Studying in Germany part 1

Studying in Germany part 1

Image credits: Saarschleife by Christopher Jenal

Some comments about Germany’s education as I experienced it.

  • A short introduction
    Hi, I’m Karl, a Computer Science student currently studying in Germany’s Saarland, and I’d like to provide a little essay on my experience of the first semester as a master’s student in Saarland- Germany. I’m convinced that this will be of help to other students, mainly internationals which are contemplating a study career in Germany as I get tons of questions regarding it.

When I first thought about studying in Germany in 2014, I did a lot of online research and found little information in general. At least in narrative styles like blog posts and first-hand information from students. Most of the information I found was from comments on places like Quora, Reddit, and other online forums. But again very limited in detail.

So with this post hopefully I will provide a little more detail on a variety of subjects on the topic of studying in Germany. Let’s dive in.

Application & Enrollment

As you might have heard, Germany has a reputation for having a remarkable bureaucracy. At least that was what people always told me, and resulted being accurate. As my Romanian friend said, he thought the application in his home university was a mess but compared to this it was a piece of cake. But more on that next lets first start with the basics.

The enrollment process is different from university to university. Therefore assuming you are a student interested in applying to a University in Germany, the first thing to check out should be the admission page of the University of your choice. You will find the required documents, grades (GPA)[]) and dates to reach in your application. I sent in my documents about six months earlier and got admitted about five months before the start of the semester; that gave me enough time to organize my travel and place to stay. It was a little challenging to get a room at one of the Studentenheime (student hostel) though. I guess because there are a lot of international students in Germany the queue is long, but eventually, you will get something just make sure to start searching earlier. To start just google something like Studentenwohnung or Wohnheim + [Location of your University].

Once I arrived at the welcome center of the University, I got a paper with the necessary steps to prepare my documents, enroll, get accommodation, register at the Administrative Office for Citizens, and open a bank account.

List of all practical steps in order

  1. Documents needed:
    a. A copy of your letter of acceptance
    b. Passport picture
    c. A scan of your passport
    d. Scan of all diplomas

2. Online enrollment

Enrollment was an unnecessary hassle in my opinion because we had to resend 90% of the documents we already had sent in on our application via the Internet. Wouldn’t it be better if the University just forwarded the already present documents to all required departments instead of forcing the students to submit the same documents over and over again at every campus department?

I had some issues because there were outdated links in the admission paper which I had received previously. I got a lot of confusion about where exactly to upload the documents as the instructions were somewhat contradictory and I didn’t find a single place (page) with clear instructions to follow. After having resolved my issues thanks to the very helpful lady at the welcome center I finally managed to upload all my documents in the correct online form for the enrollment process to proceed.

3. Accommodation

Ideally, you should’ve already applied to a room at the Studentenwerk from your home country, if this is not the case just get some guidance at the Accommodation Service (Welcome Center).

4. Registering at the Administrative Office for Citizens

Once you obtained your accommodation, you have to visit the city registration office (Bürgeramt) to register as a citizen. Make sure to take your passport or ID card with you as well as your rental contract.

5. Opening a bank account

To open your bank account, you need your passport, proof of registration at the city and student card.

7. Visa / Residence Permit

The Bürgeramt will transfer your information to the visa office once you’ve registered there. The visa office will contact you by mail with details about the appointment to extend your residence permit. If your visa expires within 3-4 weeks, ask for an appointment at the Administrative Office for Citizens or the Welcome Center of your University.

As you might have seen these steps need to be done in exact order because they are interdependent, if you get stuck in one step you will be delayed in the rest.

Academic experience

Overall I was very satisfied with the educational content. Class material was very up to date, covered a broad range of subjects while also going into great depth. My only complaint here is that some of the tools used in AI class, in my opinion, were not industry standard, and suboptimally documented which made them tedious to use. I will try to give you a better picture of this digging a deeper in two of the classes for which Saarland University is famous for, AI (Artificial Intelligence) and Security.

From my previous research, I knew that AI was one of the areas for which Saarland University is famous for; later I also learned that it is one of the best Universities in Security (anecdotal evidence) while drinking some beers on the night bar at one of the local Weinheim’s. Academic ratings are not always accurate/representative especially on international rankings; German Universities tend to miss points because a lot of the research is done at research institutes.

Even if these research institutes are located on the University campus as it is often the case, the University rarely earns the credits from them because research papers get signed with the name of the research institute instead of the name of the university. So even if there are tons of quality research coming from the University, the International Rankings may not reflect this fact. At least this is what I learned from some information digging on the Internet. If someone knows more details on this topic, please let me know in the comments.

Artificial Intelligence

The AI course has three main segments; the lectures, Homeworks, and modeling sheets. Lectures typically lasted 1:45 hours I guess this is pretty much international standard. The lectures are excellent explaining the concepts in depth with high-level examples and more practical algorithms accompanied by images. The professor made great use of quizzes to make the class more interactive. Only occasionally slides were somewhat ‘overloaded’ which made them harder to read and of less quality in my opinion. I found that a lecture at King’s College was substantially more digestible. Of course, we got into more details like proofs and the like, but certain slides were a little too much for my gusto as I live by the motto: “Simplicity is the ultimate sophistication.”

And this is perhaps one of the things which differentiate German or perhaps more generally, European Education from the academics in the US. Although I didn’t study personally there, I did participate in a lot of Online Courses from Universities like Stanford. I feel like I should dedicate an entire Blog Post to this topic.

Don’t get me wrong the education is superb. The professor of the AI lecture is one of the top experts in this field. His Ph.D. even is mentioned in the “Bible of AI” due to his contributions in Planning a crucial branch of AI that involves the realization of strategies or action sequences. The course dealt with the very leading edge of research. So I don’t feel the need to babysit the strengths and will rather dig into some critics because that is where after all, improvement can happen.


You might have heard about the ECTS credit points. Basically, its aim is to “make programs and the performance of students of higher education more transparent and comparable European-wide and to replace or complement the different local (national) standards within Europe.” In other words, it is a rough estimate of the time required to earn one credit point. The workload per credit point varies somewhat between countries despite the effort of making a standard. In the case of Germany, it happens to be close to 30 hours per point.

Let’s do some math, the core lectures at Saarland University have 9 credit points. This would mean 930 = 270 hours of workload per core lecture in theory. With a semester duration of 12-14 Weeks, students have to dedicate about 20 hours per week of their time on that lecture. In the case of AI at least from my experience, this is a vast underestimation because merely the weekly homework sheets took that amount of time. Not bearing in account the modeling sheets which were absolute time killers. Basically, they consisted of solving a series of problems with some SAT solver, and every week the solver to be used changed. It would have been less painful if those somewhat exotic solvers had a better documentation, but usually, the documentation was obscure and of little use. Students had to do a lot of guesswork or experimentation, and that in itself took up a substantial amount of time. Time which students who need to make roughly 30 credit points per semester just don’t have. Theoretically, the math adds up correctly, 30 points * 30 hours divided by ~13 weeks ==> (3030)/(13*7) = about 10 hours per day.
But in practice, the story is different. Sometimes you can get stuck with exercises. You might ask yourself how students even manage to complete the weekly assignments in time? Well, many don’t, and my guess is what saved the majority was the fact that we could build groups and distribute the work among its members. Groups could have a maximum amount of three participants, dividing the work in 3 made the time more manageable. Poor students which didn’t make it into a group. I’m not sure if this is done on purpose in order to enforce “collaboration” but in my opinion, students should be able to complete the course on their own. Dividing the work also means that it is likely that not every student gets familiar with all of the content and thus the purpose of having that homework in the first place is flouted.

The exercise sheets although being a good way to practice, were time-consuming and often tedious as we frequently had to compute algorithms by hand. Personally, I get very bored by doing algorithms by hand as I think this is the task of computers. Humans should try to use the creative power of their brain to create and improve those algorithms. But I get that this might be harder to grade so probably that’s why performing algorithms by hand still form a significant part in CS education nowadays.

Modeling sheets were even more tedious as they involved dealing with extrinsic tools and languages which as mentioned were difficult to get familiar with and often felt more like a series of hacks rather than mature frameworks. As one fellow student said, “If this is the industry standard than God help all SAT Solvers and Planners.” The idea of learning to use tools for SAT solving and Planning in itself is something I strongly support. But I think the focus should shift to a place where we develop intuitive and useful open source frameworks preferably hosted on a platform like Github. Doing that would help start growing a bigger user community which ultimately would contribute to the creation of comprehensive user guidelines, documentations, and overall improved tools into mature projects with potential use around the world. Rather than little-known tools buried somewhere on local university file servers, destined to hardly ever grow, improve and become simpler to use. Perhaps I’m being a little too idealistic, call me a dreamer I’m ok with that.


As for the security course. Basically, we covered the most important papers from the beginning of relevant security implications of the information age to the top of newest discoveries. In one occasion the professor apologized because the most recent paper he touched upon on lecture wasn’t yet published. But dealing with papers which were only weeks old was the norm through the course, which I loved. The only pain was a cryptographic protocol analyzer named Proverif, which is an impressive tool in itself but the documentation is poor and hence once again we had to choose between perform brute force guessing and reading a 50 page or so PDF documentation, only to end up confused and lose track of things. So most of the time, guessing until it worked was the preferred strategy.

The rest of the course was great fun. Homework 1 consisted of performing buffer overflows, cross side scripting, cross side request forgery, SQL injection, writing a worm and making a phishing attack. In homework 2 we had to use Proverif to analyze different types of protocols and scenarios. For example trusted server authentication, commitment schemes, prove Observational Equivalence and anonymity by using zero-knowledge proofs.

I would rate the difficulty of the homework as challenging based on the comments of other students as well as my own appraisal.

Grading: Norm-Referenced Grading at it’s best?

The AI exam consisted of about 6 pages of exercises which weren’t that difficult but rather time consuming. With a hard time limit of two hours, there was little time for hesitation. Students had to compute every algorithm by hand quickly to finish on time although the difficulty was moderate.
One unexpected thing happened, though. After the exam evaluation, it was announced that the minimum amount of points to pass was changed from 50% to 60% resulting in the following mapping of points-grades.

= 97.5: 1.0
= 95: 1.3
= 92.5: 1.7
= 90: 2.0
= 85: 2.3
= 80: 2.7
= 75: 3.0
= 70: 3.3
= 65: 3.7
= 60: 4.0
< 60: failed

Many students were caught off guard by the change because everyone expected 50% to be the threshold for passing the exam. The cause of raising the bar was that the exam was supposedly easier than in previous courses and therefore not raising the bar would be “unfair to other students”. Let me quote an excerpt from the professor:

It is very difficult to design exams at the same level of difficulty every year. Usually it more or less works out, and we set the passing threshold to 50%, like most other courses do as well.

However, this year, our design of the exercises turned out to be substantially skewed to the easier side. The point distribution is very clearly and significantly higher than in previous editions of this course. To be precise: The point average of the first exam in last year’s edition was 48.72. The point average in this exam is 67.78.

So what do you think is fair? Let everyone with 50% pass just so that we stick to “half of the points”? Or adapt the scaling of grades to achieve at least some level of consistency across course editions?

To me, the answer is very clear. With a difference of averages of 19 points, raising the threshold by 10 points is definitely justified.

As expected the change caused a wild debate in the forum. Many students expressed their disappointment with this adjustment. The frustration is comprehensible as the problem with relative grading is that it depends on how the faculty interprets the performance of the individual relative to previous courses and/or other students. Despite an apparent lack of transparency in the grading scheme, Norm-Referenced Grading was clearly used. In this case, norm meaning the “normal” performance or difficulty of the course through history. In my opinion, statistics like grade distributions/histograms from the actual course and previous courses should be publicly available since this would confer the possibility for students to evaluate the relative grading adjustments made by the faculty and point out potential flaws in the system.

Personally, I don’t like relative grading or grading on a curve. In my opinion, it doesn’t help motivating students but rather has the contrary effect. I recommend reading this article for further details from Psychology Professor Dr. Marty Covington who has spent several decades investigating the effects of classroom competition on academic motivation at the University of California, Berkeley.

Covington believes that every student should have the opportunity >to earn an A. This does not mean that they will, but that they have >the opportunity to. Precisely that is what grading on a curve prevents from happening.

And with regards to relative or norm-referenced grading towards other students and courses, is this really an effective strategy? Consider a scenario where we have three hypothetical students, Ed Eda and Edy. These tree students participated in a course where only one of them, Eda, happened to pass the exam. The next semester we have tree students again, but this time, for whatever reason, all three of them happened to be exceptionally talented. All three students successfully complete most of the exercises and confidently wait for the next big challenge. But hey, wouldn’t that be unfair towards Ed Eda and Edy? They too worked hard but only one of them succeeded in passing the exam. The faculty decides to raise the bar and ditches two super gifted students who were aiming high in order to pursue their dreams. After all, this will cause them to be even more competitive and the University to be more >elite< right? As passing the course represents a “big challenge” with only a few possible winner. I hope your sarcasm detector rang while reading some of the last few sentences.

