There’s a concept that’s been around since 3.5, and maybe before then – maybe you’ve heard of it, and maybe you haven’t – that goes by many names. It has been called the bag of rats, the bag o’ cats, the death sack, and, in more modern times, the kitten test. In case you’re one of today’s ten thousand (hi!), the bag of kittens refers, in 3.5 and other D20 systems, to exploiting a trigger-based mechanic by having a large number of helpless animals/orphans/living things on hand. The most basic example involves a mechanic that triggers “when a creature dies” or similarly, though other uses are certainly possible; back in 3.5, liches with the Contagious Paralysis feat (Libris Mortis) sometimes kept bags of cats around so they could reach in, grab a then-paralyzed cat, and throw it as a stun grenade at people.

A lot of times the bag of kittens is used as a reason a mechanic is bad. And, to be fair, it often is. But what we’re going to be talking about today is why and when it is bad, why in some cases it is not avoidable, and the cases where it’s an option, but not a problem. So, let’s open up with this from the player end, shall we?

Acquiring Your Cats

Did you know that there are no actual defined stats for a kitten in 3.5 or Pathfinder? The closest is applying the Young template to the Common Cat, which does make a certain amount of sense. However, that means that there are still certain rather important things left undefined about getting your bag of cats, rats, or puppies around, like how much they cost and where in the hell you’re getting them anyway. One could always round them up off the streets, of course, though that may cause you to get strange questions from the authorities about why you’re dumping an astonishing amount of cats into your handy haversack and whether those cats have owners.


Adventurers get gouged on prices when they’re defined; it does not get better when the prices aren’t.

Storing Your Cats

So now you have a large pile of helpless, adorable animals that look to you for their every need, and you’re looking forward to killing them pitilessly to fuel your adventuring career (it is not suggested you do this near the party druid). Congratulations! However, now you need some method of keeping them alive until you actually want them to die. This is not as easy as it sounds at first glance. Using a bag of holding or handy haversack runs into the problem of the cats running out of air unless removed regularly, and more mundane storage runs into the problem of space and/or escaping cats. Food is relatively trivial to solve (you’re an adventurer, you’re rich), as is water, but shielding them from death-by-environmental-hazard is a bit more tricky. What happens when one needs to swim across a lake or gets thrown into a pit filled with sizzling acid?

Once these problems are resolved, we reach the next step of this process:

Using Your Sack of Cats

So, you’ve acquired your cats, you’ve found a method of storing them, and you’ve prevented them from being melted, destroyed, raised as undead, or otherwise brutally annihilated by the various frequent hazards of your adventuring lifestyle. Now you actually get to kill the things! But how? That question depends on two other major factors:  how many cats do you need dead at one time, and how fast can you actually kill them?

To demonstrate, we’ll use three different scenarios.

In our first scenario, we have a necromancer who has researched a new spell. This spell provides him with a +1 boost to his caster level whenever a creature dies within 60 ft. of him. This guy? He’s ready to roll with dead cats. Opening up his storage space and dropping a fireball into it gives him instant power for as long as his boosting spell lasts, and with methods of extending it he can afford to keep around multiple discrete bags of cats so that he can retain his god-like infusion of kitten souls. Clearly, we have a problem here.

For our second scenario, I would like to introduce Countess Felis von Orphanpuncher:

Countess Felis von Orphanpuncher

The sad part is that this outfit is not the craziest idea adventurers come up with.

Von Orphanpuncher here is a harbinger (Path of War: Expanded); harbingers recover 1 maneuver when they Claim a creature as a swift action, and recover maneuvers equal to their Intelligence modifier when a Claimed creature is reduced to 0 or fewer hit points. Sounds ideal for dead cats, yes? Well, sure, they’re an option, but they’re actually kind of a crappy option. If Von Orphanpuncher keeps the cats in a bag, she has to spend a move action to produce a cat (dropping it is a free action), her swift to Claim it, and then a standard to kill the thing – more than a full-round action. Congrats, Von Orphanpuncher, you’ve taken the most efficient recovery method in the product line and made it the worst one! Now, as you can see from the picture, she’s got another idea in mind – strap the cats to her body so she just has to spend the swift and the standard. That’s pretty efficient (assuming hazards aren’t a problem), but it means that she still can’t use her boosts, counters, or swift-action items, to say nothing of having wasted her one-and-only attack option on killing the cat. By recovering maneuvers this way, she’s spent at least 1 round not doing her job (delivering damage and debuffs) at all. The bag of cats isn’t a great option for her.

Our last example instead involves a soul hunter (Path of War), chosen because their method of recovery is similar to, but distinct from, the harbinger illustrated above. So, how do soul hunters recover their maneuvers? They spend a swift action to Claim a creature they’ve dealt damage to, gaining 1 maneuver when the creature is Claimed and maneuvers equal to their Wisdom modifier when it dies. In the specific case of the kittens, that means they can swing once for a swift + standard and get (Wis mod + 1) maneuvers back. Easy, breezy, beautiful. Right? The trouble is, again, that our soul hunter is not doing their job. They gain back a lasting resource (maneuvers), but they could have gained that resource back while still delivering damage and battlefield control instead of only being able to move when otherwise brutally murdering their cat, to say nothing of the unavoidable 5% failure chance on swinging at the cat (those natural 1s are a thing). Bag of cats? Still not a great option, especially not when they can still recover maneuvers while bringing the fight to the enemy.

