gayle replied on Mar 09, 2015 06:49

@gayle I do not doubt that at all, and not just because “smart” is fuzzily impossible to nail down. tqbf replied on Mar 09, 2015 06:50

@gayle I have to think about whether I can pluralize the word “hundred” in my experience doing this job, FWIW. tqbf replied on Mar 09, 2015 06:50

@gayle Cost of bad hire is enormous; a 90% success rate is still dreadful. Meanwhile… tqbf replied on Mar 09, 2015 06:52

@gayle … meanwhile, the real problem is how many mispriced extremely capable people are rejected by the interview process. tqbf replied on Mar 09, 2015 06:53

@gayle mis-identifying people as smart who aren’t smart is a problem, but not the big one. tqbf replied on Mar 09, 2015 06:53

@tqbf Yeah, that will happen with any interview process though. Fortunately companies don't use all use the same process. gayle replied on Mar 09, 2015 06:54

@gayle Take what you said about “giving people a new, hard problem”… tqbf replied on Mar 09, 2015 06:54

@gayle You leave out, “in an incredibly hostile environment, with cortisol retarding judgement and recall”. tqbf replied on Mar 09, 2015 06:55

@gayle Also: remember, your experience is (in large part) as Google, which has a very distinctive hiring pipeline… tqbf replied on Mar 09, 2015 06:55

@gayle … selecting for people who are extremely conversant in high-pressure selection processes. tqbf replied on Mar 09, 2015 06:56

@gayle But Google doesn’t in reality benefit from hiring mainly people who are excellent at high-pressure interviews! tqbf replied on Mar 09, 2015 06:56

@tqbf The goal is not to select for that. On the contrary, Google trains interviews to be nice & supportive. Still some pressure, yes. gayle replied on Mar 09, 2015 06:59

@gayle As I wrote: we did material, overt things to mitigate interview hostility; great candidates still overwhelmingly stressed out! tqbf replied on Mar 09, 2015 07:00

@gayle Interviews are an intrinsically hostile experience. One of the most hostile things normal people EVER experience. tqbf replied on Mar 09, 2015 07:00

@tqbf Absolutely agree with you there. gayle replied on Mar 09, 2015 07:01

@gayle So, respectfully: why do you expect candidates to handle “new, hard problems” under those circumstances? tqbf replied on Mar 09, 2015 07:01

@gayle You don’t expect them to do their day to day work with intensely elevated cortisol levels. tqbf replied on Mar 09, 2015 07:01

@tqbf It'd be great to somehow entirely remove stress from interviews. But how (without other consequences)? gayle replied on Mar 09, 2015 07:04

@gayle Respectfully referring you back here: t.co/YtCMS6DDWj tqbf replied on Mar 09, 2015 07:04

@tqbf Yes. Read it. gayle replied on Mar 09, 2015 07:05

@tqbf Work sample tests/homework generally can't identify problem-solving/algorithm skills & can put unreasonable time burden on candidate. gayle replied on Mar 09, 2015 07:06

@gayle I’m not sure I’m clear on how any algorithm skill could be better demonstrated in a conference room than at candidate’s home. tqbf replied on Mar 09, 2015 07:07

@gayle Can you help me understand which problem solving skill is best demonstrated in prospective employer’s offices? tqbf replied on Mar 09, 2015 07:07

@gayle Unreasonable time burden is a problem. So we set up our work samples to take less time than F2F interview. tqbf replied on Mar 09, 2015 07:08

@tqbf Because candidates cheat. gayle replied on Mar 09, 2015 07:08

@gayle No, they don’t. tqbf replied on Mar 09, 2015 07:08

@gayle Over all candidates — a decently large number, incl. everyone we didn’t hire — we had ZERO instances of cheating. ZERO. tqbf replied on Mar 09, 2015 07:08

@tqbf Then you're unaware or defining cheating differently. Give a candidate an algorithm problem and they'll google / ask friends for help. gayle replied on Mar 09, 2015 07:09

@gayle 100% retention after hire. No fires, no quits. EXTREMELY demanding workload. No, we did not miss cheaters. tqbf replied on Mar 09, 2015 07:10

@gayle You can’t give candidates problems they can simply Google answers for! Those are terrible questions regardless of setting! tqbf replied on Mar 09, 2015 07:10

@gayle That’s not a work-sample test, it’s a trivia question. Work-sample must mirror the actual work candidates deliver. tqbf replied on Mar 09, 2015 07:10

@gayle The post you read gave an example of how to do that for standard web dev job. Tell me how you’d cheat on it? tqbf replied on Mar 09, 2015 07:11

@tqbf @gayle if a company the size of google uses a question for a week it becomes googlable. I don't see this process scaling. aarondufour replied on Mar 09, 2015 10:58

@aarondufour @gayle Google interview questions are open secrets already. tqbf replied on Mar 09, 2015 11:01

@tqbf @aarondufour Not really. There are no "google interview Qs." There's no list. There are many thousands of Qs that have been asked. gayle replied on Mar 09, 2015 11:12

