Sleep vs. Wererats, Dire Rats, and Rat Swarm. Which is affected first?

[Jay R]

The party in my 3.5e game will soon be attacked by a couple of wererats, some dire rats, and two rat swarms. If the wizard casts sleep, which goes to sleep first?

"Creatures with the fewest HD are affected first." The wererats have 3 hit dice, the dire rats have 1 hit die, and the swarm has 4 hit dice. So by the rules, 4 1HD dire rats fall asleep before a 4HD rat swarm. If there are no dire rats in the AoE, then two 2HD wererats fall asleep, and the 4HD swarm is still awake.

But the swarm isn't a 4HD creature; it's a bunch of ¼HD creatures that are more easily simulated as a single entity. So it makes the most sense to me that they would fall asleep first. And since there are 4HD of them, that uses up the sleep spell, and nothing else would be affected.

How would you rule on this?

[loky1109]

"A swarm is immune to any spell or effect that targets a specific number of creatures (including single-target spells such as disintegrate), with the exception of mind-affecting effects (charms, compulsions, phantasms, patterns, and morale effects) if the swarm has an Intelligence score and a hive mind."
Sleep: One or more living creatures within a 10-ft.-radius burst

Sleep doesn't work on swarms at all.

[Jay R]

"A swarm is immune to any spell or effect that targets a specific number of creatures ..."

Sleep doesn't work on swarms at all.

Sleep doesn't target a specific number of creatures; it targets a varying number of creatures based on hit dice. That could be 4 orcs, 2 horses, or 1 blink dog. I assume that this rule includes spells like hold monster that target one creature, or mass suggestion that target one creature / level, but I'm not convinced it applies to sleep.

Besides, a swarm is a bunch of individual rats, and individual rats are susceptible to sleep. The idea that they become immune to sleep because there are a lot of them makes no sense to me.

But I'm also trying to learn. Everybody, add this question: Does sleep affect swarms?

[Tzardok]

If you were to encounter seventeen rats, sleep would put sixteen to sleep, but not the seventeenth. But suddenly, when there are 300 rats at once, it can affect all of them? Seems illogical.

Edit: So yeah, sleep counts as one of those "only a certain number of targets" spells.

[loky1109]

"One or more" sounds specific enough for me.
And no, swarm isn't a bunch of individual rats, swarm is swarm.
DM, if he don't love his sanity, can set on the table rat swarm and 300 rats side by side. And them will be completely different creatures, working under different rulesets.

About "why rats in the swarm are immune to sleep".
They aren't. Swarm is. 16 rats fall asleep, for swarm changes nothing. As like as 5 rats are killed by Magic Missile.

[Wintermoot]

If you have the need for verisimilitude around this, just tell yourself "it doesn't work on the rat swarm because they are swarming over and around each other, so they just jostle each other awake."

But yeah, doesn't work on the swarm. Hits the 1 HD dire rats first, 3 HD Wererats and ignores the swarm due to type invulnerabilities.

[Inevitability]

The argument that "One or more" is a specific number makes zero sense to me. If a term can evaluate to any possible natural number, then it is not 'specific'!

As for the difference between swarms and individual creatures... the same argument you'd use to explain any such disconnect. Why doesn't my burning hands kill the swarm when it can kill any individual rat? Some shield others with their bodies. Why doesn't the swarm die after this fall when any individual rat would? Their corpses cushion others. Just say something like 'the spell semi-intelligently knows where to direct its magic energy, and swarming rats present a different set of signatures than normal ones'.

[InvisibleBison]

It makes sense to me that sleep would have no effect on a rat swarm. A swarm is, after all, a rules abstraction designed to make it easier to run an encounter with a large number of individually weak creatures. Thus, sleep would make sixteen of the rats in the swarm fall asleep, and the remaining 284 rats would continue to function as a swarm.