Submit your papers where they are likely to be accepted

It came to me tonight. A simple idea, really.

All scientists should submit their papers where they are likely to be accepted. Oh! We all need a critical review. Sometimes we even need to be told to rewrite our paper and come back. However, scientists should not play Russian roulette.

Sending a paper to a venue where you…

  • need to include or exclude certain words or concepts;
  • have a good chance (say 50%) to see your work turned down (irrespective of this quality);
  • must cite certain people;
  • have to exaggerate your results;
  • must force yourself to be more fashionable

is just wrong. We should never do it.

We should work hard. Do our very best to write excellent and honest papers, then ship them and have them published (after some revision). This should be the normal process. Period. Nothing more, nothing less.

We need less boxing, more science.

Note: I have had good luck getting my papers published this year, so I do not write this post out of spite.

Published by

Daniel Lemire

A computer science professor at the University of Quebec (TELUQ).

6 thoughts on “Submit your papers where they are likely to be accepted”

  1. My professor always insists on submitting papers to weird journals, he is a full professor in his 70s so he is after very classy journals! and our papers get rejected! and he replies a little more improvement!!

    PS. I do research at Berkeley

  2. “have a good chance (say 50%) to see your work turned down”

    I’d say this rule is too conservative? The top conferences in my subfield (network security) accept 10-15% of papers. I’m a relatively new researcher, but I get about half of my papers accepted to such places; most people consider this excellent.

  3. I’d restate your position as:

    * avoid venues with arbitrary acceptance criteria

    * submit work that you feel would be accepted on the merits unless there simply weren’t enough slots

    It seems to me that, if the reviewing process is reliable enough, you should be able to get a good predictor of acceptance from your peers. And, if it isn’t, that should decrease the expected ROI for submission.

    Of course, the problem is that, if a prestigious venue fails this test, then you may still be incented to submit to it despite the lower probability of acceptance–and, collectively, researchers perpetuate the prestige of such venues. It’s a prisoners’ dilemma.

  4. Also there is an interesting issue with point 2: do you mean a 50% acceptance rate or that your subjective appreciation is that you have 50% chance of getting in.

    I mean the latter.

  5. I’m not sure why you relate the acceptance rate to the other 4 points.

    It seems to me that you would want to avoid your points 1, 4, 5, and to a lesser extent 3, irrespective of the acceptance rate (point 2).

    In fact even that may be dependent on your academic situation. You (and I) can conceivably ignore more prestigious venues if we are uncomfortable with some of the reviewing or acceptance practices. Not so for Grad students or even post-docs: in addition to advisor pressure (cf. comment from SIS), there is the need to build a CV and reference list, where prestigious conferences play a big role, mostly because when you are applying for an academic job, most if not all the people who evaluate your file will not be familiar with your domain and therefore rely on acceptance rate or other perceived signs of prestige.

    Whew, that was a long sentence.

    Also there is an interesting issue with point 2: do you mean a 50% acceptance rate or that your subjective appreciation is that you have 50% chance of getting in. As noted by Steven, when you submit good work, your chances of getting accepted may be much higher than the acceptance rate, simply because a good chunk of what is submitted to many conferences is crap. (and I don’t mean this as a criticism of Steven’s achievement, which sounds impressive)

Leave a Reply to Anonymous Cancel reply

Your email address will not be published. The comment form expects plain text. If you need to format your text, you can use HTML elements such strong, blockquote, cite, code and em. For formatting code as HTML automatically, I recommend

You may subscribe to this blog by email.