@tqbf @gayle but there are so many of them that you cant memorize them prior to an interview. Standardization or work sample eliminates that aarondufour replied on Mar 09, 2015 11:03

@aarondufour @gayle Properly designed work-sample tests are much harder to cheat on than lame algo questions. tqbf replied on Mar 09, 2015 11:05

@tqbf @aarondufour They're not testing the same thing though. And work tests have their own flaws. Works great for some but not all. gayle replied on Mar 09, 2015 11:24

Person I’d like to debate: @gayle, regarding this comment, which I categorically disagree with: t.co/eAHngQzG7s tqbf tweeted on Mar 09, 2015 06:44

@tqbf "Categorically" is awfully strong, and logically impossible :). gayle replied on Mar 09, 2015 06:46

@gayle Sorry. I was imprecise. I meant, “there was no argument made in that comment I agree with at all.” tqbf replied on Mar 09, 2015 06:46

@tqbf So you disagree with "The interview process isn't perfect." AND "it actually works fairly well."? gayle replied on Mar 09, 2015 06:48

@gayle The interview process is fatally flawed, not merely imperfect. It does not work at all. tqbf replied on Mar 09, 2015 06:48

@tqbf Having done hundreds of these interviews AND talked with their managers, the people I identify as smart tend to in fact be smart. gayle replied on Mar 09, 2015 06:49

@gayle … meanwhile, the real problem is how many mispriced extremely capable people are rejected by the interview process. tqbf replied on Mar 09, 2015 06:53

@tqbf Yeah, that will happen with any interview process though. Fortunately companies don't use all use the same process. gayle replied on Mar 09, 2015 06:54

@gayle But Google doesn’t in reality benefit from hiring mainly people who are excellent at high-pressure interviews! tqbf replied on Mar 09, 2015 06:56

@tqbf The goal is not to select for that. On the contrary, Google trains interviews to be nice & supportive. Still some pressure, yes. gayle replied on Mar 09, 2015 06:59

@gayle Interviews are an intrinsically hostile experience. One of the most hostile things normal people EVER experience. tqbf replied on Mar 09, 2015 07:00

@tqbf Absolutely agree with you there. gayle replied on Mar 09, 2015 07:01

@gayle You don’t expect them to do their day to day work with intensely elevated cortisol levels. tqbf replied on Mar 09, 2015 07:01

@tqbf It'd be great to somehow entirely remove stress from interviews. But how (without other consequences)? gayle replied on Mar 09, 2015 07:04

@gayle Respectfully referring you back here: t.co/YtCMS6DDWj tqbf replied on Mar 09, 2015 07:04

@tqbf Work sample tests/homework generally can't identify problem-solving/algorithm skills & can put unreasonable time burden on candidate. gayle replied on Mar 09, 2015 07:06

@gayle I’m not sure I’m clear on how any algorithm skill could be better demonstrated in a conference room than at candidate’s home. tqbf replied on Mar 09, 2015 07:07

@tqbf Because candidates cheat. gayle replied on Mar 09, 2015 07:08

@gayle Over all candidates — a decently large number, incl. everyone we didn’t hire — we had ZERO instances of cheating. ZERO. tqbf replied on Mar 09, 2015 07:08

@tqbf Then you're unaware or defining cheating differently. Give a candidate an algorithm problem and they'll google / ask friends for help. gayle replied on Mar 09, 2015 07:09

@gayle You can’t give candidates problems they can simply Google answers for! Those are terrible questions regardless of setting! tqbf replied on Mar 09, 2015 07:10

@tqbf @gayle if a company the size of google uses a question for a week it becomes googlable. I don't see this process scaling. aarondufour replied on Mar 09, 2015 10:58

@aarondufour @gayle Google interview questions are open secrets already. tqbf replied on Mar 09, 2015 11:01

@tqbf @gayle but there are so many of them that you cant memorize them prior to an interview. Standardization or work sample eliminates that aarondufour replied on Mar 09, 2015 11:03

@aarondufour @gayle Properly designed work-sample tests are much harder to cheat on than lame algo questions. tqbf replied on Mar 09, 2015 11:05

@tqbf @gayle maybe not enough detail in the example; I dont see how it helps. There's a question and a solution, both can be put on the web. aarondufour replied on Mar 09, 2015 11:08

@aarondufour @gayle Bad work sample test: “show us an implementation of an AVL tree”. tqbf replied on Mar 09, 2015 11:10

@tqbf @aarondufour Some companies want to test algorithmic problem-solving. You don't (which is fine). At-home tests don't work for that. gayle replied on Mar 09, 2015 11:11

Person I’d like to debate: @gayle, regarding this comment, which I categorically disagree with: t.co/eAHngQzG7s tqbf tweeted on Mar 09, 2015 06:44

@tqbf "Categorically" is awfully strong, and logically impossible :). gayle replied on Mar 09, 2015 06:46

@gayle Sorry. I was imprecise. I meant, “there was no argument made in that comment I agree with at all.” tqbf replied on Mar 09, 2015 06:46

