Excellent, excellent article by one of the new staff writers on SRK, MyLifeIsAnRPG.

In the article, he discusses very interesting concepts to learning such as chunking and scaffolding and how you can apply that to your self-development in fighting games.

Here’s a healthy excerpt below, and I highly recommend to everyone to go to SRK, check the full article out, and perhaps leave a comment or thought.

“What Is Chunking?

Believe it or not, the human brain actually has a RAM limit. It can think about approximately seven things at once (give or take depending on the person.) However, what a “thing” is, is pretty amorphous, so the human brain saves mental RAM by clustering several objects together as one “thing.” This is called Chunking, and here is an example. Highlight and read the string of random numbers below. Then un-highlight and try to recall it.

13568453674

How did you do? Was it difficult?

Now do the same with this phone number.

1-(356)-845-3674

Easier? Here’s why. The random string of numbers gives your brain twelve things to remember, which is more than its RAM can hold. The phone number, on the other hand, only has four things to remember, in this case four groups of numbers, making it much easier to remember.

What does this all have to do with fighting games? Combos my friend. Combos. Even the most basic Marvel 3 air-combo (L, M, H, S, jump, M, M, H, S) breaks the brain’s RAM limit. However, to compensate we use chunking to group series of inputs together as individual items to remember. In psychology, Chunks in motor learning are defined as everything that takes place between pauses in successive action. The super jump cancel in this case is the pause, and so we think of LMHSjump and MMHS as two separate chunks. So how can we use this to our advantage?

Well first of all, never try to learn a combo that breaks the brain’s RAM limit all at once. This is failing proposition from the get go. Instead, identify where the pauses are in a combo string. Jump cancels, slow and easy links, wall and ground bounces, and supers or other cinematic actions like command hits/throws are all areas where you experience a pause in your inputs, and these are the areas the brain is most likely to develop chunks. Practice these areas individually, and your brain will develop muscle memory far quicker than if you were trying to learn them all at once.

Secondly, identify areas that aren’t chunks and treat them as such. When the brain attempts to recall an input that it doesn’t actually have in working memory, the result is panicked mashing because the brain doesn’t actually have a muscle memory queued up to recall. So no matter how perfect your DP execution is normally, you’ll still screw it up a few times when you integrate it into a combo.

You probably had an experience like this when you were first learning how to FADC. You would perform a move, input the FADC command, and then mash the next move out like crazy hoping that it hit, and this probably didn’t work too well. This is because an FADC isn’t a chunk, and you are treating it as if it were, i.e. you were treating the FADC itself and the move that came after it as two separate mental objects. There is no pause in inputs from the move that started the FADC, the FADC itself, and the move that follows the FADC, and so the whole command must be treated as one smooth movement in muscle memory. Similarly, canceling a special into a super is also one fluid movement, and should be treated as its own chunk. By trying to think of each move individually, you will simply reduce your chances of hitting both.”

Great stuff sir, looking forward to the next part.