We discovered that slashd (think slashdot's creation of cron way back in slashdot's early days) did not run to completion overnight when it was updating mod points.
NB: Rehash (the code that runs the site) is a fork of a version of slashcode. That code was open-sourced a couple years before we grabbed it. It also needed a great deal of work to get into an actually-functioning state.
It actually handed out some mod points... that delayed our understanding of what had happened. I restarted slashd several hours ago. Problem solved, right? Nope.
Apparently things were left in an awkward state where some users got some mod points and others did not get any. Of course, I did not realize *that* until a bunch of people had used mod points.
So there was no obvious way to give "make up" mod points to those who missed out. (my mysql skills are limited). Nor, apparently, to any other staff who was around.
Rather than mung (Mung Until No Good) things even further, in discussion with other staff, it was decided to leave well enough alone. Things should clear themselves up on their own the next time mod points are scheduled to be re-issued (starting at 00:10 UTC).