Heist Logic: Why the Guards Stand Down in a TTRPG

False Machine had a nice post about the Thief video game series, its emphasis on stealth, and what aspects of the game can (and cannot) be applied to TTRPGs.

A useful comment by Kyana focuses on a particular issue for TTRPGs; how and why does security “reset” after an incursion? 

Thief (1981) Michael Mann

I haven’t played the Thief games, but I’ve certainly seen the same problem in other video games, like Metal Gear Solid. You’re spotted in a secure area by multiple antagonists, the whole place goes on high alert, you’re caught in a firefight, grenades are exploding, soldiers are dying. But after 90 seconds spent hiding in a locker, the game reverts to its baseline state, and the guards resume normal patrols, as if nothing happened.

Video game players accept this artifice because it facilitates good gameplay. It would certainly be realistic if you had to wait hours (or days, or indefinitely) for the guards to stand down from high alert, but it would also certainly be boring. So the game speeds past the boring part and gets back to the fun part.

TTRPGs also endeavor to speed past the boring part and get to the good stuff, but they also have a generally higher expectation of verisimilitude than video games do. This is, in part, because players in a GM-led game expect the GM to dynamically react to PC action in a way that is impossible (thus far, at least) for a video game to match. 

So say you’re running a heist or infiltration in D&D or a similar game – basically, any game that isn’t like Blades in the Dark, which has built-in procedure for heist complications and failure states. You need a plan for how and why the guards are going to stand down or otherwise redirect attention in a reasonable amount of time, because you don’t want a single guard sighting to botch the entire enterprise. What kind of circumstances and conditions can we add to prolong the heist without resorting to video game-level suspension of disbelief?

1. Misdirected response. The guards react, but deploy their resources in the wrong place. Perhaps they cut off exits while the players are still delving deeper into the facility, or move to protect the big boss when the players are really after the MacGuffin. 
2. Expecting someone else. A twist on the above. The defenders suspect a completely different adversary is behind the incursion. Either directly (if they didn’t get a good look at the PCs) or indirectly (they presume the PC party is a decoy, or merely an illusion, or similar). It might be a faction or NPC that the PCs know, or even a group they weren’t already aware of. Either way, the defenders’ reaction to the perceived “true” threat wastes time or resources that give the PCs a renewed opportunity.
3. Multiple infiltrators. Twisting the above in a different direction, there actually are other infiltrators, unconnected to the party, whether here for the same prize, or something unrelated. The defenders catch one of the other individuals or groups infiltrating, taking some heat off the PCs. 
4. They don’t appreciate what they have. Most of these options presume that the defenders have a good idea of what they’re protecting. But if they don’t – if they’re unaware of the value of what they have, or don’t even realize it is within their area of control – the nature of the heist changes, and it’s more plausible that their reaction would be delayed, misdirected, or ineffectual.
5. A crisis is also an opportunity. There’s some internal conflict within the defenders’ ranks. Maybe a second-in-command wants a shot at leadership, or a sub-faction wants to leverage the situation against a rival sub-faction. Resources are spent primarily to advance this goal, rather than respond in full to the threat presented by the PCs.
6. Environmental distraction or complication. “Environmental” in this instance is just shorthand for something happening independent of any faction action. Something in the scenario outside the defenders’ control hinders their response. It could be bad weather, wild animals, invasive plants, stellar emissions, localized tremors, or even some supernatural effect of the MacGuffin itself. 
7. The heist is counterintuitively helpful in some way to the defenders of the location, and tacitly allowed to proceed. The aforementioned Metal Gear Solid game does this. The antagonist actually needs the protagonist to succeed, at least in part, to advance their overall plan, so some suboptimal efforts by the guards can be interpreted as an intentional ploy. But be careful with this one, as it can stray into gotcha-style GMing. The antagonists’ goal should probably be orthogonal to whatever the PCs are trying to do, rather than a direct negation of their success, so that circumstances out of their control or awareness can’t rob them of a win if they complete the heist.
8. Dumb but dangerous. This is essentially what Patrick recommended in the False Machine post with his ogre guards. Aggressive, loyal servants with goldfish memories are good antagonists during a heist. 
9. Programmed guards. Some or all of the guards act programmatically. They are undead, golems, trained animals, robots, mind-controlled servants, or similar. They can patrol, pursue, and attack, but they don’t have the capacity to react in complex, adaptive ways to PC action. The PCs can take advantage of this to continue the heist even if they’ve been spotted once.
10. A wizard did it. Similar to the above, but more open-ended. Perhaps the wizard or other magical antagonist is so paranoid that they dose their guards with amnesia-inducing chemicals. Perhaps the guards are all charmed, and some of the enchantments breaking (due to accident or PC intervention) disrupt an organized response to an alarm.
11. Magical passage of time. Remember how we said it would be realistic but boring to wait a really long time for the guards to stand down from a high alert? In a fantasy world with magic, that’s not a hard limitation on PC solutions. Some kind of magical item that allows them to do a duration-extended Rope Trick or similar effect could go a long way here. In this instance, the heist goes on, but every failure requires waiting some long period of (in-game) time, with possible complications in the outside world. This trick is the twist to a certain heist movie, where the protagonist waits out the guards for an implausibly long time. Hiding for spoiler purposes, but I'm talking about Inside Man (2006).
12. Life is complicated and mistakes happen. The simplest explanation of all. If you read accounts of real-world revolutions, battles, and other pivotal historical moments, it’s amazing how often the fate of nations and peoples hinges on situations where people simply make a lot of mistakes, fail to communicate, or organize suboptimally, and those errors domino out of control. Defending a secure location is quite complex, and some security failures are going to boil down to this kind of thing.