Monster AI, Intelligence Cheat Sheet?

[Bjarkmundur]

Argh! I saw on tiktok someone using a cheat sheet that determined base monster behavior based on their int score, and now I can't find it again! It was basically a threshold for tactics.

It had maybe four columns, and showed how smart a creature has to be in order to use certain tactics, such as focusing a caster, having an escape plan ready etc.

I know we can make one easily together, but it's gonna bug me forever if I can't find the original one.

Are you familiar with the tiktoker or the sheet I'm referring to?

[Durazno]

Is this what you're looking for?

It's at least a similar concept.

[ad_hoc]

Intelligence is a measure primarily of ability to learn and remember facts.

It is not a measure of tactical acumen.

Beasts, esp. predatory ones, are very tactical.

[PhantomSoul]

If you have access to The Monsters Know What They're Doing (book), that's a great spot to get baselines (with some moderation by desired feel, which as I recall is even in the book). The website has some relevant pages, e.g., https://www.themonstersknow.com/why-these-tactics/, but there's a fair bit more in the book to use as a launching point. It's not strictly INT-based (and I haven't seen a TikTok one), but it might still have the sort of content you're looking for!

[Unoriginal]

If you have access to The Monsters Know What They're Doing (book), that's a great spot to get baselines (with some moderation by desired feel, which as I recall is even in the book). The website has some relevant pages, e.g., https://www.themonstersknow.com/why-these-tactics/, but there's a fair bit more in the book to use as a launching point. It's not strictly INT-based (and I haven't seen a TikTok one), but it might still have the sort of content you're looking for!

This particular author has peculiar views on how INT affect behavior, though. Like he's commented that an 8 INT NPC would be barely capable of putting a blade on a pole and calling it a glaive, for example.

[PhantomSoul]

This particular author has peculiar views on how INT affect behavior, though. Like he's commented that an 8 INT NPC would be barely capable of putting a blade on a pole and calling it a glaive, for example.

Yeah, moderation definitely needed... :)
(I remember the attacks vs. saves stuff more fondly, but also stuff like thinking about goals and inferring combat "styles" from stat blocks [e.g. skirmisher vs. brute vs. melee/range vs. ambush]. Details and exact implementations more subject to debate, but the "ability contour" as a baseline is still a nice concept. I did appreciate explicitly highlighting that Int ≤ 7 might mean less adaptable, but not inefficient or ineffective with their abilities, contra some common white-room talk! Underdeveloped survival instinct at Wis ≤ 7 seems a bit... much, though.)

[sithlordnergal]

I'll be honest, I don't base NPC tactical ability on their Int score, I prefer to use their Wisdom over their Int. After all, a pack of wolves knows how to pull off an ambush, knows when to retreat, knows how to flank/grapple, and knows that working together lets them take down larger kills. And yet, despite knowing all of that, a Wolf's Int is only 3. If Int scores really did determine how tactical a creature is, wolves shouldn't be able to do any of that

[Greywander]

I'll be honest, I don't base NPC tactical ability on their Int score, I prefer to use their Wisdom over their Int. After all, a pack of wolves knows how to pull off an ambush, knows when to retreat, knows how to flank/grapple, and knows that working together lets them take down larger kills. And yet, despite knowing all of that, a Wolf's Int is only 3. If Int scores really did determine how tactical a creature is, wolves shouldn't be able to do any of that

I think INT and WIS represent two different qualities of a creature's mental ability. Wolves have some decent tactical execution, but lack the intelligence of a human. Wolves know how to flank, but would not recognize a weapon. They're crafty, but easy to fool because they lack the ability to comprehend in the same way a more intelligent creature can.

With this in mind, you could expect a high WIS creature to employ some more advanced tactics, like flanking, baiting into a trap, and so on. But with a low INT, it will hamper their ability to make sounds tactical judgments. A wolf won't have a concept of a "mage", and may instead target someone like the barbarian as the most dangerous looking enemy on the field. A wolf doesn't understand concepts like spells or concentration, and thus will make no effort to attack a mage that is concentrating on a dangerous spell, instead reacting to the spell as if it were a natural and independent phenomenon.

Which isn't to say wolves couldn't be trained. You can train a wolf to recognize weapons and attack someone who brandishes one. You can also train them to target mages first, especially when a flashy spell is in effect. However, this doesn't change their intelligence; they would only do these things because they've been trained to react a certain way to certain stimuli. They would still lack the intelligence to truly understand why such tactics are effective, and you could subvert their training by using something they don't recognize (e.g. a type of weapon or spell they've never seen before, putting the mage in armor, etc.).

[ad_hoc]

I think INT and WIS represent two different qualities of a creature's mental ability. Wolves have some decent tactical execution, but lack the intelligence of a human. Wolves know how to flank, but would not recognize a weapon. They're crafty, but easy to fool because they lack the ability to comprehend in the same way a more intelligent creature can.

With this in mind, you could expect a high WIS creature to employ some more advanced tactics, like flanking, baiting into a trap, and so on. But with a low INT, it will hamper their ability to make sounds tactical judgments. A wolf won't have a concept of a "mage", and may instead target someone like the barbarian as the most dangerous looking enemy on the field. A wolf doesn't understand concepts like spells or concentration, and thus will make no effort to attack a mage that is concentrating on a dangerous spell, instead reacting to the spell as if it were a natural and independent phenomenon.

