Hacker News new | past | comments | ask | show | jobs | submit login
Non-Hermitian topology in rock-paper-scissors games (nature.com)
35 points by RafelMri on March 6, 2022 | hide | past | favorite | 8 comments



That abtract has big hermeneutics of quantum gravity[1] energy for me. Probably I'm just ignorant, but is it actually necessary to the march of science for academics to write this way?

[1] https://physics.nyu.edu/sokal/transgress_v2/transgress_v2_si...


It's unfortunately an artifact of the peer review process. Many venues only accept 15-20% of submissions, and so otherwise good papers get rejected for not following the formulaic structure.

The issue is really that reviewers (who are volunteers) don't have a lot of time to read the manuscripts. Many of them are overworked grad students. Thus, as silly as it seems, outside of this style it's easy to be rejected because the reviewer didn't correctly identify the contributions when they skimmed the paper.

It is worth noting though as a difference between the linked works that games are readily accepted to be tied to more broad dynamical systems mathematically, while (having no knowledge of the area) I'd speculate the quantum gravity social science is not as established.


The intention of this journal (Nature Scientific Reports) was to peer review only for technical correctness and not judge "interesting-ness". I don't think it's really panned out that way but nevertheless this journal has a pretty high acceptance rate. I'm sure there are still systemic pressures to follow a certain formula, but I don't think one needs to do so to publish in Scientific Reports.


It's not so much the "interesting-ness" aspect. It's about recognizing where the papers fit and if there is a novel contribution in the first place.

Even though they emphasize technical correctness, it's very easy for a reviewer to misread the intention of the paper and assert that it, for instance, uses too simple of a model for a different field, or didn't cite appropriate literature for that field. It obviously doesn't happen all the time, but the fact that there is that risk there encourages the general formulaic writing (at least for me).


> reviewer didn't correctly identify the contributions

yeah, OR a creeping culture of slavish mimicry prevails over valuation of unique insights, one paper at a time. no pure answers in psychology


I mean, maybe there's an aspect of pointless mimicry (though I haven't seen evidence of it), but my comment represented my general experience as someone who works the dynamical systems area and has reviewed for that journal in the past.

I have yet to receive a negative review based on using a less standard technique/perspective beyond the occasional "this work doesn't belong to this field, but instead to [insert adjacent field here]" which seems to usually be ignored by the editor.

I have, multiple times, had one of the three reviewers write a negative review based on a (to me at the time) seemingly conceivable misunderstanding of contribution of the paper because I had done something like include a simple example to better illustrate an idea. The editor sometimes rejects submissions based on one negative review, even if the other reviews are positive.

I think it's relatively easy to speculate that the consistency in form is from the feedback of the system:

Assume a researcher wants to maximize his chances of getting the paper accepted, and that, if accurately reviewed, the content of the paper would merit acceptance. Furthermore, assume that there are a finite number of distinct styles of writing in scientific papers, and that the likelihood of accurately reviewing a paper increases monotonically with the number papers you've seen in the given style (the source of which are papers that have successfully made it through the process).

Then the researcher would aim to match the most prevalent writing style. Even if they aren't always successful at it, it skews the input distribution towards the most prevalent writing style. This, in turn makes it more likely for that style to be accepted in the future.


>the researcher would aim to match the most prevalent writing style

My easy method to determine if a paper is legit:

   1. Open PDF in Acrobat.
   2. Select "properties".
   3. Choose Font tab.
   4. Does it use Computer Modern fonts?
   5. Yes --> it's real; No --> it's a crank.


I have a doctorate in mathematics, I didn't get those vibes when I read this title




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: