Mind over Meta #56 - Conditioning

Welcome back to another installment of Mind over Meta! This week, we’re going to be dipping into some basic Psychology by talking about Conditioning! Sit back and enjoy the ride!

 

momline1.png

 

ARCHIVE

 

momline2.png

 

SKINNER BOXES

 

Conditioning, in is most basic sense, is learning to do something based upon certain stimuli in one’s environment. In Smash, and most cases with a competitive game, almost any conditioning that takes place is known as operant conditioning. Operant conditioning is defined as “a learning principle in which environmental contingencies—or more specifically, discriminative stimuli (antecedents) influencing its consequences—are controlled and manipulated to change behavior”. In a more intuitive sense, operant conditioning involves conditioning someone to perform an act on their own, receiving a reward or punishment for doing said act. When you’re a kid and your parents reward you for doing chores, or when they punish you for, that’s operant conditioning.

 

The concept of operant conditioning was created by a psychologist named B.F Skinner, known for his famous Skinner Box experiment (also known as an operant conditioning chamber). In this experiment, an animal is put in a box or cage which has a switch or lever. Hitting this switch will release food, water, or some other positive reinforcement. In this experiment, Skinner had found that animals would often hit this switch to receive the positive reinforcement, and if they had stopped receiving the positive reinforcement, they would stop hitting the switch. This showed that the animals had learned that they could gain the positive reinforcement by performing an act, and learned to stop performing the act when the positive reinforcement had stopped.

 

Now, when it comes to the reinforcement (increasing a desired activity/response) and punishment (decreasing a desired activity/response), there’s 4 options: Positive reinforcement, which is the addition of something wanted (in the case of the Skinner Box, the food was positive reinforcement), negative reinforcement, which is the removal of something unwanted, (such as doing something to avoid being punished), positive punishment, which is the addition of something unwanted (such as a parent making their child do chores as punishment), and negative punishment, which is the removal of something wanted (such as a parent taking away privileges as a punishment). In Smash, and again, most competitive games, you’ll usually be dealing with positive punishment, although all 4 of the different types of reinforcement/punishment can be present.

momline3.png

 

LATHER, RINSE, REPEAT

 

When it comes to Smash, many people consider conditioning to be a bit of a risky strategy. On one hand, you can make the opponent play into your hand, but on the other hand, it takes quite an effort to condition someone to do something. You have to make them think that what they’re doing is safe, when it’s really not. What you’re tricking them into doing can be anything from approaching a certain way in neutral, recovering at a certain height, or even just DI’ing something a certain way. In fact, most DI mixups can be manipulated with conditioning. For example, if you use one specific throw every time you grab someone, then your opponent will start DI’ing for that throw. Now, you’ve conditioned them to DI a certain way whenever you grab them by taking away the punishment that would come with bad DI (negative reinforcement), so you can mix it up by doing another throw and punish their DI. Now, that’s not to say that’s necessarily a good thing to do, because there’s still a chance they can read your mixup, and you’re throwing away a potential punish every time you use that throw knowing they’ll DI it correctly.

 

Of course, conditioning is not bad to use. There’s still plenty of great uses for it, and you’ll often condition people in small ways without realizing it. Every time you whiff a punish on someone in a specific scenario, you’re conditioning them to either think they’re safe in that scenario, or think that you don’t know that they’re not safe in that scenario. In the same way, every time you punish someone for doing something in a specific scenario, you’re conditioning them to avoid that scenario (positive punishment). For example, if I continue to punish you for a certain kind of approach in neutral, I can essentially bait you into trying a different approach next time by conditioning you to know that the first approach was not safe.

 

That brings me to my next point: Baiting.

 

momline4.png

 

BAITED AND OUTSMARTED

 

A classic example of conditioning is baiting, as, in Smash terms, they’re very similar. The process of baiting (which you can read more about in our article here: https://www.reddit.com/r/SSBPM/comments/33ykpf/mind_over_meta_18_gr8_b8_m8_presented_by/) is essentially making someone do something predictable because they think it will work. While conditioning is similar, conditioning requires one of the 4 previously mentioned processes of positive and negative reinforcement and punishment, while baiting doesn’t always need that. Something like dash dancing and reacting to something that the opponent does is baiting, not conditioning, while making someone perform a certain action because you’ve made them think that it’s safe through the aforementioned 4 processes is conditioning AND baiting. You’re essentially baiting someone into doing something by conditioning them.

 

In many sets in which a player baits the other in a specific scenario, there’s conditioning behind it. Like in this example in Brawl, ZeRo punishes the Snake’s attempt to land next to him twice. He’s used positive punishment to make the Snake think that landing next to him is not safe, so the Snake goes for a different option. ZeRo baits this option out by conditioning him to think that landing next to him is not safe, and punishes it with up b.

 

An example of baiting like this in PM would be with Toon Link. If you use a lot of bombs as Toon Link, your opponent may try and use the bombs against you by catching them and throwing them back. You can let your opponent do this until you find a pattern, and then punish. Do they catch it with a glide toss upward? Punish their upward shift in momentum by catching them with a boomerang. Do they z catch it and throw it back right away? Catch it and throw it back again. The same goes for Links bombs, and even ROB’s gyro and Diddy’s banana.

 

Another quick example could be found in Dedede’s neutral B. I find that whenever I play against someone, the first time I spit them out off stage, they almost always jump out of the star. Dedede can easily punish this with forward air , so once I do this once or twice, most people will opt to go low. When this happens, I can just as easily jump off stage and fast fall forward air to punish their low recovery instead. In this scenario, I’ve conditioned them to think that recovering high is unsafe, and then baited them into recovering low, effectively putting them in an even worse position.

 

Taking another example from Dedede, his up throw can often be followed up with up tilt at lower percents, especially on fast fallers. At a certain percent, many characters will be able to jump out of this follow up. If Dedede doesn’t punish this at lower and mid percents, he’s using negative reinforcement by removing the punish (up tilt) and making them think that jumping out of up throw is safe. Suddenly, at higher percents, Dedede can chase you by jumping up and using up air, which can kill on some characters. This type of mixup exploits good conditioning.

 

Every character can do something like this. While it may not be as clear or as easy, every single character in Project M has the tools they need to bait and condition people in certain scenarios, or even in neutral. You’ll often see some characters camp in neutral, using positive punishment to condition the other player to think that not approaching is unsafe, effectively forcing them to approach.

 

momline2.png

 

RECAP

 

So, what have we learned today?

 

  • Conditioning in Smash and most games is considered to be Operant Conditioning

  • There are generally 4 types of reinforcement and punishment

  • Baiting doesn’t require conditioning, but conditioning usually ends with a bait.

  • Baiting and conditioning is common in many scenarios in Project M, and all fighting games for that matter