@tqbf So you disagree with "The interview process isn't perfect." AND "it actually works fairly well."? gayle replied on Mar 09, 2015 06:48

@gayle The interview process is fatally flawed, not merely imperfect. It does not work at all. tqbf replied on Mar 09, 2015 06:48

@tqbf Having done hundreds of these interviews AND talked with their managers, the people I identify as smart tend to in fact be smart. gayle replied on Mar 09, 2015 06:49

@gayle … meanwhile, the real problem is how many mispriced extremely capable people are rejected by the interview process. tqbf replied on Mar 09, 2015 06:53

@tqbf Yeah, that will happen with any interview process though. Fortunately companies don't use all use the same process. gayle replied on Mar 09, 2015 06:54

@gayle But Google doesn’t in reality benefit from hiring mainly people who are excellent at high-pressure interviews! tqbf replied on Mar 09, 2015 06:56

@tqbf The goal is not to select for that. On the contrary, Google trains interviews to be nice & supportive. Still some pressure, yes. gayle replied on Mar 09, 2015 06:59

@gayle Interviews are an intrinsically hostile experience. One of the most hostile things normal people EVER experience. tqbf replied on Mar 09, 2015 07:00

@tqbf Absolutely agree with you there. gayle replied on Mar 09, 2015 07:01

@gayle You don’t expect them to do their day to day work with intensely elevated cortisol levels. tqbf replied on Mar 09, 2015 07:01

@tqbf It'd be great to somehow entirely remove stress from interviews. But how (without other consequences)? gayle replied on Mar 09, 2015 07:04

@gayle Respectfully referring you back here: t.co/YtCMS6DDWj tqbf replied on Mar 09, 2015 07:04

@tqbf Work sample tests/homework generally can't identify problem-solving/algorithm skills & can put unreasonable time burden on candidate. gayle replied on Mar 09, 2015 07:06

@gayle I’m not sure I’m clear on how any algorithm skill could be better demonstrated in a conference room than at candidate’s home. tqbf replied on Mar 09, 2015 07:07

@tqbf Because candidates cheat. gayle replied on Mar 09, 2015 07:08

@gayle Over all candidates — a decently large number, incl. everyone we didn’t hire — we had ZERO instances of cheating. ZERO. tqbf replied on Mar 09, 2015 07:08

@tqbf Then you're unaware or defining cheating differently. Give a candidate an algorithm problem and they'll google / ask friends for help. gayle replied on Mar 09, 2015 07:09

@gayle The post you read gave an example of how to do that for standard web dev job. Tell me how you’d cheat on it? tqbf replied on Mar 09, 2015 07:11

@tqbf How long did your test take candidates, out of curiosity? gayle replied on Mar 09, 2015 07:12

@gayle 2-3 hours each. tqbf replied on Mar 09, 2015 07:12

@gayle Battery of three tests. Extremely abbreviated in-person interview (less than 3 hours), immediate decision. tqbf replied on Mar 09, 2015 07:12

@tqbf Did you invite people onsite regardless of their performance on the test? Or just the good people? gayle replied on Mar 09, 2015 07:14

@gayle Work-sample challenges are a downselect. tqbf replied on Mar 09, 2015 07:14

@gayle Candidates prefer them, sharply; people we passed on wrote us explaining why: tqbf replied on Mar 09, 2015 07:15

@gayle - challenges were interesting and, because not on-site, not stressful, and… tqbf replied on Mar 09, 2015 07:15

@gayle - total time commitment lower, and (importantly) MORE PREDICTABLE than with interviews, and… tqbf replied on Mar 09, 2015 07:15

@gayle - scheduling flexibility and no requirement to dress up and find their way to office for job they wouldn’t get tqbf replied on Mar 09, 2015 07:15

@tqbf That's great. It too has flaws, but it works very well for some companies. Not all. Same with in-person algorithmic interviews. gayle replied on Mar 09, 2015 07:26

@gayle Just to be clear: I’m arguing that in-person algorithm interviews don’t work well for ANYONE. tqbf replied on Mar 09, 2015 07:27

@gayle At the very least, they bias selection, and also raise cost for employer, who pays premium for subset of devs… tqbf replied on Mar 09, 2015 07:27

@gayle … which is a subset they don’t really care about (again: nobody makes devs work under interview-type pressure day-to-day) tqbf replied on Mar 09, 2015 07:28

@tqbf The success of Google, Facebook, Amazon, Dropbox, and all these other companies is remarkable then, with such a broken process. gayle replied on Mar 09, 2015 07:28

@gayle I agree, it is remarkable. tqbf replied on Mar 09, 2015 07:30

@gayle @tqbf there's two ways to fail at recruiting - over-rejecting and over-accepting. Some can afford to over-reject. mik235 replied on Mar 09, 2015 07:55

@mik235 @gayle Notably, they can only fail to over-reject if their serious competitors ALSO fail to over-reject. tqbf replied on Mar 09, 2015 07:56

@tqbf @mik235 But you're arguing that their interview processes isn't better than random. gayle replied on Mar 09, 2015 07:58

