Are you happy being a Physicist?

[twofish-quant]

Now you have 200-300 papers per scientist which are totally useless. I mean you publish one paper that says: "with x=1 and y=2 you get z=3" and then you publish another identical(!) paper that says: "with x=1.1 and y=2.1 you get z=3". I mean what the hell?

In the case of theoretical astrophysics, "me too" papers are extremely important. All computer programs have bugs, and having another team try to replicate your results with a different code base is pretty essential to figuring out what is going on. What happens invariably when you run two (or more) different simulations is that the results don't match completely and part of the point of doing independent runs is to figure out which results are "robust" and which ones are sensitive to the simulation.

If you don't find out sth very important you shouldn't publish a paper about it.

Except that you don't know if it's important or not. The other thing is that most of astronomy are "I pointed my telescope at object X and I saw Y."

One problem with "publish only if you find something important" is publication bias. For example (and this is a real example), you do a statistical test to either look for a particle or see if a drug cures cancer. You run into major, major problems if only "important" results get published. Because the one lucky (or unlucky) experimenter that cures cancer by some statistical fluke publishes where as the 999 people that "find nothing" don't.

Another problem is that if you publish 70% of the submissions, then the editorial boards really don't have that much power that they can abuse. If you publish 1% of the submissions then then editorial boards become super-powerful and then can just kill any line of research that they are ideologically opposed to. This is a big problem with economics, management, and finance journals.

Then there is the "dumb luck factor." You just can't decide where nature is weird. The people that showed that the universe was accelerating did really solid work, but it would not have been a spectacular result if it turned out that the universe was working the way everyone thought it would.

If it's helpful for other scientists who do the same experiment you just can post you results on a website.

You mean like

http://adswww.harvard.edu/ and http://arxiv.org/

Astrophysics publication went online about a decade ago.

I'm ok if supercollider won't show us anything important because it happenes.

One thing about LHC and the supercollider is that they are *going* so show us something important. If LHC does a search of energies between 100 and 200 GeV and finds no Higgs boson, that's extremely important information.

Similarly, if LIGO doesn't find any gravity waves that's an extremely important result.