Tuesday, January 12, 2021

How to Overcome Skill Plateaus in Smash (Or Any Other Skill)

It’s the winter of 2007. It's snowing outside, and I'm on winter vacation. Thirteen year old me boots up a clunky Windows XP desktop to enjoy a full day of Runescape. I log in to find myself surrounded by 2,000 other players standing in the middle of Falador Park. This is world 2 - where all Runescape players went to buy or sell items. I take a deep breath and I begin to furiously mash on the keyboard. 

Flash2:Selling nats 200 ea - JJROCKETS

Flash2:Selling nats 200 ea - JJROCKETS.

Flash2:Selling nats 200 ea - JJROCKETS. 

In a crowd of this size, the only way to be heard (seen) is to spam your sale in the chat over and over as fast as possible. If you can’t keep up with the others your message is drowned out in a sea of faster typists. For those that didn’t have the pleasure of experiencing this firsthand growing up this might sound like an exaggeration, but I assure you it is not. It is through this trial of fire that I learned to type as fast as I do today.

That was almost 15 years ago. Since then I’ve written God only knows how many messages  on AIM, IRC, and Facebook chats. I’ve gone through all of high school and college and typed more papers and essays than I can count. I wrote blog posts, matchup notes for every character in Smash 4, embarrassing posts on forums. I’ve done it all. Yet in those 15 years, my typing speed has not improved whatsoever. Why is that? Today we explore what is known as the OK plateau - the reason my typing speed isn’t improving. The reason most people don’t get any better at driving as they grow older. And most importantly - the reason you aren’t getting better at Smash.

The "OK Plateau"

In 1967, two psychologists named Paul Fitts and Michael Posner set out to answer this question of why skill levels tend to plateau after a certain point. They showed their findings in a three phase model of skill acquisition.

From Furley and Memmert (2010), The Role of Working Memory in Sport

The first phase of their model is the cognitive phase. This is the phase where you start with a completely blank slate. You are first just acquiring information and trying to understand how the skill is performed. In a context of Smash or fighting games, this would be where you first learn how to play. You learn how the characters can attack, how the characters can move, how you can shield and dodge attacks. This is the absolute basics.

In the cognitive phase, you go from knowing nothing to having a decent grasp on the game fairly quickly as your brain starts to figure out how the game is played. I like to think about this like leveling up in an RPG. As your level increases, the amount of experience required to reach the next level increases. Low levels = faster level ups

As you get better at a skill, your improvement rate slows down.

Next is the associative phase. At this point the problem to be solved moves from gathering information to actually putting that information into practice. Now you are building muscle memory for complicated button inputs. You learn combos, have cleaner movement, and start to have full control of your character. Improvement slows down now, but it is still consistent and you have a clear path on what you can learn or practice next.

Eventually you reach a certain point where there is no obvious next step. You have full control of your character. You can execute bread and butter combos with ease. Nothing in the game surprises you anymore. At this point the experience points, or effort, required to increase your skill level becomes much higher. Your brain unconsciously decides “I’m okay with how good I am.” and it pushes your play from consciousness to unconsciousness.

This is the third and final stage that Fitts and Posner called the autonomous stage. From a biologic evolutionary standpoint, this shift to unconsciousness makes sense. When it comes to a matter of survival, your brain is better off using energy on learning new and unfamiliar things rather than marginally improving at what you are already capable of. Why waste mental energy on going from level 98 to 99 fire making when you could go from level 70 to 80 hunter with the same effort? When you reach this level that your unconscious mind puts the game on the back burner, you have reached the OK Plateau.

Overcoming the OK Plateau

Ok, you understand the OK Plateau. Now what? How do you escape from the plateau and speed up improvement again? Improvement in any skill in the autonomous phase is much harder and much slower than improvement in the cognitive phase. In order to break past the plateau you must revert back to the cognitive phase and stop yourself from auto piloting.

Joshua Foer, a science journalist who first coined the term "OK Plateau," summarized the findings of many case studies on experts across various fields into a four step strategy to stay out of the autonomous stage and to turn off auto piloting.

The first step is to practice outside your comfort zone and to study your failures. Let's go back to the typing speed example to see how to apply this. Most people have a comfortable typing speed that they tend to type at all the time with very little variation. People stay in that comfort zone because attempting to type any faster will result in more mistakes and ultimately slow down the speed at which they type the entire message. However, if you stay in that comfort zone you will never learn to type faster without mistakes.

You want to practice in the area just outside the reach of your abilities.

Foer suggests that to improve your typing speed you should consciously type at a speed above your normal and pay close attention to where you are making mistakes. As you spend more time typing at faster speeds speeds, you will start to notice patterns on what is causing your mistakes and you can focus your practice on correcting those.

Step two is to watch and analyze experts in your field. Luckily for fighting games and esports players this is incredibly easy nowadays. With the prevalence of tournament streams and VODs on Youtube, anybody can find a plethora of videos of top players of their game.

