Looking through the logs of the review queues, I sometimes see very long chains of reviews from the same user, frequently all with the same decision (leave open, no action needed, looks OK, etc.); occasionally, these take place over very short periods of time.

For example, in the First Post review, I have recently observed a user make $17$ reviews over the course of $1$ minute and $9$ seconds - this is an average of about $4.7$ seconds per review. Looking through the history, I can see other periods of rapid reviewing from the same user ($5$ reviews in $24$ seconds, $3$ reviews in $11$ seconds, $3$ reviews in $5$ seconds, and so on). The vast majority of these reviews select "No Action Needed." These chains of reviews include at least one passed and at least one failed audit, but also contain "No Action Needed" reviews for posts that were later closed / deleted via other queues.

In my experience, $5$ seconds is too short a time to make a reasoned judgement about whether a first post is good or not, and since a single "No Action Needed" removes a post from the queue, inattentive first post reviewing seems to be more potentially harmful than other reviews.

I have seen similar behaviour from other users in other queues, too - including vastly more egregious instances of robo-reviewing.

Now for the questions:

How fast is too fast when reviewing (am I just a particularly slow reviewer, or is $5$ seconds sufficient)? If one observes scattered instances of such reviews, what is the appropriate action (e.g. flagging the moderators with a report)?

  • 38
    $\begingroup$ My usual course of action is a sigh of disappointment mixed with futility. Sometimes I roll my eyes as well. $\endgroup$
    – Asaf Karagila Mod
    Commented Jun 18, 2014 at 8:17
  • 19
    $\begingroup$ Facepalming is also appropriate, although moderation is called for to avoid facial damage. $\endgroup$
    – rschwieb
    Commented Jun 18, 2014 at 14:41
  • 11
    $\begingroup$ @rschwieb Care also has to be taken to remember to palm forehead rather than glasses. $\endgroup$
    – Justin
    Commented Jun 18, 2014 at 22:49

1 Answer 1


Flag, flag, flag, flag, and flag.

If you see something wrong, tell the mods.

If you see your flag declined, you will know that you may be a bit too sensitive. If you see your flag marked as helpful, you will know that you've done a good thing.

For what it is worth moderators can temporarily ban users from review queues. We have used it a few times here and there when we caught someone "breezing" through queues like you described. Though we prefer to act when the evidence is a bit more concrete: for example, selecting "No Action" on a post that is clearly spam, or failing outrageous instances of the review audit.

As to how fast is too fast:

it really depends on the question. An obvious spam (as in actual advertisement) can be identified in less than 3 seconds. A longer tirade against the Catholic church (just as a random example of something definitely off-topic) may take a while to read and make sure.

While I don't review on MSE (because of mod superpowers it makes my votes count a bit more), on MathOverflow I think I often take 30+ seconds to review each instance. And quite frequently after thinking about it I will decide I don't have enough of an opinion to say anything other than "skip". So no, you are not particularly slow...

  • $\begingroup$ Thank you, I've flagged one of the reviews I was talking about. I too will frequently have to skip after reading the question. $\endgroup$
    – user61527
    Commented Jun 18, 2014 at 15:35
  • 12
    $\begingroup$ I don't know, I could probably discriminate between a long tirade against the Catholic Church and a legitimate math post in about 3 seconds or less, also. $\endgroup$
    – Emily
    Commented Jun 18, 2014 at 17:55
  • 5
    $\begingroup$ @Arkamis The Catholic Church should have had you examining Galileo's work ;) $\endgroup$ Commented Jun 24, 2014 at 11:16
  • 2
    $\begingroup$ @TrevorAlexander I might have better served Galileo ;) $\endgroup$
    – Emily
    Commented Jun 24, 2014 at 13:38

You must log in to answer this question.