July 13, 2026
Peer review or paper purgatory?
Counting ArXiv Delays
Researchers blame the queue, but the comments say the real chaos is moderation purgatory
TLDR: One frustrated researcher checked years of arXiv data after their paper was delayed and found no big overall slowdown, suggesting the problem may be category-specific moderation instead. In the comments, people split between "this is ridiculous" and "good, someone has to stop AI slop."
A scientist got stuck in arXiv’s dreaded “on hold” pile and responded in the most internet-brained way possible: by scraping years of data to see whether the site had quietly turned into a waiting-room nightmare. arXiv, for the uninitiated, is the giant online bulletin board where researchers post papers before or alongside journal publication. After digging through archived pages and matching submission dates to public listings, the verdict was unexpectedly spicy: there’s no giant sitewide slowdown. The overall delays look weirdly stable over time, which means the author’s personal hold-up probably wasn’t caused by some recent system-wide meltdown after all.
And that’s where the comments turned into a group therapy session with side-eye. One user begged for an even more dramatic chart showing how many papers were delayed by at least a certain number of days, then shared their own tale of category-flipping “purgatory” after arXiv changed their paper’s label without asking. Another commenter came in hot with the harshest take of the thread: with AI-made junk papers flooding the internet, why is anyone surprised there are barriers now? Translation: better a slow gate than a slop avalanche. But others were not having it, saying the delays feel absurd even for papers already accepted by top journals. The funniest mini-plot twist? In the middle of all the queue rage, one commenter just gushed that the charts were gorgeous. Even in drama, the data nerds stay on brand.
Key Points
- •The article analyzes arXiv submission-to-announcement delays after the author's own paper was placed on hold.
- •To build the dataset, the author combined Internet Archive snapshots of arXiv announcement pages with submission dates from arXiv paper pages and recent data from arXiv catchup pages.
- •The sample covers 63,847 papers from 2015-01-28 to 2026-04-04.
- •Delay was measured in working days between submission and announcement, excluding weekends and holidays.
- •The article finds no sustained overall increase in delays over time and instead points to variation across subject categories as the likely explanation for specific long waits.