Symptoms of Success ... or of Achievement

I realized early that controlling the things that I could control
gave me greater freedom to address the things that I could not control.
Markt Twight (famous mountaineer)

Inevitably, in your career, you will hit the pea-counters. They will look at your CV and will search for a quantifiable weakness. Similarly, there will be comparative evaluations of you with your peers where ‘’some'' people will look at numbers. The items which I summarize here as “Symptoms of Achievement” give them an easy way to look at you. By ensuring that these quantifiable symptoms are within the normal range, you can make sure that your supporters have an easier time helping you and your opponents a harder time of overtaking you in the race. Henceforth, controlling these numbers carefully is important.

However, also don’t fool yourself! Having great numbers does not guarantee anything. People with terrible symptoms have been very successful - both amazing and terrible scientists. For example, the incremental scientist, who never moves far from his Ph.D. thesis and only searches for simple additional applications of the method which he developed (ideally with only 2 to 3 degrees of freedom), has an easy job of maximizing his or her key figures while never soaring scientifically. However, researchers who are good in their own field usually look beyond such silly arguments as whose publication list is longer or better. They will quickly realize whether the Research Agenda of a scientist is sound, whether the scientist is capable of a bigger picture combined with his own vision (or understands only the tiny project which he is pursuing) and can really tell the whole narrative from a new idea to an impressive application. The incremental scientist on the other hand is usually not taken serious by such kind of guys…

Obviously, the paragraphs above indicate that every place where you interview will have a different weighted mix of these indicators. For example:

  • The top 10 US universities will barely look at these numbers and really trust the recommendation letters (both for hiring and tenure decisions).
  • The other 600 US universities will potentially even tell you what key numbers they expect from you.
  • Places like the TU Munich or the KIT in Germany have often done their decisions with the measuring tape to simplify counting (Hint: always make sure your publications in your publication list are numbered! Same for talks, media appearances, etc.)
  • Numbers in the UK need to be optimized for their national evaluation and the candidate list is frequently optimized on that measure. Here, a single Neuroscience/Medicine publication in a 10+ impact factor journal will allow you to make it to the top candidate … even if you only did the statistics as a helper (and irrelevant co-author) or it was only thanks to your 10 year old Master’s thesis that you had this publication.
  • French places frequently only count journal papers for hiring.
  • Canadian places often completely discount conference papers in tenure decisions.
  • People who are aware of the quality of conferences will give an unjustifiably high weight to ERA A+ conferences … or sometimes just to their personal favorites among these.

That in mind, let’s start with the numbers.

Publications

We all want publication success. What success does really count?

  • ‘''Journal Papers:’’’ With high likelihood, you will need journal papers (note: at the US top 10 schools, they are somewhat less relevant than ERA A+ conference papers, but in most other schools they are quite crucial for tenure). Ideally, these journal papers should have an impact factor of 2 or higher … and you may wanna discount journals with am impact factor below 1. Stefan Schaal once told me to aim at a career average of about one to two journal paper per you (before tenure). Obviously, you can’t do all of these alone. While it is key that publish 3 to 5 journal papers as first author, you can also start documenting your ability to supervise research by last author journal papers. Importantly, figure out what journals are relevant for your long term agenda; our list on Impact Factors? may potentially help. Additionally, make sure to get as many little extra journal articles as you can … your friend the medical doctor or the neuroscientist can make your start in Britain much easier…
  • ‘’’Conference Papers:’’’ There is no doubt that computer science cares about conference papers. Particularly, just a few ERA A+ conference papers can get people to notice your research and bring you up to the top of the your field. Look at Conference Quality to check what conferences could potentially matter. However, do note that only your first couple of NIPS or R:SS papers will give you a huge boost, the returns quickly diminish if the conference is not “your community”. Hence, it is crucial that