@gayle @mik235 If everyone pushes candidates through a PRF, nobody is disadvantaged by it, until someone stops. tqbf replied on Mar 09, 2015 07:59

@gayle @mik235 Shorthand: if GOOG AAPL FB AMZN literally used monkeys to select candidates, as long as they all did, they’d be fine. tqbf replied on Mar 09, 2015 08:00

@tqbf @mik235 That is a very strange position to take. I doubt a set of randomly selected candidates could build a system like google. gayle replied on Mar 09, 2015 08:02

@gayle @mik235 Once again: success of Google with poor process doesn’t vindicate process; nobody yet competing on hiring process. tqbf replied on Mar 09, 2015 08:03

@gayle @mik235 That is going to change, very soon; it will change before immigration law does. tqbf replied on Mar 09, 2015 08:03

@tqbf @mik235 Nearly all top companies have built excellent systems with people hired through this process. Good evidence of effectiveness. gayle replied on Mar 09, 2015 08:08

@gayle @mik235 The market is just now starting to demand that companies do better than this process. tqbf replied on Mar 09, 2015 08:09

@gayle @tqbf @mik235 I have a tiger rock here. Never been mauled by a tiger while holding it. I'll sell it to you for lots-of-bux. edropple replied on Mar 09, 2015 08:09

@edropple @tqbf @mik235 That has about zero relationship to what I said... gayle replied on Mar 09, 2015 08:23

@gayle @tqbf @mik235 Top companies do X. Top companies do Y. Therefore, X implies Y. edropple replied on Mar 09, 2015 08:23

@edropple @tqbf @mik235 Didn't say X causes Y. But X clearly isn't getting in the way of Y. gayle replied on Mar 09, 2015 08:30

@gayle @tqbf @mik235 Can you prove that? edropple replied on Mar 09, 2015 08:31

@edropple @tqbf @mik235 Yes... Very successful companies use this process. Therefore, this process does not prevent building such companies. gayle replied on Mar 09, 2015 08:35

@gayle @tqbf @mik235 Asserting that you're at a local maximum is real hard when few are looking for better. edropple replied on Mar 09, 2015 08:37

@gayle @tqbf @mik235 I think we can agree that it works to an extent, but thats not evidence of effectiveness. There are conflating factors. edropple replied on Mar 09, 2015 08:37

@gayle @tqbf @mik235 You're assigning credit to something without ruling out that, say, they get a TON of applicants and can miss on most. edropple replied on Mar 09, 2015 08:38

@gayle @tqbf @mik235 That's not a credit to the process, that's a credit to the company's (employment) market position. edropple replied on Mar 09, 2015 08:38

@edropple @tqbf @mik235 No. You're missing where this discussion stemmed from - @tqbf 's statement that it's no better than random. gayle replied on Mar 09, 2015 08:40

@gayle @edropple @mik235 Every single company relying on interviews could be improved by replacing them. tqbf replied on Mar 09, 2015 08:41

@gayle @tqbf Thomas's rhetoric is muddying that you're making a positive statement of the effectiveness of X. edropple replied on Mar 09, 2015 08:41

@edropple @tqbf If you're arguing that the interviews are no better than random, then it's meaningful that top companies do it. gayle replied on Mar 09, 2015 09:15

@gayle @tqbf That assumes that they know what they're doing any better than anybody else. Or, in other words, "they haven't died yet." edropple replied on Mar 09, 2015 09:16

@gayle @tqbf If a horrific process with great supply can be differentiated from a great process with a great supply, you haven't proven it. edropple replied on Mar 09, 2015 09:18

@edropple @tqbf But if it's no better than random, supply is irrelevant. gayle replied on Mar 09, 2015 09:27

@gayle @edropple It really bugged you that I used the word “random”. Sorry! tqbf replied on Mar 09, 2015 09:28

@tqbf @edropple It's not the word. It's the concept that no interview is better than algorithms interview. Seems absurd & extreme position. gayle replied on Mar 09, 2015 09:42

@gayle @edropple Yes, I DEFINITELY think no interview better than algorithm interview. tqbf replied on Mar 09, 2015 09:42

Person I’d like to debate: @gayle, regarding this comment, which I categorically disagree with: t.co/eAHngQzG7s tqbf tweeted on Mar 09, 2015 06:44

@tqbf "Categorically" is awfully strong, and logically impossible :). gayle replied on Mar 09, 2015 06:46

@gayle Sorry. I was imprecise. I meant, “there was no argument made in that comment I agree with at all.” tqbf replied on Mar 09, 2015 06:46

@tqbf So you disagree with "The interview process isn't perfect." AND "it actually works fairly well."? gayle replied on Mar 09, 2015 06:48

@gayle The interview process is fatally flawed, not merely imperfect. It does not work at all. tqbf replied on Mar 09, 2015 06:48

@tqbf Having done hundreds of these interviews AND talked with their managers, the people I identify as smart tend to in fact be smart. gayle replied on Mar 09, 2015 06:49