Hopefully, my little short “novel” seems as senseless to you as it does to me or at least my point became clear. In my opinion, education should be based on learning standards i.e., descriptions of what students are expected to know and be able to do at a specific stage of their education. Of course, I imagine this to be challenging, but not impossible, with some good collaboration with the industry perhaps, it is certainly achievable. Packing students in a funnel in order to let them sort out into predestined categories in sort of an artificial natural selection process is outdated in my opinion. I envision education as consisting of more inspiring nature not merely competing individuals. And I don’t mean that competition is bad I think it can be very helpful, and a big motivator. But not in all cases and for all students. Many of them are driven by curiosity and inspiration and don’t give much importance towards competition.

And doesn’t it make more sense to prepare students to actual verifiable standards out there instead of letting them compete against others in a game where a percentage is always predestined to lose? I bet most of those potentially gifted students who didn’t pass the exam, would rather do something more interesting than doing a reexamination or repeating the course.

Student activities

Parties which were organized by the AStA (Allgemeine Studierenden Ausschuss) did rock. AStA is a group of students who get elected every year to engage in representative activities for Saarland Universities students. They do a great job in a wide range of areas, and the parties were one of them. I think there were only two parties in the semester, but they were extremely well organized, with different stages, featuring different styles of music.

Also, I saw great political, social and academic engagement from them through flyers although I didn’t take the time to dig into that so I can’t comment in more detail.

Non-Academic Activities

There certainly is no lack of activities in Germany. One can easily stay up to date on all the local events through the Internet, for example with dedicated Facebook pages containing such events. What I did is connect the Calendar application on my Mac with Facebook so that all events automatically get downloaded so that every day I receive a notification about today’s events on my screen (Thanks Apple for making our live easier).

But even though all of this, I find myself spending lots of time dedicating to myself, in what some would call self-improvement. Be it reading, listening to podcasts or audiobooks, watching talks on YouTube or brainstorming ideas about my present ambitions and future goals. I never get bored as long as I have books or an Internet connection. The latter providing the books, perhaps I should name a few of my favorite time-effective methods of consuming information as lack of time is people always complain about when asked about those topics.

Not having time to read should NOT be an excuse. There are plenty of methods like the smartphone. You can install Podcast and Audio applications some of them freely available like Stitcher, other ones available with reasonable subscription fees like audible and just listen to your books/podcasts while you commute. I use to listen to audiobooks while I’m on the way to University or the supermarket. Every time I go somewhere even if it is a just a 10 minutes walk, I tend to take advantage of that time and listen to my stuff.

At night I often go for a walk, while listening audiobooks. I like to adjust audible to 1,25 or 1,5 speed to get more information in less time. Depending on the complexity of the book and the natural speed of the narrator you will have to gauge the ideal speed for yourself. In my experience, listening at a faster rate is still very understandable in general. Taking walks is one of the most stimulating activities I found in this country. I fell in love with Germany’s forest ever since I visited this country in 2014. The fresh summer nights tend to be around comfortable 12-16 degrees Celsius, perfect for outdoor activities. The fact that the sun sets late, approximately 21hs adds to this even further. Something which I wasn’t used to coming from a Latitude closer to the equator.

If you prefer the big city life, I would recommend you to apply for a university located in Munich or another one of the bigger cities. Just take into account that the cost of living can be substantially higher there. As you will realize when you visit, even though Germany has a population of 80+ million, it is spread out evenly. The effect is that population density appears to be not that high. Most of the Towns are of a small to middle size and it is likely that your university will be a little outside the city in a quieter area. I like it because life is more relaxed that way and I get to breathe this beautiful fresh air emanating from the forest all around.


There are many things I could expand on a lot more, but I prefer to wrap up here and let you decide the topic I should talk about next. Just let me know in the comments. And feel free to support this Blog by the means you choose. This would help and motivate me a lot. Cheers


Karl Niebuhr's Picture

About Karl Niebuhr

Karl is a passionate reader, thought leader, author, learning enthusiast, ex knife combat champ, student and general entropy reduction maniac

Saarbrücken Germany