[s3nnyy]

I live in Zurich and I used to code for a living. Now, I hire engineers for different startups in Switzerland.

As I got deeper into IT-recruiting, I realised that candidate filtering at the top of the funnel is fundamentally broken. Especially in Europe companies expect a CS degree and don't appreciate self-taught skills as much as in the US.

I am trying to change this. If you look for a tech-job in the most liveable city in the world, check out my story "8 reasons why I moved to Switzerland to work in IT" on http://medium.com/@iwaninzurich/eight-reasons-why-i-moved-to... or send me a mail to the mail-address in my HN-profile.

[_yosefk]

Could the difference be due to education being cheaper and thus there being fewer reasons to forego it?

[s3nnyy]

More people have degrees in Europe as it does not cost money. Also, the standard of education dropped and the exams are easier to pass.

I know from experience that a B.Sc. in Computer Science at the Technical University of Munich (considered a very good school) was harder to get before 2008 than it is now.

In the killer-subjects like discrete mathematics, theoretical computer science and algorithms we had failing-rates of 80%.

I can image the German government asked for "more CS grads" and after this the professors started to introduce one problem in each exam that anyone could pass who practised a bit and / or wasn't entirely stupid. The other two to three problems were still on a high level, and if you wanted a top-grade, you still had to solve those. As a result the failing-rate for the mentioned killer-subjects dropped from 80% to 20-30%.

[js-coder]

> As a result the failing-rate for the mentioned killer-subjects is dropped from 80% to 20-30%.

Do you have any links to back that up?

I'm not studying at TUM but I'd be really surprised if there's a 80% pass rate for those subjects.

[s3nnyy]

Around 2008 Prof. Mayr stopped teaching discrete mathematics. After that in most of the midterms there was a solvable problem. Often it was something like "Use the Dijkstra's algorithm to find the shortest paths in this graph. 10 Points" (and the exam had like 32 points).

[pc86]

> the professors started to introduce one problem in each exam that anyone could pass who practised a bit and / or wasn't entirely stupid.

What is considered a passing grade at one of these schools?

At my private liberal arts college in the US it was usually (but ultimately up to professor discretion):

  00-59% - Fail (0.0 for GPA)
  60-67% - D    (1.0)
  68-69% - D+   (1.3)
  70-72% - C-   (1.7)
  73-77% - C    (2.0)
  78-79% - C+   (2.3)
  80-82% - B-   (2.7)
  83-87% - B    (3.0)
  88-89% - B+   (3.3)
  90-92% - A-   (3.7)
  93+    - A    (4.0)

You could count any class with a D or better toward graduation, but to actually graduate you needed a 2.0 or better GPA. If you did not have it you could take additional courses to bring your GPA up, but adding on 3-credit courses when you've got 120 credits built up with a sub-2.0 GPA is typically a losing proposition. Most transferred if they were sub-2.0 by the end of their Sophomore year.

I've seen other grading scales with E's in addition to/instead of F, or minor variations on the percentages. It seems popular to give Honors courses an extra point (e.g. an A- is 4.7 instead of 3.7), particularly in US High Schools.

[schwarzmx]

Works similarly in Germany, with the difference that 4.0 is the lowest passing grade and 1.0 is the highest.

See https://en.wikipedia.org/wiki/Academic_grading_in_Germany

[schwarzmx]

I don't see how one professor makes the entire program.

FWIW I did my masters there and in my opinion the way they evaluate students is suboptimal anyway - but I don't know if that's a TUM thing or a German thing.

[s3nnyy]

He doesn't; I just gave the example of one course but the pattern happend for other courses, too.

[saiya-jin]

similar story of "greatest filter in whole CS 5 year path quietly removed" here, albeit ours was actually non-programming related. It was theoretical electronics, don't ask me why we had to go through that guy - actually other students, focused on electronics had much nicer guys. the guy was firing people out of school with passion (literally for a single dot missing in Fourier equations at one place, often with deep personal insults, nobody liked him, not even from professors.

yeah, my uni was pretty bad, all the useful stuff I learned and use daily came from my own learning, none from university. campus was a fun place and great experience though :)

[jacquesm]

No, it just saves a bunch of time and you later won't hear (when someone messes up) 'he doesn't even have a degree, what did you expect'?

It's the recruiting equivalent of buying IBM (or Microsoft I guess).

If the applicant pool is large enough any quick way to discriminate that skews positively for applicant capability is going to be used. It is also much easier to see if someone has a degree than to actually test if they have the required skills.

[pc86]

> If the applicant pool is large enough any quick way to discriminate that skews positively for applicant capability is going to be used.

I think a lot of self-taught developers (myself too earlier in my career) fail to realize this.

If you've got 1,000 applicants and your job is to schedule 10 interviews from that pool, you will do anything that won't absolutely destroy candidate quality. It's not about the finding the best, it's about finding someone who wants the job and will be able to do it. The name of the game is not minimizing false negatives (disregarding good people), but minimizing false positives (interviewing shitbirds).

So if you work for BigCo you say the person need three years of corporate experience. You're down to 800 applicants. One year of experience in the JavaScripts. 750 applicants. Maybe the phrase "SQL Server" has to be on their resume. 400 applicants. College degree. 320 applicants. Computer Science degree. 120 applicants. Maybe you want to filter by your preferred recruiter, because they only charge you 11% of base salary instead of 18% like that other firm. 22 applicants. Now you can actually read the resumes and pick the top half to interview. Anyone who actually wrote a cover letter and is in this pile is pretty much guaranteed an interview.

[_yosefk]

That might make sense, but why the difference between Europe and the US, if in fact it exists as the grandparent claimed?

It's true that it's costly and hard to test one's skills. The trouble with grades though is some people cheat brilliantly on their tests and others, some of them really good, test badly (if they're bored they don't prep, screw the grade average, etc) and some of the latter are actually extremely pragmatic on the job. I don't quite understand it but it's there.

[simi_]

Hey Iwan - I just shot you an email over at your fastmail address. Looking forward to start a discussion there.