@gayle … meanwhile, the real problem is how many mispriced extremely capable people are rejected by the interview process. tqbf replied on Mar 09, 2015 06:53

@tqbf Yeah, that will happen with any interview process though. Fortunately companies don't use all use the same process. gayle replied on Mar 09, 2015 06:54

@gayle But Google doesn’t in reality benefit from hiring mainly people who are excellent at high-pressure interviews! tqbf replied on Mar 09, 2015 06:56

@tqbf The goal is not to select for that. On the contrary, Google trains interviews to be nice & supportive. Still some pressure, yes. gayle replied on Mar 09, 2015 06:59

@gayle Interviews are an intrinsically hostile experience. One of the most hostile things normal people EVER experience. tqbf replied on Mar 09, 2015 07:00

@tqbf Absolutely agree with you there. gayle replied on Mar 09, 2015 07:01

@gayle You don’t expect them to do their day to day work with intensely elevated cortisol levels. tqbf replied on Mar 09, 2015 07:01

@tqbf It'd be great to somehow entirely remove stress from interviews. But how (without other consequences)? gayle replied on Mar 09, 2015 07:04

@gayle Respectfully referring you back here: t.co/YtCMS6DDWj tqbf replied on Mar 09, 2015 07:04

@tqbf Work sample tests/homework generally can't identify problem-solving/algorithm skills & can put unreasonable time burden on candidate. gayle replied on Mar 09, 2015 07:06

@gayle I’m not sure I’m clear on how any algorithm skill could be better demonstrated in a conference room than at candidate’s home. tqbf replied on Mar 09, 2015 07:07

@tqbf Because candidates cheat. gayle replied on Mar 09, 2015 07:08

@gayle Over all candidates — a decently large number, incl. everyone we didn’t hire — we had ZERO instances of cheating. ZERO. tqbf replied on Mar 09, 2015 07:08

@tqbf Then you're unaware or defining cheating differently. Give a candidate an algorithm problem and they'll google / ask friends for help. gayle replied on Mar 09, 2015 07:09

@gayle The post you read gave an example of how to do that for standard web dev job. Tell me how you’d cheat on it? tqbf replied on Mar 09, 2015 07:11

@tqbf How long did your test take candidates, out of curiosity? gayle replied on Mar 09, 2015 07:12

@gayle 2-3 hours each. tqbf replied on Mar 09, 2015 07:12

@tqbf Did you invite people onsite regardless of their performance on the test? Or just the good people? gayle replied on Mar 09, 2015 07:14

@gayle - scheduling flexibility and no requirement to dress up and find their way to office for job they wouldn’t get tqbf replied on Mar 09, 2015 07:15

@tqbf That's great. It too has flaws, but it works very well for some companies. Not all. Same with in-person algorithmic interviews. gayle replied on Mar 09, 2015 07:26

@gayle Just to be clear: I’m arguing that in-person algorithm interviews don’t work well for ANYONE. tqbf replied on Mar 09, 2015 07:27

@tqbf The success of Google, Facebook, Amazon, Dropbox, and all these other companies is remarkable then, with such a broken process. gayle replied on Mar 09, 2015 07:28

@gayle @tqbf there's two ways to fail at recruiting - over-rejecting and over-accepting. Some can afford to over-reject. mik235 replied on Mar 09, 2015 07:55

@mik235 @gayle Notably, they can only fail to over-reject if their serious competitors ALSO fail to over-reject. tqbf replied on Mar 09, 2015 07:56

@tqbf @mik235 But you're arguing that their interview processes isn't better than random. gayle replied on Mar 09, 2015 07:58

@gayle @mik235 Shorthand: if GOOG AAPL FB AMZN literally used monkeys to select candidates, as long as they all did, they’d be fine. tqbf replied on Mar 09, 2015 08:00

@tqbf @mik235 That is a very strange position to take. I doubt a set of randomly selected candidates could build a system like google. gayle replied on Mar 09, 2015 08:02

@gayle @mik235 Once again: success of Google with poor process doesn’t vindicate process; nobody yet competing on hiring process. tqbf replied on Mar 09, 2015 08:03

@tqbf @mik235 Nearly all top companies have built excellent systems with people hired through this process. Good evidence of effectiveness. gayle replied on Mar 09, 2015 08:08

@gayle @tqbf @mik235 I have a tiger rock here. Never been mauled by a tiger while holding it. I'll sell it to you for lots-of-bux. edropple replied on Mar 09, 2015 08:09

@edropple @tqbf @mik235 That has about zero relationship to what I said... gayle replied on Mar 09, 2015 08:23

@gayle @tqbf @mik235 Top companies do X. Top companies do Y. Therefore, X implies Y. edropple replied on Mar 09, 2015 08:23

@edropple @tqbf @mik235 Didn't say X causes Y. But X clearly isn't getting in the way of Y. gayle replied on Mar 09, 2015 08:30

@gayle @tqbf @mik235 Can you prove that? edropple replied on Mar 09, 2015 08:31

@edropple @tqbf @mik235 Yes... Very successful companies use this process. Therefore, this process does not prevent building such companies. gayle replied on Mar 09, 2015 08:35