Which isn't to say wolves couldn't be trained. You can train a wolf to recognize weapons and attack someone who brandishes one. You can also train them to target mages first, especially when a flashy spell is in effect. However, this doesn't change their intelligence; they would only do these things because they've been trained to react a certain way to certain stimuli. They would still lack the intelligence to truly understand why such tactics are effective, and you could subvert their training by using something they don't recognize (e.g. a type of weapon or spell they've never seen before, putting the mage in armor, etc.).

I typically have wolves go after the weakest.

Makes trouble if the PCs are escorting.

[Angelalex242]

That only goes up to 13. As it gets ever higher, towards 20 like pretty much every PC wizard, you almost have to metagame to account for how smart 20 really is.

[Chronos]

In the most recent game I DMed:
"Wait, at Int 5, is a water elemental really smart enough to pull off a tactic like that?"
"No, but the Int 20 archmage commanding it is."

But then, of course, it went south when one of the water elementals tried to keep on following its orders, even when the tactical situation had changed to the point where that no longer made sense, and the archmage wasn't able to give it new orders.

[Greywander]

In the most recent game I DMed:
"Wait, at Int 5, is a water elemental really smart enough to pull off a tactic like that?"
"No, but the Int 20 archmage commanding it is."

But then, of course, it went south when one of the water elementals tried to keep on following its orders, even when the tactical situation had changed to the point where that no longer made sense, and the archmage wasn't able to give it new orders.

Now that's some truly advanced tactics (though likely your players did it on accident). Baiting the enemy into a certain strategy, predicting what that strategy will be, and then cutting off their communication so that the enemy continues to execute that strategy even after the situation changes (because they don't have a backup plan, and the one who could come up with a backup plan can no longer communicate with them). Of course, maybe they do have a backup plan, so you need to account for that such that their backup puts them in an even worse position.

TBH, I've seen stuff like this in anime, and it tends to strain credulity. Death Note has a lot of this, for example. It's hard not to think about how a much simpler plan would have probably been more effective, or how a single mistake could have caused the whole plan to collapse on itself (it almost works better when the heroes stumble into something by accident, acting as more of a spanner in the works of the villain's plan, rather than an elaborate counter-plan designed to outsmart the villain). But of course, something like this could actually happen, and in a D&D game it would be a bit more credible (unless the DM is deliberately tailoring things to appear that way), and it would feel amazing if you could actually pull it off.

[dana96]

Is this what you're looking for?

It's at least a similar concept.

Thank you. I was looking for that info recently too.

[da newt]

"With this in mind, you could expect a high WIS creature to employ some more advanced tactics, like flanking, baiting into a trap, and so on. But with a low INT, it will hamper their ability to make sounds tactical judgments. A wolf won't have a concept of a "mage", and may instead target someone like the barbarian as the most dangerous looking enemy on the field. A wolf doesn't understand concepts like spells or concentration, and thus will make no effort to attack a mage that is concentrating on a dangerous spell, instead reacting to the spell as if it were a natural and independent phenomenon."

I like the idea of trying to figure out how a creature would think/act based on it's INT and WIS, but I think we often fall into the trap of basing this on what a group of people would do in order to win a battle (or even worse - a game), and forget that a pack of wolves have no concept of trying to win X - they are predators, their goal is to kill something so they can eat it while expending the least amount of energy and exposing themselves to the lowest risk possible.

Predators will isolate and gang up on the one weakest prey item in order to kill it quickly and take it away to be eaten. A pack of wolves would never attack the "most dangerous looking enemy" or even attack all the members of a party - to do that would be counter to their survival.

[Chronos]

Unfortunately for the party, from a wolf's point of view, a wizard looks like "the weakest member of the party".

And my archmage had ordered his two water elemental minions to move on top of two of the party's casters, and to whelm them if they tried to cast a spell. It worked to shut down the cleric, for several rounds, but the one assigned to the sorcerer just couldn't quite reach him right away, and then he used a Misty Step to get far enough away that it never managed to close the distance. So it just kept on chasing the sorcerer it couldn't reach, even when other party members were within range.

[MoiMagnus]

Unfortunately for the party, from a wolf's point of view, a wizard looks like "the weakest member of the party".

Quite interestingly, this is something which depends on worldbuilding.

In my setting I usually consider that magic can somewhat be "felt" by peoples and creatures (in cases where it isn't used to dissimulate itself). In such a world, non-magical predators would likely learn by natural selection to fear the "feeling" of magic, hence avoid spellcasters instinctively.

[Joe the Rat]

I wouldn't use INT for having tactics, but rather for the range of options, and flexibility in implementation. Do you have different approaches best suited to the known situation? Can you adapt to unexpected complications? That sort of thing.

This particular author has peculiar views on how INT affect behavior, though. Like he's commented that an 8 INT NPC would be barely capable of putting a blade on a pole and calling it a glaive, for example.

Indeed. And it's oddly just the mental that are down-counted. The physical analysis and incorporation of abilities and motivations makes for a good read. (Animals are hungry)