So What’s This Mean?

The bag-of-cats trick is only really viable if one or more of three things is true: you benefit from a large number of dead cats, you can kill lots of cats all at once, or the bonuses granted from the dead-cat trigger are large. In our first example, the necromancer has all three things true: a bonus to CL is major, he has an uncapped benefit from dead cats, and he can kill lots and lots of cats at the same time. Spending a round butchering a hundred and fifty cats is a round well spent for him, even in combat. Countess Felis von Orphanpuncher and the soul hunter, on the other hand, need 1 cat dead at a time, under precise conditions, and get bonuses for doing so that they’d get for fighting their enemies anyway. It’s just not worth the investment. As in most things, a player looking into the bag of cats as an option (druids, please, do not apply) needs to ask herself how efficient an option it’s going to be and decide accordingly.

From a Design Perspective

The bag of cats has two problems that a designer does not like: it’s absurd, and it’s a relatively easy exploit that can break an otherwise-sane mechanic. Addressing it doesn’t have to be difficult, but the method you choose influences how the mechanic you’re putting controls on will function. The first step, of course, is:

Identifying the Problem

The bag of cats is not always an issue. Notice how Von Orphanpuncher comes out behind attempting to use the bag of cats trick? If you can math out the action cost and the bag of cats is less efficient than using the ability as intended, it’s not a problem that needs to be solved. If it breaks even with going legit, then it’s still not a problem that needs to be solved. It needs additional consideration only when it starts being more efficient than using the ability as it is intended to be used.

Fixing the Problem

So: you’ve identified that the bag of cats can work, that it’s efficient, and now you want to solve it. If you need the mechanic to remain a trigger-based mechanic for whatever reason (and there’s lots of good reasons), then installing restrictions is probably the cleanest method of solving it. If the ability comes online at fifth level or later, then installing something like this in the death trigger – “whenever a creature (with CR at least equal to ½ your character level) dies[…]” – solves your various problems in one go.

But despite this, part of the reason I suggest ignoring the bag of cats for low-level abilities, especially first level abilities, is that there’s no real way to exclude kittens that won’t also exclude, say, orcs and goblins. Excluding CR as a method of control falls apart in the first three levels when you can reasonably expect to face threatening creatures in swarms, with any individual member’s CR being a fraction of the players’ or even a fraction of 1. It’s for this reason that harbinger doesn’t have a control on what it can Claim, and why soul hunter is shortly going to be losing theirs in errata; at first level, a goblin with a CR of less than 1 is still a problem and you need the tools you use to solve that problem.

You may want to additionally, or even instead, place a cap on the bonus granted. Using a cap instead of a condition means the kittens work, but they only work so well, leaving the exploit installed for campaigns where it’s appropriate or for players cunning enough to utilize it efficiently. The stronger the bonus gained from the trigger is, the more controls you want on it. In our necromancer example, such a spell would want to exclude lower-level creatures (the CR control) and also have a tight cap, because caster level increases are scary, horrifying things that do bad things to good people. If the spell instead granted temporary hit points that stack with themselves, one might set the cap a bit more generously – half of their normal max hit point total, for instance – and maybe even permit the kittens to work, because temporary hit points, while nice, just aren’t that big a deal.

If you want the cats to not work for thematic reasons or because you think it’s silly or whatever, CR exclusion on higher-level abilities is the way to go. But for lower-level ones, there’s not a way to exclude the cats while including all the other creatures that need to be targetable unless you want to be as ham-fisted as possible and specifically call out kittens, rats, puppies, freshly-hatched chicks, and any other of the giant list of small helpless creatures casually available in any humanoid settlement. One could try excluding the Young template, of course, but then you have fun times where a GM uses that template to represent an actual threat and your ability stops working for no reason. No one particularly wants that, yes?

In Conclusion

The bag of cats exists, and it’s never going to go away, partially because it’s hilarious, but mostly because it’s famous. The argument is often made that no reasonable GM would allow it, but our job as designers is to make the mechanics function as we want to anyway, so as to lift some of the massive burden of work already on the GM. With that in mind, I’m not convinced it’s as big a problem as it’s made out to be, and it’s certainly not always a problem; that is, the availability of dead cats for bonuses does not automatically mean the mechanic is bad.

But it should be kept in mind. Even if the actual bag of cats is banned, death-trigger mechanics do get better when fighting swarms of weaker enemies, so even if everyone is playing it straight and no kittens are involved you might have a problem worth examining. But if you’ve examined the consequences and costs and it’s just not a big deal? Leave it alone. Save yourself some work and some headaches.

Jade Ripley

Freelancer, Dreamscarred Press

Artwork provided by Domochevsky of

No kittens were harmed in the making of this article.

Liked it? Take a second to support Andreas Rönnqvist on Patreon!