@gayle @tqbf @mik235 Asserting that you're at a local maximum is real hard when few are looking for better. edropple replied on Mar 09, 2015 08:37

@gayle @tqbf @mik235 I think we can agree that it works to an extent, but thats not evidence of effectiveness. There are conflating factors. edropple replied on Mar 09, 2015 08:37

@edropple Apparently @tqbf believes it doesn't even work "to an extent." No better than hiring at random, in his opinion. gayle replied on Mar 09, 2015 08:38

@gayle @edropple That’s true. I think it’s a net negative. tqbf replied on Mar 09, 2015 08:39

@gayle @edropple I’m not trying to make this argument easier to win for myself. I’m pretty sure I’m right. Interviews are terrible. tqbf replied on Mar 09, 2015 08:40

Person I’d like to debate: @gayle, regarding this comment, which I categorically disagree with: t.co/eAHngQzG7s tqbf tweeted on Mar 09, 2015 06:44

@tqbf "Categorically" is awfully strong, and logically impossible :). gayle replied on Mar 09, 2015 06:46

@gayle Sorry. I was imprecise. I meant, “there was no argument made in that comment I agree with at all.” tqbf replied on Mar 09, 2015 06:46

@tqbf So you disagree with "The interview process isn't perfect." AND "it actually works fairly well."? gayle replied on Mar 09, 2015 06:48

@gayle The interview process is fatally flawed, not merely imperfect. It does not work at all. tqbf replied on Mar 09, 2015 06:48

@tqbf Having done hundreds of these interviews AND talked with their managers, the people I identify as smart tend to in fact be smart. gayle replied on Mar 09, 2015 06:49

@gayle … meanwhile, the real problem is how many mispriced extremely capable people are rejected by the interview process. tqbf replied on Mar 09, 2015 06:53

@tqbf Yeah, that will happen with any interview process though. Fortunately companies don't use all use the same process. gayle replied on Mar 09, 2015 06:54

@gayle But Google doesn’t in reality benefit from hiring mainly people who are excellent at high-pressure interviews! tqbf replied on Mar 09, 2015 06:56

@tqbf The goal is not to select for that. On the contrary, Google trains interviews to be nice & supportive. Still some pressure, yes. gayle replied on Mar 09, 2015 06:59

@gayle @tqbf FWIW, I've interviewed at Google twice and neither time would I call it "nice and supportive." Maybe "brusque and jerkish." edropple replied on Mar 09, 2015 07:48

@edropple @tqbf Right. Google trains people otherwise, but unfortunately doesn't really exclude bad interviewers. gayle replied on Mar 09, 2015 08:00

@gayle @edropple Does Google 100% standardize interviews, so that interviewers A, B, and C all give the same interview X? tqbf replied on Mar 09, 2015 08:00

@tqbf @edropple No, it does not. 100% standardizing leads to bigger problems. gayle replied on Mar 09, 2015 08:04

@gayle @edropple Are the interviews *in any way* standardized, and if so, how? tqbf replied on Mar 09, 2015 08:06

Person I’d like to debate: @gayle, regarding this comment, which I categorically disagree with: t.co/eAHngQzG7s tqbf tweeted on Mar 09, 2015 06:44

@tqbf "Categorically" is awfully strong, and logically impossible :). gayle replied on Mar 09, 2015 06:46

@gayle Sorry. I was imprecise. I meant, “there was no argument made in that comment I agree with at all.” tqbf replied on Mar 09, 2015 06:46

@tqbf So you disagree with "The interview process isn't perfect." AND "it actually works fairly well."? gayle replied on Mar 09, 2015 06:48

@gayle The interview process is fatally flawed, not merely imperfect. It does not work at all. tqbf replied on Mar 09, 2015 06:48

@tqbf Having done hundreds of these interviews AND talked with their managers, the people I identify as smart tend to in fact be smart. gayle replied on Mar 09, 2015 06:49

@gayle … meanwhile, the real problem is how many mispriced extremely capable people are rejected by the interview process. tqbf replied on Mar 09, 2015 06:53

@tqbf Yeah, that will happen with any interview process though. Fortunately companies don't use all use the same process. gayle replied on Mar 09, 2015 06:54

@gayle But Google doesn’t in reality benefit from hiring mainly people who are excellent at high-pressure interviews! tqbf replied on Mar 09, 2015 06:56

@tqbf The goal is not to select for that. On the contrary, Google trains interviews to be nice & supportive. Still some pressure, yes. gayle replied on Mar 09, 2015 06:59

@gayle Interviews are an intrinsically hostile experience. One of the most hostile things normal people EVER experience. tqbf replied on Mar 09, 2015 07:00

@tqbf Absolutely agree with you there. gayle replied on Mar 09, 2015 07:01

@gayle You don’t expect them to do their day to day work with intensely elevated cortisol levels. tqbf replied on Mar 09, 2015 07:01

@tqbf It'd be great to somehow entirely remove stress from interviews. But how (without other consequences)? gayle replied on Mar 09, 2015 07:04

