And OUCH on “referee 1″ = “Prof Joe Doe”. That kind of behavior just sucks ass.

The post Julianne kindly linked to is very much intended to be tongue in cheek - that time of year, y’know. But, as with most attempt at humour, there is also some truth to it.

For what it is worth, the comments I chose are illustrative but not verbatim examples of some actual extremes. The translations are also not all that far off the mark…

The refereeing doesn’t get to be unreasonable until you realise that “referee 1″ was actually “Prof Joe Doe” in the second example.

There are agencies that are not the NSF where a 15% funding rate would be wonderfully high. And, yes, most good proposals can not be funded.

This is not the situation at all in any of the review panels that I’ve been on. It is invariably the case that there are many proposals that deserve funding that can’t be funded because of the ~15% NSF funding rate. With such a small fraction of proposals to be funded, almost any tiny error in a proposal will get it rejected. Plus, many good proposals with no errors still won’t get funded because the acceptance rate is so low. Basically, your proposal has to be judged to be really exceptional to get funded, and many very good proposals are rejected without anyone necessarily being shafted.

And amused — carry on with the discussion as you wish!

For example, a black hole cosmology proposal related, or relatable, to quasinormal modes could mention how Shahar Hod’s work has (AIUI) overturned earlier studies concluding that these always evolve to the same form.

Postdocing does not equal serfdom. When it does, this is a problem with either the standards of treatment in the field (for example, biology postdocs have less independence than astronomy postdocs) or the behavior of individual PIs. It is difficult to change people’s behavior and there are relatively few consequences to people who behave badly though. I fantasize that granting agencies should somehow compile and anonymize followup interviews for grad students and postdocs that they have paid for.

translation: 12 to 19.2 hours a day, 7 days a week.

But, the mistake is thinking that all science funding agencies are focused on curiousity driven blue sky science, most are not.
In particular the major funding agencies for the physical sciences are mission driven to a large extent. This generally implies there are sequences of tasks that need to be carried out by people with particular expertise - for example there may be a need to build some equipment, or to analyse a particular data set, or to model a particular scenario, or to attempt to calculate a known unknown.

Often such tasks are beyond the capabilities of graduate students, at least on the time scale the problem is to be done on, the expectation is that the research requires someone with a PhD. Arguably the faculty involved should do some of these tasks themselves, rather than hire postdocs to do so, but often the labour involved in doing the task on the time scale required is more than is available from the faculty PI - some things require a 50-100% time committment to be done in a timely manner. (Buying out faculty from teaching and service duties is possible but discouraged in many cases and cost-ineffective in other cases).

So… there are grants for postdocs to work on specific pre-defined tasks.
Typically they require 50-80% time commitments, notionally, the rest of the time being available for exploratory independent research.

We can argue about whether this is a good system, but right now that is the system the US uses. It is relatively cost effective and flexible, it is arguably damaging in the long run.
In the short run, faculty apply for grants to hire postdocs, and soak up a large part of the supply of recent PhDs in the process.
There are worse things than being a postdoc.