Basics principles of peer-reviewing

You submit your work. A few people read it, usually the magic number of reviewers is 3, and they are supposed to be experts in the topic of your paper. These people say what they think about your paper. Your paper gets accepted or rejecting according to the reviewers’ opinions and possibly a sanity check by the general chair of the conference.

What do reviewers look for in a paper?

Usually, they focus on novelty and correctness. This means that they want to first understand why what you did matters, and how “big” your contribution is; then, they will move to checking that your work is sound (no errors or flaws). Remember that reviewers are humans, they make errors, may be swayed to praise more works closer to their line of work, have a personal taste. Also, it is hard to define what novelty actually means, and too often, lately, novelty is confused with hardness. There’s a nice post by Michael J. Black that discusses novelty in computer science, I think many of his claims apply to the cryptographic community as well.

How do I write a good review?

Be objective, be fair, try to be constructive: if you find a flaw or something that is not ‘up to the bar’, think what you would do to improve it, and provide suggestions to the authors. This constructive review technique is fairly common in mathematics, where it is not rare that reviewers become authors of re-submissions of the work. Be honest when you self-judge your ‘reviewer’s expertise’, this is meant to be in the specific topic of the paper you are reviewing. This is a very good resource for structuring reviews and using coherent terminology, and how not to write a review.

Other related topics

You can have a look here for acceptance rates in IACR Conferences. The modus operandi within cryptography is that the reputation of a publication = 1/probability to get your paper accepted in that venue. I personally do not believe this is the best judging tool, and especially, it should not be the only criteria to judge research quality (reed more about it on this other post).