@gayle Respectfully referring you back here: t.co/YtCMS6DDWj tqbf replied on Mar 09, 2015 07:04

@tqbf Work sample tests/homework generally can't identify problem-solving/algorithm skills & can put unreasonable time burden on candidate. gayle replied on Mar 09, 2015 07:06

@gayle I’m not sure I’m clear on how any algorithm skill could be better demonstrated in a conference room than at candidate’s home. tqbf replied on Mar 09, 2015 07:07

@tqbf Because candidates cheat. gayle replied on Mar 09, 2015 07:08

@gayle Over all candidates — a decently large number, incl. everyone we didn’t hire — we had ZERO instances of cheating. ZERO. tqbf replied on Mar 09, 2015 07:08

@tqbf Then you're unaware or defining cheating differently. Give a candidate an algorithm problem and they'll google / ask friends for help. gayle replied on Mar 09, 2015 07:09

@gayle That’s not a work-sample test, it’s a trivia question. Work-sample must mirror the actual work candidates deliver. tqbf replied on Mar 09, 2015 07:10

@tqbf okay, so how do test algorithmic problem solving at home? gayle replied on Mar 09, 2015 07:11

Person I’d like to debate: @gayle, regarding this comment, which I categorically disagree with: t.co/eAHngQzG7s tqbf tweeted on Mar 09, 2015 06:44

@tqbf "Categorically" is awfully strong, and logically impossible :). gayle replied on Mar 09, 2015 06:46

@gayle Sorry. I was imprecise. I meant, “there was no argument made in that comment I agree with at all.” tqbf replied on Mar 09, 2015 06:46

@tqbf So you disagree with "The interview process isn't perfect." AND "it actually works fairly well."? gayle replied on Mar 09, 2015 06:48

@gayle The interview process is fatally flawed, not merely imperfect. It does not work at all. tqbf replied on Mar 09, 2015 06:48

@tqbf Having done hundreds of these interviews AND talked with their managers, the people I identify as smart tend to in fact be smart. gayle replied on Mar 09, 2015 06:49

@gayle … meanwhile, the real problem is how many mispriced extremely capable people are rejected by the interview process. tqbf replied on Mar 09, 2015 06:53

@tqbf Yeah, that will happen with any interview process though. Fortunately companies don't use all use the same process. gayle replied on Mar 09, 2015 06:54

@gayle But Google doesn’t in reality benefit from hiring mainly people who are excellent at high-pressure interviews! tqbf replied on Mar 09, 2015 06:56

@tqbf The goal is not to select for that. On the contrary, Google trains interviews to be nice & supportive. Still some pressure, yes. gayle replied on Mar 09, 2015 06:59

@gayle Interviews are an intrinsically hostile experience. One of the most hostile things normal people EVER experience. tqbf replied on Mar 09, 2015 07:00

@tqbf Absolutely agree with you there. gayle replied on Mar 09, 2015 07:01

@gayle You don’t expect them to do their day to day work with intensely elevated cortisol levels. tqbf replied on Mar 09, 2015 07:01

@tqbf It'd be great to somehow entirely remove stress from interviews. But how (without other consequences)? gayle replied on Mar 09, 2015 07:04

@gayle Respectfully referring you back here: t.co/YtCMS6DDWj tqbf replied on Mar 09, 2015 07:04

@tqbf Yes. Read it. gayle replied on Mar 09, 2015 07:05

@gayle OK, because to me, it’s pretty much all about that. tqbf replied on Mar 09, 2015 07:05

Person I’d like to debate: @gayle, regarding this comment, which I categorically disagree with: t.co/eAHngQzG7s tqbf tweeted on Mar 09, 2015 06:44

@tqbf "Categorically" is awfully strong, and logically impossible :). gayle replied on Mar 09, 2015 06:46

@gayle Sorry. I was imprecise. I meant, “there was no argument made in that comment I agree with at all.” tqbf replied on Mar 09, 2015 06:46

@tqbf So you disagree with "The interview process isn't perfect." AND "it actually works fairly well."? gayle replied on Mar 09, 2015 06:48

@gayle The interview process is fatally flawed, not merely imperfect. It does not work at all. tqbf replied on Mar 09, 2015 06:48

@tqbf Having done hundreds of these interviews AND talked with their managers, the people I identify as smart tend to in fact be smart. gayle replied on Mar 09, 2015 06:49

@gayle … meanwhile, the real problem is how many mispriced extremely capable people are rejected by the interview process. tqbf replied on Mar 09, 2015 06:53

@tqbf Yeah, that will happen with any interview process though. Fortunately companies don't use all use the same process. gayle replied on Mar 09, 2015 06:54

@gayle But Google doesn’t in reality benefit from hiring mainly people who are excellent at high-pressure interviews! tqbf replied on Mar 09, 2015 06:56

@tqbf The goal is not to select for that. On the contrary, Google trains interviews to be nice & supportive. Still some pressure, yes. gayle replied on Mar 09, 2015 06:59

