Behaviorism is a sniper strategy that relies less on watching the hard tells spies generally need to do (bugging, swapping, purloining), and more on what an AI looks like and will likely do. They try to figure who the spy is by asking themselves, "Why would a spy do that? Would a spy even do that?" and using those deductions to reason out who is probably an AI and which of their suspects is the spy.
This style of sniper play is what most players attempt when they are new to SpyParty, but usually they grow out of it as the beginner vs. beginner maps reward camping much more than behaviorism. This style also requires a lot of knowledge about the game to be played proficiently. This is not to say camping doesn't require one to know a lot about game mechanics, but behaviorism can require knowledge about much more miscellaneous parts of the game that many player might not be aware of.
The Second Book of SpyParty Strategy[]
The Second Book of SpyParty on how to approach a behavioralist sniper:
Behavioralists, as the name suggests, are more concerned with the general behavior of the partygoers than they are with specific tells. Behavioralism tends to be used only by advanced players since it requires considerable experience and knowledge to use effectively. They are relatively more likely to shoot you for walking around from mission site to mission site than they are because they saw you actually complete a mission.[1] The goal of the behavioralist is to take maximum advantage of all the data presented by the party by building increasingly accurate dossiers on the likely guilt of this partygoer or that. They do this by asking at the outset "what kinds of things will the spy do, and what kinds of things will the AI do?" Then they analyze the party looking for those things. They do look for actual tells, but they sacrifice a great deal of coverage in order to do their behavioralist data mining. This means that they often don't need to see actual tells - they know what a spy completing their missions "looks like" without ever seeing the actual tell. Likewise, they know what an AI "looks like" - they are often able to get many accurate lowlights, which lets them narrow their focus. To be sure, different behavioralists pay attention to different things- some concentrate more on finding AI-like behaviors, some concentrate more on looking for spy-like pathing, and so forth, but all of them share the same general focus and relative indifference to actual mission tells.
Against a behavioralist, your most potent weapons are two P's - Pacing and Pretending to be an AI. Of these, pretending to be an AI is the most important. The greatest weakness of the behavioralist game style, besides the light guard placed upon actual missions, is the risk of a type II error, the false negative. You must be able to put on a convincing show of being an AI to take advantage of this, and you must do it against behavioralists, who are expert in this subject, since they spend all their time staring at AIs and trying to figure out what they look like. In order to beat such a sniper, you must be very well-practiced in pathing, idling, and all the little do's-and-don't's.... Further, you must put these ideas effectively into practice while still accomplishing missions, which means breaking your cover, at least temporarily. Choosing your moments to actually do missions and other actions that require you to break your "I'm an innocent AI" act is critical and difficult.
The difficulty of accomplishing even soft tell spy actions against a behavioralist[2] is one of the reasons pacing is so important against them. Good pacing will minimize the difficulty to you of the moments when you have to "break cover" by concentrating those moments in a short period of time, though of course you risk the sniper noticing your clump of mission-like behaviors. Another reason pacing is important here is because of the way a behavioralist tries to make progress in the game. Recall from the previous section that rushing missions either at the beginning or end of a game is an attempt to destroy or subvert, respectively, the sniper's advantage of improving the accuracy of their data about the partygoers as time goes on. Because a behavioralist relies more than usual on this advantage, rushing can be very effective against them. Rushing missions early denies them any time at all to collect partygoer data and improve their game. Rushing missions late gives you a great chance to be lowlit/otherwise considered to be an AI before you try to complete your missions, and it creates a good chance for a bad shot. Rushing missions both early and late takes advantage of the primary problem with behavioralism - that it leaves the actual missions relatively unguarded. Just keep in mind two things: 1 - truly behavioralist sniping is very difficult and only very advanced players ever manage it successfully. Such players tend to adapt very well to varied spy tactics, and 2- Like anything else, rushing will get you killed if you do it too often.
- ↑ They still see hard tells, of course, they are just relatively less likely to notice them because of their broader focus.
- ↑ Because they do not look for actual tells, but instead "mission-like" behaviors, soft tells are no harder than hard tells to guard for behaviorists; sometimes they are easier.