When watching top players you should be asking yourself Why? Why did they make the decisions they did? Why did they use that attack in that one position? What would have happened if they did something different? Learning how to learn from better players' VODs is a skill itself, one that I don't have time to cover in this video. Later in the series I will do a more in depth look at how to study VODs, but for now just know that it's best to stick to asking "why?"

Next on the list is that you need to have immediate and constant feedback to adjust your play. This is actually a trap that many Smash players fall into. They see the results of the game as the constant feedback, but I would argue that game results aren't consistent. When you play a match you only have control over 50% of the game. You could play the best game of your life, but if your opponent is playing just a little bit better they will win anyways. Looking at just the result of the game as a loss doesn't tell the whole story.

Instead of focusing the result of the game, you should concentrate on what are called SMART goals. This stands for specific, measurable, achievable, relevant, and time bound goals. Let's break down each part to show how to create an effective smart goal.

S-Specific - When you are trying to improve at something, setting a specific goal will help you achieve that. Something like "Get better at Smash." is so broad that it won't be of any use in your practice. Instead try to narrow down one specific aspect of the game that you want to focus on at a time, for example ledge trapping, combos, juggling, or escaping from disadvantage.

M-Measurable - Even if you have a specific goal like "Get better at ledge trapping", it is still difficult to measure whether you succeeded in achieving the goal. You must find a way set your goal such that you can measure exactly how successful you were. Now we can further adjust our goal to "cover the opponent's ledge jump with a Z-drop spike." (Flash info image of what Z-drop spike is)

A- Achievable - Your goal must be achievable. Something like "take every stock ever with Z-drop spike" is obviously not going to happen. You should try to avoid setting goals that are too difficult or unrealistic to achieve.

R-Relevant - This one is simple. Is your goal relevant to the skill you are trying to develop? For our example we want to improve at Smash, so our goal of landing z-drop spikes is very relevant.

T-Time bound - Last, your goal should be time bound. This is important for getting immediate feedback as well as ensuring that you aren't wasting your time trying to achieve a goal that isn't as achievable as you thought. Now we will further narrow down our goal to "Cover the opponent's ledge jump with a Z-drop spike in this best of five set."

Using this SMART goal, I guarantee you that by the end of that best of five set you will at least be better at finding situations that you can use the Z-drop spike. By focusing on this one aspect of the game during that set, you will be forced back to the cognitive phase where you are constantly thinking about ways you can achieve the SMART goal. In addition, you won't be tunnel visioning on the result of the games. In practice it is more important to work on expanding your skill set than it is to win every single game. With constant improvement, the results will soon follow.

Finally the last point is that you want to treat the skill you are learning like a science. You want to collect data, analyze it, and create and test theories. Try to figure out what the right options are for every single interaction in the entire game. I personally have experience doing this in Smash 4 where I kept and maintained match up notes for every character in the game that I could reference right before playing an important match. Many other top players do the same; the most well known example is probably Dabuz and his laptop.

To demonstrate the effectiveness of putting all these strategies together, I spent one week practicing typing to try to overcome my OK plateau in my typing speed. At the start of the week, I took an average from 3 typing tests online to find my base WPM. Throughout the week, I spent approximately 30 minutes per day practicing typing above my usual speed. Each day I took another average of 3 tests to see if my practice was working. By the end of the week, typing at my original 76 WPM average felt incredibly slow and I was making far fewer mistakes when typing at that speed. My final average was 92 WPM, a 20% increase over a one week period after being stagnant for 15 years.

In one week I saw a 20% increase in my typing speed after not improving my typing speed for 15 years.

Obviously this isn't an exact science and there were many sources of error in my experiment, but I think it does decent job demonstrating that with the right changes to the way you practice that you can overcome skill plateaus with relative ease. By studying your failures and practicing outside of your comfort zone you can start to expand the reach of your ability. Watching and studying other experts will help you see which areas you can still improve in, and getting immediate and and constant feedback while you practice will allow you fine-tune your skills. Finally by treating the skill like a science you can round out your skill set.


  1. As the gaming style is a Satta Matka style which is a banned game system in India. fatafat

  2. It s a very useful page. Thank you. 0df116c03debfff8291f082571084066

  3. Congratulations on your article, it was very helpful and successful. 36b68f4d8f7b9282a5b56543f611b1c9
    website kurma
    sms onay
    numara onay

  4. Thank you for your explanation, very good content. b6a3394833a75bcc25e88a5ce96b48f8
    define dedektörü

  5. This allowed the system to manage the extent of payout by stopping the drums at positions it had determined. If the payout channel had filled up, the payout grew to become extra generous; if nearly 카지노사이트 empty, the payout grew to become much less so . With microprocessors now ubiquitous, the computers inside modern slot machines allow manufacturers to assign a unique chance to each symbol on every reel. To the player, it'd seem that a profitable symbol was "so close", whereas actually the chance is way decrease. A computerized model of a basic slot machine, video slots tend to incorporate particular theming components, such as themed symbols, as well as|in addition to} bonus games and additional methods to win.

  6. Thank you for your explanation, very good content. c8162bce28a84a0d3e353b278c284a62
    evde iş imkanı