@gayle Interviews are an intrinsically hostile experience. One of the most hostile things normal people EVER experience. tqbf replied on Mar 09, 2015 07:00

@tqbf Absolutely agree with you there. gayle replied on Mar 09, 2015 07:01

@gayle So, respectfully: why do you expect candidates to handle “new, hard problems” under those circumstances? tqbf replied on Mar 09, 2015 07:01

@tqbf Not all good candidates can (which is unfortunate). But enough can. gayle replied on Mar 09, 2015 07:03

Person I’d like to debate: @gayle, regarding this comment, which I categorically disagree with: t.co/eAHngQzG7s tqbf tweeted on Mar 09, 2015 06:44

@tqbf "Categorically" is awfully strong, and logically impossible :). gayle replied on Mar 09, 2015 06:46

@gayle Sorry. I was imprecise. I meant, “there was no argument made in that comment I agree with at all.” tqbf replied on Mar 09, 2015 06:46

@tqbf So you disagree with "The interview process isn't perfect." AND "it actually works fairly well."? gayle replied on Mar 09, 2015 06:48

@gayle The interview process is fatally flawed, not merely imperfect. It does not work at all. tqbf replied on Mar 09, 2015 06:48

@tqbf Having done hundreds of these interviews AND talked with their managers, the people I identify as smart tend to in fact be smart. gayle replied on Mar 09, 2015 06:49

@gayle I have to think about whether I can pluralize the word “hundred” in my experience doing this job, FWIW. tqbf replied on Mar 09, 2015 06:50

@tqbf Sorry, what'd you mean by that comment (the "pluralize" comment)? gayle replied on Mar 09, 2015 06:52

@gayle “I’ve done fewer interviews than you, I’d guess, but still have done many many”. Sorry. I don’t think you are a crank, not snarking. tqbf replied on Mar 09, 2015 06:53

@tqbf Have you gotten to correlate interview performance (for both good & bad performance) with job performance? gayle replied on Mar 09, 2015 06:54

@tqbf To be clear: I don't think this is THE right interview process. I think it's one that works well for certain companies. Not all. gayle replied on Mar 09, 2015 06:55

@gayle VERY YES. tqbf replied on Mar 09, 2015 06:56

@gayle I wrote the article that TechCrunch piece links to. Happy to answer questions about what we did. tqbf replied on Mar 09, 2015 06:57

Person I’d like to debate: @gayle, regarding this comment, which I categorically disagree with: t.co/eAHngQzG7s tqbf tweeted on Mar 09, 2015 06:44

@tqbf "Categorically" is awfully strong, and logically impossible :). gayle replied on Mar 09, 2015 06:46

@gayle Sorry. I was imprecise. I meant, “there was no argument made in that comment I agree with at all.” tqbf replied on Mar 09, 2015 06:46

@tqbf So you disagree with "The interview process isn't perfect." AND "it actually works fairly well."? gayle replied on Mar 09, 2015 06:48

@gayle The interview process is fatally flawed, not merely imperfect. It does not work at all. tqbf replied on Mar 09, 2015 06:48

@tqbf Having done hundreds of these interviews AND talked with their managers, the people I identify as smart tend to in fact be smart. gayle replied on Mar 09, 2015 06:49

@gayle … meanwhile, the real problem is how many mispriced extremely capable people are rejected by the interview process. tqbf replied on Mar 09, 2015 06:53

@tqbf Yeah, that will happen with any interview process though. Fortunately companies don't use all use the same process. gayle replied on Mar 09, 2015 06:54

@gayle Also: remember, your experience is (in large part) as Google, which has a very distinctive hiring pipeline… tqbf replied on Mar 09, 2015 06:55

@tqbf Sure. But I'm not arguing it's perfect for every company. Just that it's possibly the least bad for some. gayle replied on Mar 09, 2015 06:56

Person I’d like to debate: @gayle, regarding this comment, which I categorically disagree with: t.co/eAHngQzG7s tqbf tweeted on Mar 09, 2015 06:44

@tqbf "Categorically" is awfully strong, and logically impossible :). gayle replied on Mar 09, 2015 06:46

@gayle Sorry. I was imprecise. I meant, “there was no argument made in that comment I agree with at all.” tqbf replied on Mar 09, 2015 06:46

@tqbf So you disagree with "The interview process isn't perfect." AND "it actually works fairly well."? gayle replied on Mar 09, 2015 06:48

@gayle The interview process is fatally flawed, not merely imperfect. It does not work at all. tqbf replied on Mar 09, 2015 06:48

@tqbf So you think Google and Facebook would be equally successful hiring at random? gayle replied on Mar 09, 2015 06:48

@gayle Do I think it is literally worse than monkeys throwing things? I’d have to think about that. tqbf replied on Mar 09, 2015 06:49

@gayle I guess, interviews vs. monkeys, would depend a lot on the top of the funnel and how outreach is done. tqbf replied on Mar 09, 2015 06:49

@gayle For some funnels, yes, I think monkeys might actually outcompete. tqbf replied on Mar 09, 2015 06:49